1 The term “law review,” used throughout this study, has no precise general meaning. Sometimes it is used loosely to identify all learned legal journals, sometimes it refers only to law journals published by law schools. I use it in the latter sense.Google Scholar

2 Harvard: Harvard Law Review, Harvard International Law Journal, Harvard Journal on Legislation, Harvard Civil Rights–Civil Liberties Law Review.Google Scholar

Chicago: University of Chicago Law Review, Journal of Law and Economics, Supreme Court Review, Journal of Legal Studies.Google Scholar

Source: November 1972 issue of the Index to Legal Periodicals.Google Scholar

3 This number of titles was indexed, according to the list given in the November 1972 issue of the ILP. The number may have varied slightly during the year because titles are occasionally added to or dropped from the ILP.Google Scholar

4 Textual material, as used in this paper, refers to pages in periodicals that are numbered with Arabic numerals and does not include preliminary pages, indexes, tables of cases, and the like that are generally identified by Roman numerals and letters of the alphabet. This is not a perfect measure because Roman numerals occasionally run into tables of cases and other such matter. However, these deviations from the usual practice are uncommon. The 278 titles comprise the sample used in this study. Note 21 gives a detailed description of it.Google Scholar

5 As far as the law reviews are concerned, it has been said that “[w] hereas most periodicals are published primarily in order that they may be read, the law reviews are published primarily in order that they may be written.” Harold C. Havighurst, Law Reviews and Legal Education, 51 Nw. U.L. Rev. 24 (1956). Another commentator agrees that the principal value of law reviews is for the students who work on them, but then goes on to observe:Google Scholar

[I]t has not been satisfactorily explained why it is any more necessary to publish the product of a student legal writing program than to publish the briefs written for intramural moot court competitions. It is difficult not to conclude that publication of a law review is in truth a quest for prestige in a form socially acceptable in the law school world. Prestige is derived from an unprestigious publication by operation of the elaborate inter-school protocol that precludes calling a spade a spade.”Google Scholar

Geoffrey C. Hazard, Jr., Research in Civil Procedure 135 (New Haven CT: Walter E. Meyer Research Institute of Law, 1963).Google Scholar

If law reviews are indeed primarily instruments of education and indicators of status, then it is understandable why there are so many of them and, perhaps, why they tend to be bulky.Google Scholar

6 Approximately 61 percent of the titles in the sample used in this study are law reviews. Their share of the pagination is about 68 percent. The sample is described in note 21.Google Scholar

7 Stanley H. Fuld, A Judge Looks at the Law Review, 28 N.Y.U.L. Rev. 918 (1953).Google Scholar

8 William O. Douglas, Law Reviews and Full Disclosure, 40 Wash. L. Rev. 227 (1965).Google Scholar

9 Id. at 228.Google Scholar

10 Hazard, supra note 5, at 60.Google Scholar

11 For a sample of such work sec: Association of American Law Schools, 1954 Proceedings 147 (library standards); Gerhard O. W. Mueller and Jerome H. Skolnick, Bar Reactions to Legal Periodicals: The West Virginia Survey, 11 J. Legal Ed. 197 (1958) (evaluation of one title); Cameron Allen, Duplicate Holding Practices of Approved American Law School Libraries, 62 L. Lib. J. 191 (1969) (budgeting). What is perhaps the most ambitious effort to evaluate the law reviews was published as The Law Review–Is It Meeting the Needs of the Legal Community? 44 Denver L.J. 426 (1967). The basic purpose of the latter study was to find out what the content of law reviews should be to make them optimally useful to the profession. On the basis of a 40 percent return from a national random sample of 1,000 judges, lawyers, and law professors, 102 law reviews were ranked according to a 100-point scale of “helpfulness.” More interestingly, the study then grouped law reviews according to median LSAT scores of entering students at the schools where they were published. A significant correlation of the rankings on the “helpfulness” scale and rankings by LSAT score was observed. (The extent of the correlation is not given.) If it is granted that LSAT is a valid measure of a law school student body's quality, then here is some proof that the quality of a law review is directly related to the quality of the law school student body where it is produced.Google Scholar

12 Consider the following as examples. (1) A survey of law review holdings in law school libraries (Allen, supra note 11) shows that in 1968, on the average, 4,584 copies of each issue of Yale Law Journal and 6,274 copies of each issue of Georgia Law Review were distributed through subscriptions or otherwise. Does this mean that Georgia Law Review was used substantially more than Yale Law Journal? (2) A questionnaire survey (Mueller and Skolnick, supra note 11) states, on the basis of a 20 percent return from all West Virginia attorneys of record, that 24 percent of the lawyers who returned the questionnaire “read” the West Virginia Law Review from cover to cover and that an additional 62 percent “usually read” at least one contribution. The text of the study does not say whether reference is to a single issue, to a volume, or to some other counting unit. But more important, one wonders how the respondents interpreted the word “read” and how they compare with the 80 percent who did not return the questionnaire, not to mention the effect of such bias as overreporting of “reading” by those who felt a need to put themselves in a good light as learned lawyers. On the prevalence of this type of bias see generally Derek L. Phillips, Abandoning Method (San Francisco: Jossey-Bass, 1973).Google Scholar

13 I am not suggesting that citation frequency is an absolute or even very precise measure of worth. Clearly it is not, if only for the reason that not all legal periodicals have the same objectives. If law reviews, generally, attempt to provide scholarly, critical analysis of legal issues, then bar association journals as a rule serve as a forum for the exchange of ideas about the bar and the issues facing it. Of course they also contain articles on law, but these tend to be of the how-to-do-it kind. Law reviews do not cite many bar journals, and bar journals do not include many citations to law reviews or anything else. Commercial legal periodicals print mostly “practical” material and do not cite many periodicals. A tax practitioner may very properly insist that some commercial journal that keeps him current on tax law is to him more valuable than any number of prominent law reviews. In addition to this it is surely reasonable to assume that there is some indiscriminate citing and that an article can have impact without being cited.Google Scholar

14 In the present context it is very important to draw a line between citation as a quantum of use and citation as something that is to be analyzed and weighed according to its significance and impact in particular instances. I suggest that it is very nearly impossible to weigh the value of particular citations with any degree of objectivity. An interesting attempt to do this is found in Chester A. Newland, The Supreme Court and Legal Writing: Learned Journals as Vehicles of an Anti-Antitrust Lobby? 58 Geo. L.J. 105 (1959). Newland concludes that law journal articles had influenced some Supreme Court opinions that he reviewed. In a more general way he observes that “an examination of the opinions and the sources cited usually reveals some interaction between the reasoning of the Justice and the author of the article in question; but the impression produced in most cases is that the Justice cited law reviews to support a view which he had arrived at more or less independently of the reasoning in the source referred to.”Id. at 143. I am not quite convinced by Newland's proof of the influence of certain articles and find it easier to agree with his more general conclusion. In fact, the reading of Newland's article reinforced my own view–that it is very nearly impossible to weigh the impact or influence of a particular article on a writer unless the writer provides a simple declaration to the effect that he had no idea what he was about until he came upon that particular article. I would be very much surprised indeed ever to come upon such a declaration.Google Scholar

15 See Eugene Garfield, Citation Analysis as a Tool in Journal Evaluation, 178 Science 471 (1972).Google Scholar

16 Comment, Legal Periodicals: Their Use in Kansas, 7 U. Kans. L. Rev. 490 (1959).Google Scholar

17 Douglas B. Maggs, Concerning the Extent to Which the Law Review Contributes to the Development of the Law, 3 So. Cal. L. Rev. 181 (1930); Chester A. Newland, Legal Periodicals and the United States Supreme Court (published doctoral thesis in the University of Kansas Government Research Center, 1958).Google Scholar

18 Quoted material is from the title page of the 1974 volume of the Shepard's Law Review Citations.Google Scholar

19 For an example of this see id. at 331.Google Scholar

20 See p. 228 supra.Google Scholar

21 A publication year generally corresponds to a “volume.” Some journals, however, do not use volume numbering and merely publish several issues each year. This is the reason for the use of the more awkward description “publishing year.”Google Scholar

Whenever possible, I used volumes or issues that were dated 1972. When this was not possible–because of a fall to spring publishing schedule-I used the 1972/73 volume or issues. In some instances it was necessary or desirable to go back to 1971/72. A few titles, for example, were so far behind schedule that their 1972/73 volume was not completed by late summer of 1974. However, it is accurate to say that almost all issues in the sample are dated from early 1972 to the middle months of 1973.Google Scholar

The November 1972 issue of the ILP lists 285 American titles. The present sample includes 278 of them. The following titles were not included because they had ceased publication on the dates given in parentheses: Antitrust Law Symposium (1971); Idea: The Patent, Trademark & Copyright Journal of Research and Education (1971/72); Inter-American Law Review (1966). Publishing, Entertainment, Advertising, and Allied Fields Law Quarterly and the University of Chicago Law School Record were not included because virtually all material in them, and all material with citations, were reprints from indexed titles. Federal Rules Decisions was left out because it is cited both for its court opinions and articles and it would have been very time consuming to attempt to differentiate between the two during counting.Google Scholar

Finally, Revista de Derecbo Puertorriqueno had to be dropped because I could not locate a complete volume for the required period in Chicago.Google Scholar

The sample includes some titles that are not listed in the November 1972 issue of the ILP. This may seem odd until one remembers that periodical titles change, journals are absorbed by other journals, and publications are discontinued or suspended and then resume publication again. Here, what appear to be earlier forms of a current title were considered to be such only if a clear, formal continuity existed between them. Two examples make this clear. Wyoming Law Journal, published by the University of Wyoming College of Law, ceased publication in 1965. In 1966 volume 1 of Land and Water Law Review was published by that school. Citations to the Wyoming Law Journal were not counted. Beginning with volume 43, the University of Detroit Law Journal became Journal of Urban Law. Citations to University of Detroit Law Journal were included with citations for the Journal of Urban Law.Google Scholar

22 In addition to footnotes, “references” and “notes” attached to the ends of articles were counted, provided they were numbered like footnotes to refer to specific loci in the text. Reading lists and bibliographies were excluded.Google Scholar

Short form citations (“ibid.”“supra,”“Chicago Study”) and cross-references (“see note 22 supra”) were not included. Their inclusion would have increased the cost of the project enormously. Without backtracking it was possible to do an average law review issue in about 20 minutes. It would have taken at least twice that time if we had gone back to see what “note 22 supra” referred to and so forth.Google Scholar

23 American Bar Association Journal, Business Lawyer, American Journal of International Law.Google Scholar

24 Journal of Taxation, New York University Institute on Federal Taxation, Taxes, Antitrust Bulletin, Labor Law Journal.Google Scholar

25 The citation counts are also surely overweighted in favor of older journals. It is clear that a title that is 1 year old has less chance of being cited than one that is 121 years old. (These figures represent the age range of titles in this study.) It is, however, very difficult to adjust for this. The principal reason is that older articles are cited significantly less often than newer ones. Age of citation is analyzed in some detail on page 247. Here it is sufficient to observe that about 75 percent of citations in this study are to materials not over 10 years old and about 10 percent are to those over 20 years old. The average age of titles by impact group is: high impact–56 years; medium impact–32 years; low impact–21 years. While this suggests that there should not be, on the average, substantial overweighting of the citation count in favor of any impact group, it is important to note that no title in the high-impact group is less than 20 years old while in the low-impact group 80 titles are not over 10 years old and of those 39 are not over 5 years old.Google Scholar

Finally, it is important to keep in mind that the rankings presented in tables 1 and 3 are based on a one-year sample and are thus subject to any errors that may introduce. A different sample year, for example, 1971 or 1974, may have produced somewhat different rank orders. The point is that no significance should be attached to close adjacent rankings. Indeed, it would be prudent to view all rankings except the ones where numerical disparities are large with some skepticism.Google Scholar

26 The term “impact factor” is borrowed from Garfield, supra note 15, at 474, where it is used somewhat differently.Google Scholar

27 For definition of “textual material” see note 4 supra.Google Scholar

28 Harvard Civil Rights-Civil Liberties Law Review had been published for only seven years. Its impact factor, based on seven years' pagination, is obviously not comparable with the others. While the extent of distortion in it is of course impossible to measure, it may be noted that if it were computed on the basis of actual pagination for nine years (1966–74) the journal would rank 17 instead of 16 in table 3.Google Scholar

29 Note, however, that only the 4 bottom ranking titles of the high-impact group (23 titles)–Vanderbilt Law Review, U.C.L.A. Law Review, Iowa Law Review, and George Washington Law Review–are displaced from the top 23 titles when impact factor is used for ranking. None moves very far down: Thirty-second rank, taken by Iowa Law Review, is the lowest. We must keep in mind, of course, that the impact groups were constructed arbitrarily. Thus no great significance should be attached to this observation.Google Scholar

30 One thing that stands out clearly in table 3 is that some of the most prominent upward movers in rank are modest-sized academic journals that do not publish student work: Supreme Court Review, 33 to 6; Law and Contemporary Problems, 15 to 8; Journal of Law and Economics, 50 to 12; Journal of Criminal Law, Criminology and Police Science, 28 to 13.Google Scholar

31 Self-citation index for each title is given in table 1.Google Scholar

32 It is perhaps worth noting that in the world of legal periodicals publication date printed on an issue does not necessarily mean that the issue was published on that date. It may have been published several months after it. Also, it is probably reasonable to assume that in many cases there is a time lag between reading and use of the material. These factors suggest that the age of citations, at least in the segment not over two years old, is overstated to some extent.Google Scholar

33 James P. Conant, 1 Harvard Case Histories in Experimental Science xv (Cambridge MA: Harvard University Press, 1957).CrossRefGoogle Scholar

34 Id.Google Scholar

35 Paul Weiss, Knowledge: A Growth Process, 104 Proc. Am. Philosophical Soc. 246 (1960).Google Scholar

36 Robert K. Merton and Harriet Zuckerman, Age, Aging, and Age Structure in Science, in R. K. Merton, The Sociology of Science 509 (N. W. Storer, ed.; Chicago: University of Chicago Press, 1973).Google Scholar

37 Id. at 510.Google Scholar