This Essay updates two well-known earlier studies (dated 1985 and 1996) by the first coauthor, setting forth lists of the most-cited law review articles. New research tools from the HeinOnline and Web of Science databases now allow lists to be compiled that are more thorough and more accurate than anything previously possible. Tables printed here present the 100 most-cited legal articles of all time, the 100 most-cited articles of the last twenty years, and some additional rankings. Characteristics of the top-ranked publications, authors, and law schools are analyzed as are trends in schools of legal thought. Data from the all-time rankings shed light on contributions to legal scholarship made over a long historical span; the recent-article rankings speak more to the impact of scholarship produced in the current era. The authors discuss alternative tools and metrics for measuring the impact of legal scholarship, running selected articles from the rankings through these tools to serve as points of illustration. The authors then contemplate how these alternative tools and metrics intersect with traditional citation studies and how they might impact legal scholarship in the future.
TABLE OF CONTENTS I. PREVIOUS STUDIES AND RATIONALE (SHAPIRO) II. CURRENT METHODOLOGY (SHAPIRO) III. ANALYSIS (SHAPIRO) A. The Effect of the Social Sciences on Legal Citation Analysis B. Top Authors, Top Law Reviews, and Top Schools C. Reflections IV. COMPARING SHAPIRO'S LISTS WITH MODERN METHODS (PEARSE) A. The Rise of Alternative Methods and Metrics in Legal Scholarship B. Alternative Methods for Tracking References in Published Works 1. Comparison of Google Scholar's Results 2. Comparison of Microsoft Academic Search's Results C. Rise in Database "Cited by" (or "Times Cited in Database") Features D. "Real World" Impact on Law and Practice and Beyond E. Download or "Popularity" Metrics F. "Buzz" Metrics V. LIMITATIONS OF CITATION METRICS (PEARSE) VI. THE FUTURE OF LEGAL SCHOLARSHIP AND CITATION METRICS (PEARSE) CONCLUSION (SHAPIRO AND PEARSE) I. PREVIOUS STUDIES AND RATIONALE (SHAPIRO)
This is the third in a series of studies that I have authored enumerating the most-cited legal articles--that is, the articles most often cited within other articles. (1) The two previous installments attracted considerable attention in both the legal community and the general media. Jack Balkin and Sanford Levinson wrote, "Fred Shapiro can lay claim to be the founding father of a new and peculiar discipline: 'legal citology.'" (2) The Wall Street Journal ran a front-page profile of me based on the citation rankings, (3) popularizing Balkin and Levinson's term "citology" to the point where Britain's Guardian newspaper included the term in a glossary of new words of the 1990s. (4) Herma Hill Kay, with tongue planted firmly in cheek, hailed my work:
Footnotes nowadays are not phony excrescences; they are the raw data used by the hottest new school of legal scholarship, the citation analysts. These bibliotechs have shown once and for all that nobody reads the text of other people's articles anyway. Anybody who is anybody in any field you care to name has already said the same thing in different words a dozen times before. There is nothing new under the sun. The only thing that is important is who cites whom. If you're cited, that means you're identified as a player in the game: a scholar of significance: I also published a more specific "most-cited" compilation listing the top thirty articles from the Yale Law Journal on the occasion of that law review's centennial. (6) Without claiming too much significance for citology, I described citology as more than a mere parlor game and as a potentially useful tool for studying the impact of scholarship:
Citation analysis is now extensively used by information scientists and sociologists to study the history and structure of the natural sciences and other disciplines.... ... Authors too have been evaluated through tabulation of citations to their writings. Citation counts have been utilized in assessing scholars' work for purposes of grant awards, tenure, or promotion decisions. Those using citation data for evaluative purposes have justified such use by pointing to research demonstrating a high correlation between the total of citations to a scientist's or scholar's writings and judgments by peers of the "'productivity,' 'significance,' 'quality,' 'utility,' 'influence,' 'effectiveness,' or 'impact' of scientists and their scholarly products" One investigator has gone so far as to say that "citations and peer ratings appear to be virtually the same measurement." Almost all citation analysts, however, are careful to note that citation counts measure a "quality" which is socially defined, reflecting the utility of the writing in question to other scholars, rather than gauging its intrinsic merit. Furthermore, the value of the counts may be lessened by limitations in the accuracy, coverage, or time-frame of the source data. For these reasons and others, evaluative use of citation analysis has remained controversial. Even with their acknowledged limitations, citation counts are attractive as relatively objective tools for assessing scholarly impact. They can be used not only to gauge the impact of a given author or writing, but also to identify which writings are the most frequently cited, taken to be a rough measure of the writings which have had the most extensive impact. (7) Both of my earlier studies used then-available tools to compile their rankings. The first compilation, published in the California Law Review in 1985, (8) relied on browsing through the print volume of Shepard's Law Review Citations and looking for long lists of citations. The second one, appearing in the Chicago-Kent Law Review in 1996, (9) employed searches of the online and print versions of the Social Sciences Citation Index.
Both of the prior studies also had limitations stemming from the coverage and functionality of their methodology. Because Shepard's Law Review Citations covered only citations since 1957 to articles published since 1947, the 1985 ranking excluded pre-1947 articles. Scholarship in interdisciplinary journals not covered by Shepard's was also excluded. The 1996 ranking drew on the more comprehensive data available in the Social Sciences Citation Index, which had no beginning date for cited publications and therefore encompassed older articles. The citing coverage did have a 1956 commencement date, however, so that older articles were still disfavored because pre-1956 citations to them were not counted.
CURRENT METHODOLOGY (SHAPIRO)
This third study benefits greatly from the development of online citators in law and in the social sciences. The limitations of past studies fall before the spectacular capabilities of the HeinOnline and Web of Science databases. HeinOnline, produced by the William S. Hein Company, includes the vast majority of the entire United States law review literature from the nineteenth and twentieth centuries. (10) I devised a search that retrieved virtually all of the over 1.4 million articles in that database (11) and then used the ability of HeinOnline to sort those articles by "Number of Times Cited" to generate citation totals that are both more thorough and more accurate than any previous counts.
I did not, however, take the HeinOnline totals as the final ranking of the most-cited legal articles of all time. HeinOnline, though wonderfully comprehensive in its coverage of law reviews published by law schools, omits some of the legal journals published by university presses, learned societies, and commercial entities. (12) More importantly, HeinOnline has only modest coverage of social science journals. In a legal academy that has become quite interdisciplinary, part of the definition of a legal article's influence seemingly should be its impact on scholarship outside of law. Therefore, it was necessary to add to a legal article's HeinOnline citing-number the total number of citations by social science journals to that article, resulting in a relatively complete count of law citations plus social science citations. (13)
The most balanced, precise, and structured source of citation data in the social sciences is Web of Science, Thomson Reuters' current version of the Social Sciences Citation Index. Web of Science covers 2,697 journals across fifty-five social science disciplines going back to the year 1900. I was able to search for all the articles classified in the "Law" category, sort the resulting articles by "Times Cited," and thus create a listing of the legal articles most cited in Web of Science. Looking at the articles near the top of that listing, I subtracted their citations in legal journals to avoid double-counting with the HeinOnline law citations. (14) The remaining Web of Science social science citations were then added to the HeinOnline totals to create the final totals used to rank the most-cited legal articles of all time.
The term "legal article" means not only articles published in traditional law reviews like the California Law Review and the Duke Law Journal but also articles appearing in "law and" journals such as the Journal of Law and Economics and even in purely social science journals like the American Sociological Review. For the "law and" journals and the social science periodicals, articles were designated "legal articles" if over 50 percent of citations to them occurred in law reviews or "law and" journals. As I have written before, "My theory in doing so was that I wanted to represent law-related scholarship as comprehensively as possible, and that a predominance of citations in law-related journals seems to identify an article as being law-related." (15)
Under this 50 percent rule, articles in "law and" journals by R.H. Coase, (16) Marc Galanter, (17) and George L. Priest and Benjamin Klein (18) qualified for the all-time top 100 most-cited...