Landes & Posner’s 1976 article Legal Precedent: A Theoretical and Empirical Analysis is a path-breaking study in two ways. Drawing on the analysis of capital formation and investment, they were the first to offer a realistic theory of precedent. To them, precedents produce a flow of information that decreases in value over time as society evolves. Legal Precedent is also an early example of empirical scholarship in the law schools. To measure the efficacy of a precedent, they developed the-then novel strategy of counting the number of times judges cite to a decision. Applying this approach to samples of federal court decisions they found that precedent depreciated over time.
Unlike the precedents in their study, Landes & Posner’s paper hasn’t seemed to depreciate much. In just the last year or two, it has been cited in studies on labor law, judicial behavior, corporate law, and behavioral economics; and in journals as diverse as Theoretical Inquiries in Law, International Review of Law & Economics, the Journal of Evolutionary Economics, and the Annual Review of Political Science.
Black & Spriggs’s new article, The Citation and Depreciation of U.S. Supreme Court Precedent, could be seen as just another of the many studies that, as they write, “owes its origin” to Landes & Posner—just another citation. But that characterization would be a mistake. After all these years, Black & Spriggs break new ground.
The question they ask is not just whether precedent depreciates—it surely does, as the many studies in the wake of Landes & Posner have confirmed. The question is why there might be variation in depreciation rates. To provide a quick example from Black & Spriggs’s dataset: The Supreme Court decided both United States v. Riverside Bayview Homes and Cornelius v. Nutt in 1985. The Justices continue to cite Riverside Bayview Homes but not Nutt. Why?
Landes & Posner offered some hypotheses, and other scholars have followed suit. Frank Cross speculated, and political scientists tend to concur, that ideological considerations led to depreciation of Warren Court precedents during the Rehnquist years. Many writers, including the Justices themselves, have intimated that the value of constitutional precedent, relative to most others, should decline at a faster rate. And still other scholars have suggested that precedent established in especially salient cases might depreciate more slowly.
Studies have tried to assess some of these ideas but none with the level of rigor, sophistication, and breadth as this article. Black & Spriggs test the various hypothesis using all 6,299 Supreme Court decisions (the “precedents” in the study) issued between the 1946 and 2004 terms.
The results are really interesting. First, Black & Spriggs continue to validate Landes & Posner’s basic finding: Supreme Court precedents don’t have an especially long shelf life: they depreciate by about 80% between years one and twenty. Interestingly, though, much of the depreciation occurs within the first couple of years. Within five years of the Court issuing a decision, the predicted probability of citation decreases by 25%. Between fifteen and twenty years from decision, the probability declines by only 4%.
The second and especially notable result is that almost none of the hypothesized factors affect the rate of depreciation—and those that are significantly related have a very small effect. To provide but a few examples: Although there are good reasons to suspect that statutory precedents should depreciate at a slower rate than constitutional precedents, the results show very little difference between the two. Ditto for the political scientists’ emphasis on ideology. There is virtually no difference in depreciation rates based on the ideological distance between the deciding and sitting Court.
Only variables that tap the citation history of precedents produce statistically and substantively significant results. The more the Court interprets a decision and the more central a precedent in the network of Supreme Court law, the slower the depreciation. But even these effects tend to disappear a decade or so after the Court issues its decision.
What are we to make of these findings? First, Landes & Posner were right—and right without qualification. Precedent (almost always) depreciates, period. Second, because of its “here today, gone tomorrow” quality, law professors and lawyers might (re)consider carefully the cases they emphasize in class and in the courtroom. If currency is a virtue for at least some judges, maybe it should be for them too.