The debate is over. It is empirically settled: Cantwell v Connecticut is the most influential United States Supreme Court case of all time; Branzburg v Hayes is the most well-grounded. So says a recent paper by Professor James F. Spriggs II and his coauthors entitled Network Analysis and Law: Measuring the Legal Importance of Precedent at the U.S. Supreme Court. Spriggs addressed the Workshop on Judicial Behavior Wednesday evening to discuss the paper in which he and his coauthors apply network analysis (typically used for computer science and sociology) to legal opinions. Their aim: to identify “the most legally central” Supreme Court cases.
Network analysis studies the relationships between things, be it people, information, or computers. Analysts study the connections (“links”) between the things (“nodes”) to determine the centrality or prestige of nodes within the network.
Spriggs’ paper used Supreme Court opinions as nodes and citations within opinions as links. Using this framework, the paper coded “outward citations” (all precedents a given case cites) and “inward citations” (all the opinions that cite a given case). Using this technique, the paper created a network of all Supreme Court majority opinions between 1791 and 2005.
The paper then classified opinions along two dimensions: "outward relevance" and "inward relevance." An outwardly relevant case “cites many other relevant decisions, thereby helping to define which decisions are pertinent to a given legal question.” An inwardly relevant case “is one that is widely cited by other prestigious decisions, meaning that judges see it as an integral part of the law.” These relevance scores measure more than the number of direct citations (known as degree centrality) because they also take into account the indirect linkages between cases that Spriggs says influence how embedded a case is in the legal network. In other words, relevance scores are not based on the absolute number of links; they also incorporate the inward or outward relevance of the cases on either side of those linkages. As a result, these relevance scores are dynamic and can thus change over time for each case. For example, the paper calculated inward relevance based not only on the absolute number of citations a given case has received, but also whether the cases citing a given case were themselves inwardly relevant. This intuitively makes sense: being cited by an important case should receive greater weight than being cited by an unimportant one.
Although workshop participants were generally accepting of inward relevance as a measure of legal prestige, they were less receptive to the idea of outward relevance as it related to Supreme Court decisions. It seems odd that a justice (okay, law clerk) crafting an opinion could increase the opinion’s legal relevance (or prestige) simply by citing to other Supreme Court cases that enjoy high inward relevance. As one participant noted, the Supreme Court could issue an opinion (Opinion Z) that no one ever read, talked about, or cited. But if Opinion Z cited a large number of prestigious cases (Opinions A-Y), then Opinion Z could still enjoy a high level of outward relevance. Furthermore, Opinion Z’s outward relevance would grow over time as new opinions cite Opinions A-Y.
The paper also used some questionable terminology to characterize inward and outward relevance. The paper refers to cases that are inwardly relevant as “influential," and cases that are outwardly relevant as “well-grounded." Although participants generally accepted the former characterization, they challenged the latter. Of all the Supreme Court opinions ever published, Griswold v Connecticut and Buckley v Valeo have two of the highest levels of outward relevance. To call these two of the most legally well-grounded cases ever decided is a proposition that many legal academics would certainly contest. One participant noted that it is plausible that Supreme Court opinions cite a large number of frequently cited precedents when they are on weak, rather than sturdy, legal ground. Spriggs conceded that perhaps “deeply connected” would have been a better characterization because he was referring to these cases’ entanglement within the legal network, rather than their legal basis.
Other participants were skeptical of what inward and outward relevance really demonstrated. Spriggs is currently studying the effect of inward and outward relevance on how precedents lose value over time (known as depreciation). According to Spriggs, his research shows that early in a case’s life, how outwardly relevant the case is has tremendous effect on the rate at which the case depreciates.
Perhaps participants were resistant to the idea of outward relevance because it is a network analysis tool that doesn’t seem readily applicable to a legal context. For instance, if one were studying a social group, outward relevance would clearly be a sign of importance: who I can get on the phone says something about how powerful I am to the same, if not greater, extent as who calls me. This simply isn’t the case in the legal network: citing to influential, landmark cases is “free” – any opinion can do it. As a result, the fact that Opinion Z cited Opinions A-Y does not show that Opinion Z is important: it shows that Opinion Z is a name-dropper. According to one participant, Google (which uses network analysis as part of it search engine) has experienced problems based on the way it weighs outward relevance. Web sites learned how to game the system: by linking to a large number of prestigious web sites, unknown web sites could rocket their names to the top of Google’s search results. This suggests that when outward relevance is self-determined and “free,” its value as an indicator of importance or prestige may be undermined.
Nevertheless, if Spriggs’ research is correct, and precedent “name-dropping” causes cases to depreciate slower, future SCOTUS clerks may want to take note.