Wednesday, September 04, 2013
Visualizing SCOTUS Doctrine - An Introduction
Hello Prawfs! It's Colin Starger from the University of Baltimore checking in. I'm a longtime Prawfs fan but first-time guest blogger. So many thanks to Dan and company for this wonderful opportunity.
During my time here, I'd like to introduce readers to the SCOTUS Mapping Project and explain how its method of visualizing Supreme Court doctrine works. I'll delve into the details of this software-driven project over my next few posts. By working through a couple of concrete examples of doctrinal maps, I hope to persuade folks of their practical, academic, and pedagogical value.
So what is a doctrinal map? To get the basic idea, consider Map 1 below. I explain what it means below the jump.
Doctrinal maps chart relationships between Supreme Court opinions. In the map above, the opinions are represented as triangles. The opinion's short name sits above the triangle and its author lies below. All maps are on a grid. The X-axis indicates the year an opinion was decided while the Y-axis shows the number of votes that opinion received. Map 1 thus shows that Justice Kennedy authored the majority opinion in 2013's Maryland v. King (short name = King) and that his opinion garnered five votes. The map also shows that in King, Kennedy directly cited to the 2012 Florence v. Board of Chosen Freeholders of County of Burlington (short name = Florence) majority opinion. As the map indicates, Florence was another Kennedy opinion that received five votes.
While Map 1 illustrates the basic schema, it is far too simple to be interesting. The maps become more useful when you create a chain of citations that track the origins of a certain legal proposition, concept, or perspective. Let's return then to Maryland v. King. As readers might recall, the majority in this case upheld against a 4th Amendment challenge a Maryland law that authorized the collection of DNA from defendants arrested for, but not convicted of, certain serious offenses. One of the justifications behind Kennedy's majority opinion was that the Maryland law was consistent with the Court's search-incident-to-arrest doctrine.
Map 2 depicts part of Kennedy's argument in King. Specifically, tt shows that Kennedy's claim that Maryland's arrestee DNA law comported with search-incident-to-arrest doctrine depended on his reading of Florence, which in turn cited to Belton for a key proposition about the doctrine, which in turn cited to Robinson and so on back to 1927's Marron -- and then back to the 4th Amendment itself.
A couple of quick technical observations. First, note the footnotes. Like all the maps, Map 2 was created using the custom SCOTUS mapper software program ("the Mapper"). The Mapper has a huge number of options -- one of them is to display text that gives the full names and cites of the cases depicted. I thus turned that option on when rendering this map. Second, note that the 4th Amendment "vote count" shown on the Y axis is essentially meaningless. The Mapper can plot non-SCOTUS opinion data points, but in this instance the Y value was based only on an aesthetic choice. I just thought it looked nicer.
One other substantive point bears emphasis: Map 2 is contestable. It was not generated by a computer alogorithm. Rather, it represents my interpretation of the doctrine's genealogy based on my study of the cases and their internal citations. Others might disagree with my rendering of the search-incident-to-arrest doctrine relied on by Kennedy in King. The point is that they could represent this disagreement by creating a different map that emphasizes different opions and connections. In this way, the Mapper is a flexible tool rather than a rigid method.
At the same time, there is a method that I believe works well when attempting to visualize SCOTUS doctrine. This method basically involves paying very careful attention to dissents. Dissents, I argue, are key. They can reveal much about the competing traditions on the Court. So before ending this introductory post, I want to improve on Map 2 by introducing dissents into the picture.
In this final map (for now), I posit that Kennedy's majority opinion in King is connected in a meaningful way to Justice Alito's dissent in Gant as well as to Justice White's dissent in Chimel. However, I use dotted arrows to represent these connections because neither Kennedy in King nor Stewart in Belton directly cited to the respective dissents in Gant and Chimel. The "citations" to the dissents are thus "implied" -- implied, that is, by me. In my view, the 4th Amendment interpretation urged in those dissents form part of a consistent doctrinal tradition that came to rule the day in King.
At this juncture, there is no need to dive into the actual 4th Amendment doctrine represented above. (For those interested in this particular doctrinal debate, I explore King -- using even more complex maps -- here and here). Instead, I wish only to emphasize the critical role that dissents play in the SCOTUS Mapping Project.
In my next post, I'll pick up on the theme of dissents and explore the idea of competing traditions in constitutional adjudication. Meanwhile, I welcome comments or questions. Please stay tuned!
Posted by Colin Starger on September 4, 2013 at 09:27 AM | Permalink
TrackBack URL for this entry:
Listed below are links to weblogs that reference Visualizing SCOTUS Doctrine - An Introduction:
Colin, thanks for the post. I'm looking forward to the rest. A few questions:
1) Your mapping shows relationships among Supreme Court decision, but it puts cases along the same chart even though they involve very different courts, generations, and facts. I worry that putting those cases along the same chart creates a misimpression that the decisions are comparable just because they involve the same general body of law.
2) Your mapping approach places significant emphasis on how many votes a majority decision received; it also makes dissents very important. Can you say more about why you think those features of opinions are that important? Especially given that the charts include cases from different courts on different aspects of the law, I'm not sure why the number of votes or the dissents are critical to the picture.
3) Finally, I'm not sure I agree with the emphasis on what is actually cited (or not actually cited but impliedly relied upon). Perhaps I am too much of a legal realist, but what is cited could just be a matter of what a law clerk chooses to cite, or what the briefs from the winning side cited; it may just be window dressing rather than a reflection of an actual link from case to case. Can you say more about why you see citations as significant?
Posted by: Orin Kerr | Sep 4, 2013 12:42:04 PM
Orin: Thanks for the excellent questions. I'll try to answer them in a deeper way over my next few posts. For now, however, let me offer some brief thoughts.
1) Certainly, I agree that facts vary from case to case. Yet the Court remains the Court even as generations pass. Precedent cannot be ignored on the sole grounds that it is old. Of course, I agree that the Court changes over time and so it very much matters who is sitting at any given moment. This is precisely why the maps draw attention to the authors of opinions displayed. The dominant metaphor here is genealogical. They do not suggest that the decisions "in a line" are the same -- but rather that they share familial relationships. The later cases descend from the earlier. At least, that is the argument advanced by the map's author.
2) The number of votes and existence of dissents matter for the same basic reason -- they reveal a split of Court opinion on what they law is/means/requires. My basic thesis is that these differences of opinion are not one-off occasions, but rather reflect deeper splits in doctrinal commitments between Justices. There are, in short, competing schools of doctrinal interpretation. The competing schools clash from case to case even as the facts change. The trick, as I see it, is to map out the familial lines of the competing schools. I'll try to demonstrate this in my next post.
3)I am a realist too and agree that what is cited cannot be the whole story. Indeed, this is why I introduced the "implied citation" technique -- to visualize relationships between opinions even in absence of formal citation. At the same time, I do think that subsequent citations can serve as evidence of an opinion's influence. And remember: The maps do not chart EVERY citation, just the ones the map author thinks are significant. I actually find this whole question of the significance of citations fascinating -- and have explored it in some detail in my article "Exile on Main Street: Competing Traditions and Due Process Dissent." You can find that at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1808984.
I hope my responses begin to answer your concerns. As I said, I'll try to demonstrate more in future posts.
Posted by: Colin Starger | Sep 4, 2013 2:00:07 PM
Maybe this is a simpler version of Orin's question 3, but I am not sure it got answered in your response to him, so I will ask it my way: Why do cites matter? If you are choosing not all the cites in a case, but only the ones you think are important, and in some cases cites are "implied", it seems like you could make the map look any way you wanted, with cites to unimportant cases (that are used to support minor or unimportant points) and add implied cites to cases that no one would really consider implied. In this way, it seems like your maps obscure more than they illuminate.
And at a simpler level, does it matter if one case cited another? What does that help you figure out? I agree with you that "subsequent citations can serve as evidence of an opinion's influence", but for that sort of finding, you would want to count number of cites, not just the fact that it was cited (or you thought it was implicitly cited). Does this data tell us anything, or is it just a nice picture?
Posted by: anonandoff | Sep 4, 2013 11:43:43 PM
Thanks for these probing questions.
In response, let me start by asking you to try a thought experiment. Fist, choose an area of Court doctrine that interests you. Now, pick a prominent "line of cases" from within that line. How might you represent that line when trying to efficiently teach the doctrine to a group of students? How could you structure a doctrinal discussion with colleagues unfamiliar with the area? The Mapper gives you an easy way to represent that line of cases visually.
For any line -- or for any set of competing lines -- a doctrinal map will lay out a sequence of opinions, name the opinion authors (thereby challenging the fiction of "the Court"), and indicate the relative acceptance of the opinions among members of the Court at the time. Such data cannot be gleaned from conventional Bluebook citations. Of course, that information could also be communicated through text. Yet maps will appeal more to visual thinkers and learners.
Only if the map-maker deliberately opted to depict unimportant cases would his map "more obscure than illuminate." But why would anyone want to make a map of obscure or irrelevant opinions?
Perhaps the hang-up is really about citations. To this, I should mention that the Mapper permits you to customize labels on the legend. You can therefore call the connections represented by the solid and dotted arrows by any name you like. Personally, I chose "direct citation" and "implied citation" because I have found it worked well to describe the doctrinal territories I have mapped.
As it turns out, and I don't think this is entirely coincidence, key opinions in the lines that have interested me have in fact cited to the key opinions that preceded them in the line. The general concern I have is with arguments in the "doctrinal modality" (to use Bobbitt's phrase). And doctrinal arguments tend to involve discussion of prior precedent.
Hope this helps clarify just a bit. In any event, I welcome the feedback will try to address your concerns in a different way in future posts.
Posted by: Colin Starger | Sep 5, 2013 8:10:31 AM