In Progress

And then there was dissent! Over the past few posts (Part I here, Part II here), I’ve been modifying an automatically generated citation network map focused on immigration cases. The automatic network connected 2010’s Padilla v. Kentucky all the way back to 1889’s Chinese Exclusion Case and 1893’s Fong Yue Ting at 3 degrees.  Today I switch the map’s look from a Spaeth visualization to a standard  vote view (for more on “standard vote view” see Introduction to SCOTUS Mapper 1.0). Here’s the current draft image (click to see full size):

In this visualization, the Y axis represents the number of votes an opinion receives. All opinions above the dashed middle line are thus majority opinions. Dissents likewise fall below below the dashed line (as would concurring opinions were any represented). In contrast to Spaeth’s emphasis on vote coalitions for judgment, this perspective emphasizes the relative strength and weakness of particular opinions. To grasp this difference, compare this map with the original Spaeth view and contrast the relative placement of Padilla. Note how the Spaeth view represents Padilla as a 7-2 decision (the vote coalition) while the vote view represents Stevens majority opinion as only receiving 5 votes (Justice Alito and Chief Justice Roberts concurred in judgment only).

Though reclassifying existing data points like the Padilla majority opinion is useful, the most important change with the vote view is the creation of space for dissents. Going all the way back to Ekiu and Fong Yue Ting, we can immediately see that Justice Brewer promoted a different path for immigration doctrine — one that many advocates today might support.  We also see an interesting cluster of dissents in the cases leading up to 1984’s Lopez-Mendoza. Justice Stevens manages to move from dissenting in that case to writing the majority opinion in St. Cyr and Padilla. (Technical note: Stevens dissent in Lopez-Mendoza and Douglas’ dissent in Jay both received 1 vote; I projected them at 1.5 votes in order to permit display of the other dissenting opinions).

Now this map is obviously incomplete. I have not attempted to integrate the pictured dissents into the existing “lines of opinions.” Doing this would allow me to trace the evolution of the competing jurisprudential schools with respect to immigration doctrine. Yet this is no easy task. Since dissents may exert influence without ever being cited, I would have to account for uncited dissents in constructing my competing lines. This requires careful reading of the opinions to parse out the essential ideas. Since I am not an immigration scholar, I am not inclined to do such heavy lifting at this juncture.

Instead, I offer the map above as proof of concept. The concept is that dissents belong in rich accounts of Supreme Court doctrine — and vote view maps provide a means to represent such accounts. If any reader of this blog wants to develop this map further, please let me know!

This week I’ve been working on improving a map of immigration doctrine. I created the original map by combining 3-degree citation networks connecting Padilla v. Kentucky (2010) to Fong Yue Ting (1893) and the Chinese Exclusion Cases (1889). In my last post, I updated the original map to include the “outcome votes” and “outcome directions” for the pre-1946 cases. I attempted to code these cases in a manner consistent with the Spaeth dataset, which does not include pre-1946 cases. Today, I begin the process of converting the map from a Spaeth visualization to a map that more effectively represents the competing traditions in immigration doctrine. These competing traditions form the basis of a doctrinal dialectic.

The first step in this process is to identify and display the authors of the majority opinions represented. This is an important step because it confronts and lays bare a bedrock fiction in Supreme Court discourse — the fiction that “the Court” issues opinions rather than individual justices. When referring to prior opinions, the standard convention is to say “the Court held this” or “the Court stated that.”  Yet opinions all have authors. Knowing the identity of these authors becomes important when we seek to locate competing traditions in doctrine. Here is the same Spaeth map from last time with authors included (click on the map to see in full size).

Adding authors adds value. We see, for example, that Justice Stevens wrote both modern “liberal” opinions (Padilla and St. Cyr) while Justices Field and Gray wrote the foundational “conservative” opinions (Chinese Exclusion and Fong Yue Ting). We also see that the  titans Holmes, Brandeis, Taft, and Douglas wrote the string of unanimous opinions from 1913 to 1947. The idea that this doctrine evolved as a long-running conversation among these justices becomes easier to swallow with the actual opinion authors displayed.

Yet the map needs work. Right now, it only shows half of the doctrinal story. The missing half concerns dissents. Dissenting opinions can exert massive influence over the development of Court doctrine even if they do not formally state “the law.” Thus, I believe that no truly careful study of any given are of Court jurisprudence can ignore dissents. Introducing dissents into the picture is thus the next step in the development of this map. Stay tuned!

In my last post, I vowed to upgrade a map plotting immigration doctrine to include pre-1946 vote counts and decision directions. The reason the map did not already include this information is that the Spaeth (Supreme Court Database) dataset does not cover cases decided before 1946.

The upgrade process involved examining each case previously plotted in green to figure out two things: (1) the number of votes for and against the majority outcome; and (2) the “direction” of the outcome (liberal or conservative). “Outcome” refers to judgment — who wins or loses. In the usual case, outcome votes will add up to 9. In cases where less than 9 justices actually cast votes, the opinion is plotted based on the number of dissents. (Thus, for example: a 8-0 would be plotted in the same position as a 9-0 and a 5-2 would be plotted in the same position as a 7-2.) Before I explain more, take a peek at the upgraded map (click on it for a full-sized view):

Note that Spaeth’s outcome measurement does not account for subtle hermeneutic distinctions that sometimes arise in fractured decisions. Outcome does not track whether a justice joins part of an opinions but not all. Outcome tracks votes. In the outcome paradigm, the votes of justices who concur “in judgment only” count as part of the majority outcome.

“Decision direction” is a similarly blunt measurement. Under the Spaeth system, direction is either liberal or conservative. While sometimes controversial, the direction code is best thought of as recording who won or lost at the Court. For this map, and following Spaeth for the post-1946 cases, I coded cases where the immigrant/person challenging the government WON relief as “liberal.” On the flip side, when the immigrant LOST, I coded it as “conservative.” This was not always a cut-and-dry process.

Consider 1922’s Ng Fung Ho as just one example. This case involved four Chinese persons challenging their deportations via habeas. Justice Brandeis wrote the unanimous opinion which ultimately allowed two of the individuals to challenge their detention, but denied the other two relief. I coded the opinion conservative because Brandeis offered a very strong view of sovereign power to exclude and was not at all sympathetic to the immigrants’ claims. But because relief was granted, I could have coded the decision liberal.

Such difficult-to-make decisions abound. The best way to avoid them, in my opinion, is to track doctrinal “concepts” rather than “outcomes.” Yet this is a tricky enterprise because it requires a close reading of the cases. Concepts cannot be plotted simply by leveraging the Spaeth dataset. It is a much slower and intensive process — but one that results in more subtle visualizations.

Yet the Spaeth-driven method is still very useful. Though generated quickly, this upgraded map is still quite revealing. For example, it clearly shows a long period of consensus followed by dissent. From 1913’s Bugajewtiz to 1947’s Delgadillo, the Court had nothing but unanimous decisions. Post WWII, we start to see divisions in the doctrine. Though “conservative” cases dominate, “liberal” ones sneak in there too. This is territory worth exploring in more detail. And so I will in the next post!

In the 10/20 Map of the Week, I quickly rendered an immigration network. To make it, I combined two 3-degree citation network. Both networks started with 2010’s Padilla v. Kentucky as a child. Padilla’s two network parents were the Chinese Exclusion Case (1889) and Fong Yue Ting (1892). I then filtered this combined two-parent network to opinions containing the word “immigration.” The resulting map looked like this:

Note that this look of the network employs a Spaeth visualization. Unfortunately, Spaeth only goes back to 1946. Many of the cases in immigration doctrine are much older. Thus our immigration problem: The y-axis is meaningless for the cases plotted in green.

First project then is to rehab this map by inserting proper vote counts pre-1946. I’ll have to count votes by hand.  Of course, this fix won’t solve all our immigration problems. But it will help and is a fair first step. Will report back on that step and on future steps in the next post.