Is there a formula to becoming a Supreme Court clerk?  Perhaps a certain pedigree? How about a prior federal clerkship?  The short answer to all three questions is yes.  At very least, going to a certain school and working for a specific judge significantly enhance your chances.

Much of the discussion surrounding Supreme Court clerkships has centered around “feeder judges” or federal appellate judges whose clerks frequently go on to Supreme Court clerkships.  One way to understand and visualize the process leading to Supreme Court clerkships is as a network.  Network analysis allows you to see the connections between different actors along the trajectory.

In their article “Hustle and Flow,” Daniel Katz and Derek Stafford used network analysis to gain insights into the important actors in this process for the years 1995-2004.  In that paper the authors found the most central appellate judges in this process were Michael Luttig from the Fourth Circuit, J. Harvie Wilkinson from the Fourth Circuit, and Alex Kosinski from the Ninth Circuit.  In this post I provide an abbreviated analysis of the clerk network from the beginning of the Roberts Court in 2005 through the 2016 clerk hires.

In their article, Katz and Stafford looked at all of the Article III judges in the network leading to Supreme Court clerkships. In this post I look at the judges the clerks worked for as well as at the law schools that the clerks attended.

Before looking at the main network for all of the years, the first figure is a visualization of the trajectories of the 2016 clerks (thanks to information on Above the Law‘s website).

In this figure and the next, the lines get darker as we move from less central to more central nodes (or actors) in the diagram.  The more central nodes also generally move towards the center of this diagram.  The arrows show the tracks that the clerks move along – from law school, to prior clerkship(s), to the Supreme Court.

For a quick note about what I mean by centrality – there are actually multiple types and yet I refer to one measure, eigenvector centrality, as the main measure.  Eigenvector centrality is based on a node’s connections to other important nodes in the diagram.  Nodes that are connected to more of these nodes have higher scores.  Other types of centrality that I mention in the post include closeness centrality and betweenness centrality.  Closeness centrality is a measure of the nodes proximity to other nodes (sort of like in a board game) so that a high closeness centrality means it takes fewer steps to navigate from that actor to all other actors in the diagram (relative to other actors).  Betweenness centrality is a measure that only applies to nodes that are connected in both direction (feed in and feed out).  Nodes that bridge other nodes and create shorter paths between them generate higher betweennesss centrality scores. [To learn more about network analysis Professor Matt Jackson provides an excellent introduction here]

With that we can make better sense of the diagram of the trajectories of the 2016 Supreme Court clerk hires.  Based on eigenvector centrality, Yale is the most central node in this diagram.  This means Yale is connected to more important nodes (feeder judges) in the diagram than any other node.  After Yale the most central schools are Harvard and NYU.  In terms of judges, the most central feeders are, in order of centrality, Judge Garland, Fletcher, Griffith, Gorsuch, and Rakoff (from the Southern District of New York).

Now to the full network.  The network diagram is below.

Obviously this is a much denser network than the one that encapsulated only the clerks hired for the 2016 Term.  As with the 2016 network, the central actors in this network cluster towards the center of the diagram with the less central actors towards the periphery.

What do the centrality measures tell us about Supreme Court clerkships?

• School matters: Based on their eigenvector centralities, many of the top-ranked schools top the list. The top five most central schools in order of centrality are Harvard, Yale, Stanford, University of Virginia, and University of Chicago.  Harvard and Yale were in the top five centrality scores when also including all judges.  In terms of closeness centrality, the top five schools are Harvard, Yale, Stanford, Columbia, and Georgetown.  The raw numbers back up these figures.  Based on counts of overall clerkships, Harvard was the starting point for 116, Yale for 113, and Stanford was next with 39.  Given Yale’s class size though, its relative rate of sending students along the path to Supreme Court clerkships far exceeds that of any of the other schools.
• Feeder Judges Matter: The most central node in the diagram belongs to the current nominee to the Supreme Court, Judge Garland from the DC Circuit. He also topped the list by feeding 41 clerks to the Supreme Court over this period.  Judge Garland is followed by Judges Kavanaugh from the DC Circuit, Wilkinson from the Fourth Circuit, Kozinski from the Ninth Circuit, and Boudin from the First Circuit.  In terms of closeness centrality though, two of the three highest scores go to judges outside the federal appellate courts.  The top score goes to Judge Cote from the Southern District of New York, the second highest score is for Judge Reinhardt of the Ninth Circuit, and the third highest score goes to Judge Liu from the California Supreme Court. Finally, for betweenness centrality, the top judges in order of centrality are Judges Tatel from the DC Circuit , Sentelle from the DC Circuit, Reinhardt from the Ninth Circuit, Kavanaugh from the DC Circuit, and Sutton from the Sixth Circuit.

Below is a table of the top ten most central law schools and judges based on their eigenvector centralities.

Data aggregated from Wikipedia’s list of Supreme Court Clerks  and from Above the Law’s coverage of 2016 Supreme Court Clerk hires.

On Twitter: @AdamSFeldman