SEAL 2009 @ Vanderbilt Law School



I am currently at Vanderbilt Law School for the 2009 Society for Evolutionary Analysis in Law (SEAL) Conference.  For those of you not familiar with the organization … “SEAL is a scholarly association dedicated to fostering interdisciplinary exploration of issues at the intersection of law, biology, and evolutionary theory, improving the models of human behavior relevant to law, and promoting the integration of life science and social science perspectives on law-relevant topics through scholarship, teaching, and empirical research.” The organization embraces a wide range of scholarship including those with interests in evolutionary and behavioral biology, complex adaptive systems, economics, psychology, primatology and anthropology. 

In the coming days, we will be highlighting our work on The American Legal Academy and previewing extensions of the paper Reproduction of Hierarchy? A Social Network Analysis of the American Law Professoriate.  So stay tuned for this and more… please add us to your blogrolls and tell your colleagues about the CLS Blog. 

Tax Day! A First-Order History of the Supreme Court and Tax


Click to view the full image.

In honor of Tax Day, we’ve produced a simple time series representation of the Supreme Court and tax.  The above plot shows the how often the word “tax” occurs in the cases of  the Supreme Court, for each year – that is, what proportion of all words in every case in a given year are the word “tax.”  The data underneath includes non-procedural cases from 1790 to 2004.  The arrows highlight important legislation and cases for income tax as well.

Make sure to click through the image to view the full size.

Happy Tax Day!

OpenSecrets Open Data! – Visualizing the Publicly Traded Assets of Senators in 2007

picture-36 went open with their data today.  In honor of this very significant act and its ramifications on future government transparency, I’ve decided to produce a quick visualization to answer a question I’ve long been intrigued by – the publicly traded holdings of Senators.  One good proxy to this question is the Personal Financial Disclosure data released today by Open Secrets.  In the words of Open Secrets,

Any legal ownership a person has in a company or property is classified as an asset, including brokerage accounts, corporate bonds and stocks. For the most part, lawmakers seem to have a stake in big-name, recognizable companies and properties. They need to report only assets worth more than $1,000 at the end of the calendar year, or producing more than $200 of income. (One note about mutual funds: Filers are not required to provide detail on funds’ individual holdings.) Any purchases, sales or exchanges of assets during the year of more than $1,000 must be disclosed as transactions. Reporting the value of a primary residence, unless it produces income, is not required.

This visual represents a very rough cut of this data for the holdings of Senators in 2007.  As is often the case with real form data, there are inconsistencies across the data set.  Ideally, one could simply use the asset’s description, which would accurately report the assets held in an account.  However, as these are not always provided, the “asset source” is used in their absence.  When combined with the mutual fund issues mentioned above, blind trust reporting, and these possible transcription errors, there is a compelling case against strict interpretation of this visual.  In any case, enjoy the visual, and again, thanks to Open Secrets for their monumental move forward!

Visualizing 26 U.S.C ___ : At the "Section Depth"

26 USC __


Title 26 of the United States Code is likely on the mind of many as we move toward April 15th. As a part of a project with Lilian V. Faulhaber (Climenko Fellow from HLS), we have become interested in the architecture of 26 U.S.C. ___.

As we define it, a “section depth” representation for 26 U.S.C. 501(c)(3) represents a traversal to the level of Sec. 501. While a “full depth” representation would include a mapping beyond Sec. 501 to its (c) and (3) subcomponents. In our previous post highlighting 11 U.S.C. __ (the Bankruptcy Code), we presented a traversable “full depth” representation for its structure.


Considering all 50 titles of the United States Code, 26 U.S.C. __ is among the largest of the titles in its architectural size and depth. In fact, given its size, it is not possible for us to render for public consumption, a labeled, “full depth” and zoomable representation for all of 26 U.S.C. ___.

Above we provide a “section depth” representation for 26 U.S.C __ where the terminal nodes are sections such Sec. 1031. You will notice that this section depth representation is roughly the size of the full depth representation we provide for 11 U.S.C. ___. We have colored in Green the primary Income Tax Sections under 26. The documentation is similar to that for 11 U.S.C. ___ thus please refer to this post for additional information. However, for traversal purposes, it is important to remember to start in the middle at the “26 U.S.C.” and follow the branch of the graph out to a leaf node.

We believe the comparative consideration of the architecture for these titles offers a rough first cut on questions of code magnitude and complexity. Although it is a first order approximation and we do believe layering in the relevant administrative regulations and jurisprudence to these sections would represent an improvement on the question, it still bears asking whether it would drastically alter the macro state of affairs. That is an empirical question and only time will tell.


Visualizing Contributions to the 110th Congress—House Edition (Take 2)

110th House - One Mode Projection

In our previous graph visualizations of contributions to members of the House and Senate, there have been two types of entities: Contributors and Congressmen.  This division manifests itself in the dynamics of the graph as well – Contributors give only to Congressmen, and Congressmen receive only from Contributors.  A network with this property is bipartite, and there are a number of additional ways to represent the relations contained therein.

One such representation is called the single-mode projection.  In this simplification, only elements from one of the two sets (e.g., Congressmen or Contributors) are displayed.  A relationship exists between two elements in the visual if they share a relationship with at least one member of the other group.  For instance, both Bernie Sanders and Sam Brownback  received campaign contributions from the the National Association of Realtors.  Thus, the Congressmen both have a relationship with the same Contributor, and the simplest single-mode projection would represent this as a shared relationship between the two Congressmen.  For more on this projection or other representations, see Wasserman and Faust (1994) or M. E. J. Newman, S. H. Strogatz, and D. J. Watts, Phys. Rev. E 64, 026118 (2001).

As the Sanders-Brownback example above demonstrates, however, it is relatively easy to be connected in this representation.  Thus, we enforce a threshold on the number of shared contributors for a relationship to exist – representatives must shared at least 10 contributors in order for a relationship to exist between them.  It is important to note that different thresholds may produce different graphs.  We have chosen this figure as it represents roughly half the number of major contributors to the typical representative. Obviously, alternative specifications are possible.  In future posts, we may present different thresholds or normalizations. However, for now, we believe this is a simple but appropriate representation for the underlying data.

Given the requirement of sharing at least 10 contributors with another member, the above visualization no longer contains every member of the House of Representatives.  To observe the full graph, please see our first post on the contributions to the House.

Interpreting this visual is very similar to previous visuals, and in many ways simpler.  Other than points (1) and (2) below, again refer to our first post on the contributions to the House.


(1) SIZING of the CONNECTIONS — Each Connection (Arc) between an Institution and a Member of the House is sized according to the amount of money flowing through a connection. Darker connections represent larger flows of money while lighter connections represent smaller amounts of money.


(2) COLORING of the CONNECTIONS — Each connection representing a shared campaign contributor  from between members of Congress is colored according to partisan affiliation. Using popular convention, we color shared relationships between Republican Party members as Red, and blue for shared relationships between Democratic Party members.  For relationships that span across the party lines, the color green is used.


Visualizing Contributions to the 110th Congress — The House Edition





Visualizing the Campaign Contributions to the Representatives of the 110th Congress —
The House Edition

University of Michigan
Center for the Study of Complex Systems
Department of Political Science


110th Congress = January 3, 2007 – January 3, 2009

435 Voting Members of the United States House of Representatives + District of Columbia (Eleanor Holmes Norton) + Puerto Rico (Luis Fortuno) +  Virgin Islands (Donna Christian-Green) + American Samoa (Eni F H Faleomavega) + Guam (Madeleine Z Bordallo)

Click here and here for the Senators of the 110th Congress.


Squares (Institutions) Introduce Money into the System and Circles (Congressmen) Receive Money.



Using recently published data on campaign contributions collected by the Federal Election Commission and aggregated by the Center for Responsive Politics our visualizations track large money donations to members of the 110th Congress over the 2007 – 2008 window.

It is important to note that most of these organizations did not directly donate. Rather, as noted by the Center for Responsive Politics “the money came from the organization’s PAC, its individual members or employees or owners, and those individuals’ immediate families. Organization totals include subsidiaries and affiliates. Of course, it is impossible to know either the economic interest that made each individual contribution possible or the motivation for each individual giver. However, the patterns of contributions provide critical information for voters, researchers and others.”

The Center describes its methodology here

To provide for an optically tractable view of the top contributions, we follow the CRP and impose the limiting requirement that to be included in our tally a given group’s contribution must fall within a given house members top contributor list.

We try to strike a tradeoff between information overload and incomplete disclosure.

In coming days, we will provide an additional visualization of the underlying data.  Check back soon!



2,508 of the Donors are captured in the Graph.

Total Recorded Donations Introduced into our Visualization by these Entities Total to  $113,134,698


(1) SIZING of the REPRESENTATIVE NODES — Each Circular node representing a Member of the House is sized according the amount of incoming donations. Thus, larger nodes are the recipients of larger sums of money while the smaller nodes received smaller amounts of money.



(2) COLORING and SHAPES of the REPRESENTATIVE NODES — Each node representing a Member of the United States House of Representatives is colored according their Political Party. Using popular convention, we color members of the Republican Party as Red and members of the Democratic Party as Blue.



(3) SIZING of the CONNECTIONS — Each Connection (Arc) between an Institution and a Member of the House is sized according to the amount of money flowing through a connection. Darker connections represent larger flows of money while lighter connections represent smaller amounts of money.



(4) COLORING of the CONNECTIONS — Each connection representing a campaign contribution from an entity to a member of Congress is colored according to partisan affiliation of the receiving representatives. Using popular convention, we color members of the Republican Party as Red and members of the Democratic Party as Blue.



(5) STRUCTURE OF THE GRAPH The Graph is Visualized Using Fruchterman-Reingold. This is an automated spring embedded, force directed placement algorithm often used in the network science literature to visualize graphs of this size.


(6) ACKNOWLEDGEMENTS We thank Rick RioloJon ZelnerCarl Simon, Scott Page and the Center for Responsive Politics for their comments, contributions and/or data.

Classic Model from Complex Systems: The El Farol Bar Problem


I recently attended a conference at the Santa Fe Institute.  During the trip, I made a point of eating at the El Farol Bar & Restaurant. This restaurant holds a special place in the lore of complex systems.  Thus, I thought I would take the opportunity to highlight the model on the CLS blog.  

Here is a subset of the model description…. “The bar is popular — especially on Thursday nights when they offer Irish music — but sometimes becomes overcrowded and unpleasant. In fact, if the patrons of the bar think it will be overcrowded they stay home; otherwise they go enjoy themselves at El Farol. This model explores what happens to the overall attendance at the bar on these popular Thursday evenings, as the patrons use different strategies for determining how crowded they think the bar will be.”   

The original paper written by Brian Arthur is located here. An interesting follow up paper employing reinforcement learning is located here.    This above is a screen print from the Netlogo model.  Netlogo offers an easy interface useful for exploring a variety of agent based models.  

The model will run in your browser provided you have Java 1.4.1+.  

To run the El Farol model, please go here.   

April as the Cruellest Month? Data on the Law Clerkship Tournament




Judge Wald’s classic article describing the market for judicial clerks reminds us how April was once the cruellest month.  Given the Federal Law Clerk Hiring plan has shifted the relevant window of discomfort, we thought it reasonable to ring in the spring season with some of our data on the law clerk tournament.  Using underlying information Derek Stafford and I collected for our article Hustle and Flow: A Social Network Analysis of the American Federal Judiciary, here is Federal Court Clerkship data for the period of the “Natural” Rehnquist Court.  The current offering is aimed at the US News Top 15 Law Schools.  Although this data terminates in the 2004- 2005 clerkship year, we still believe it offers useful empirical insight into the status of the law clerk tournament. 

Computer Programming and the Law — OR — How I Learned to Learn Live with Python and Leverage Developments in Information Science




One of our very first posts highlighted a recent article in Science Magazine describing the possibilities of and perils associated with a computational revolution in the social sciences.  A very timely article by Paul Ohm (UC-Boulder Law School) entitled Computer Programming and the Law: A New Research Agenda represents the legal studies analog the science magazine article.  From information retrieval to analysis to visualization, we believe this article outlines the Computational Legal Studies playbook in a very accessable manner.

Prior to founding this blog, we had little doubt that developments in informatics and the science associated with Web 2.0 would benefit the production of a wide class of theoretical and empirical legal scholarship. In order to lower the costs to collective action and generate a forum for interested scholars, we believed it would be useful to produce the Computational Legal Studies Blog. The early results have been very satisfying. For example, it has helped us link to the work of Paul Ohm.  

For those interested in learning more about not only the potential benefits of a computational revolution in legal science but also some of the relevant mechanics, we strongly suggest you consider giving his new article a read!  

Coming Next Week on CLS Blog



A Netlogo 3D screenprint of one of the classic agent based models—the Shelling Segregation Model is above. We offer it as a holdover until CLS Blog Returns Sunday Night with more exciting content…..

(1) Discussion of a New Paper: Computer Programming and the Law
(2) Visualizing the 110th Congress — The House of Representatives
(3) For Law Students and Law Professors — Data on the Law Clerk Tournament
(4) And More …..

Data Mining the News — J. Kleinberg Work Discussed in MIT Tech Review


This short but cool article from MIT Technology Review discusses recent work by Computer Scientist Jon Kleinberg and his Cornell colleagues. This very nice visualization is the byproduct of their efforts at data mining more than 1 million online news items per day in the weeks leading up to the 2008 presidential election.

With Bankruptcy on Our Minds: The Structure of Title 11 U.S.C.



We have become interesting in visualizing the structure of the law including its components and subcomponents.  In reduced form, statutes, regulations and certain other units of the law can be characterized in graph theoretical terms.  While we do not make deep inroads on the content of this above graph, we do generate a tree traversable visualization for its structure.

Much of my training in law school (particularly in the so called “code-based” classes) was focused upon developing mental models for the structure and content of graphs such as the one displayed above. In my case, I believe the usage of such a visualization early in a code-based course would have been beneficial. Thus, we offer this traversable visualization to the world for not only its research value but also for pedagogical purposes.


Start in the MIDDLE at the “11 U.S.C.__ ” Label and traverse out.

GREEN NODE LABELS =   for SECTIONS  {In the Example below, 11 U.S.C. § 101}
YELLOW ARCS — Chapter 7 of Title 11 = LIQUIDATION
BLUE ARCS — Chapter 11 of Title 11 = REORGANIZATION  (aka “Filing Chapter 11“)
RED ARCS and GREY ARCS — Balance of the Chapters under Title 11
Red Arcs are for lines which lead to terminal nodes
Grey Arcs are for lines which do not immediately lead to terminal nodes


Please feel free to PLAY AROUND and TEST IT OUT!
This is an early production version so please provide us with any feedback and/or suggestions.

Co-Sponsorship Networks– Senators of the 108th Congress


In the days and weeks to come we will turn our attenion away from Congress in favor of other institutions and substantive questions. However, given our prior posts focusing upon the structure of the 110th Congress, we thought it proper to highlight some relevant realted scholarship. James Fowler from the UCSD Political Science Department and leader of the Networks in Political Science movement has published several papers exploring the strucutre of legislative co-sponsorship.  You can find a link to these papers here. My favorite of these papers is Community Structure in Congressional Cosponsorship Networks published in Physica A by Yan Zhang, A. J. Friend, Amanda L. Traud, Mason A. Porter, James H. Fowler & Peter J. Mucha. The above figure, drawn from the paper, is a dendrogram for the legislative cosponsorship network of the Senate of the 108th Congress.  

Senators of the 110th Congress Take 2-Contributions by Industry/Sector

To view the full image, please click here.

Senator By Industry


This represents a deeper cut on campaign contributions to the Senators of 110th Congress. Again, we rely upon data from the Center for Responsive Politics.  The CRP aggregates contribution data up to the industry or economic sector. Thus, as before, we adopt their classification scheme and methodology herein.  While aggregating to the industry/sector level removes the degree of specificity we offered in our earlier post, it provides a cleaner representation for the graph.  For those interested in the other chamber, click here for the House of Representatives.

Click on the picture above and it will take you to our flash where you can zoom in and read the labels.

As you review the graph, please consider the following:

(1) Industries locate in the center of the graph because they provide significant funding to both Democrats and Republicans.

(2) Industries which generally only fund one political party are located toward the respective red/blue boundary.  For example, it is hardly surprising to observe the location of “Oil and Gas” relative to “Environmental” groups.

(3) It is important to note that we do not impose the partisan separation or the placement of party outliers apparent in the image. Rather, the algorithm places Red Senators in Blue Territory and Blue Senators in Red Territory because they receive significant sums from industries who typically fund the opposing party. For example, consider Senator Olympia J Snowe (R-ME) who is typically characterized as a moderate Republican.  Since she receives money from more industries that typically fund Democrats than Republicans, she is placed in Blue Territory by the algorithm.

(4) It is important not to over read the position of Senator Herb Kohl (D-WI).  Over the relevant time window, Senator Kohl received 94% of his resources through self-financing.