Obama Analysis

Results

Figure 1: Network Graphs of All Plaintiff States by AG Political Party. Node size represents degree centrality, and edge widths are weighted.

The network graph displayed in Figure 1 illustrates the plaintiff states, labeled by the political party affiliation of their attorney general at the time the state joined the case.

During the Biden Administration, there were nine states who had a change in the political party of their elected attorney general: Alaska, Arkansas, Mississippi, Missouri, Nevada, New Jersey, Ohio, Oklahoma, and Washington.

The degree assortativity, or the preference of nodes to attach to nodes of similar degree, is 0.15, revealing assortativity between states with similar number of connections. The political assortativity, preference of states to connect with states of the same political party is 0.2, positively assortative.

The network graph of plaintiff states is presented interactively in the plot, Figure 2.

Figure 2: State Plaintiff Network Graph, plotly visualization. Node and edge size are weighted. Node size represents scaled strength.

Table 1: Top Ten States by Weighted Degree Centrality. Column Centrality Measures include unweighted degree, eigenvector, betweenness, closeness, and strength.

State	Degree	Eigenvector	Betweenness	Closeness	Strength
TX	45	1.00	9.49	0.35	16.95
AL	44	0.65	12.50	0.40	10.88
ND	42	0.59	5.39	0.37	10.19
NE	43	0.61	17.97	0.41	10.14
WV	39	0.59	13.76	0.37	10.05
KS	41	0.57	7.24	0.39	9.49
MI	43	0.56	8.98	0.39	9.44
SC	43	0.56	9.68	0.39	9.30
OK_R	37	0.55	3.73	0.37	9.02
LA	43	0.55	10.83	0.40	8.88

Figure 3: Network Graph of Litigators. Node color and shape describes political party affiliation. Node size describes represents degree centrality. Edge size is the number of interactions between any two nodes.

Figure 3 shows the network of judges and attorneys. The edge sizes in the graph are weighted by the number of interactions between each set of two nodes. The graph contains a tight core of Republican attorneys, surrounded by lesser connected litigators and a smaller cluster of Democrat attorneys. This network has a degree assortativity of 0.05, indicating a slight preference for attorneys and judges of similar degree centrality to connect with each other.

An interactive renditions of the litigator network is accessible through Figure 4.

Figure 4: Attorney Network Visualization by plotly.

Table 2: Ten Most Central Judges with Attributes and Network Statistics

Judge Name	Title	Organization	Political Party	Degree	Eigenvector Centrality
Brett Kavanaugh	Circuit Court Judge	D.C. Circuit Court of Appeals	Republican	28.07143	0.60
Janice Rogers Brown	Circuit Court Judge	D.C. Circuit Court of Appeals	Republican	26.78571	0.52
Thomas Griffith	Circuit Court Judge	D.C. Circuit Court of Appeals	Republican	25.78571	0.61
David Tatel	Circuit Court Judge	D.C. Circuit Court of Appeals	Democrat	24.57143	0.48
Karen Henderson	Circuit Court Judge	D.C. Circuit Court of Appeals	Republican	22.35714	0.58
Judith Rogers	Circuit Court Judge	D.C. Circuit Court of Appeals	Democrat	22.14286	0.54
Cornelia Pillard	Circuit Court Judge	D.C. Circuit Court of Appeals	Democrat	18.14286	0.41
David Sentelle	Circuit Court Judge	D.C. Circuit Court of Appeals	Republican	15.64286	0.21
Robert Wilkins	Circuit Court Judge	D.C. Circuit Court of Appeals	Democrat	14.57143	0.28
Sri Srinivasan	Circuit Court Judge	D.C. Circuit Court of Appeals	Democrat	14.50000	0.32
a Network Statistics are Weighted.

Table 2 lists the judges in their organization with their respective political party affiliations, degree centrality, and eigenvector centrality. The network includes 80 judges serving across 27 districts.

The judge with the highest degree centrality value is Brett Kavanaugh, a Circut Court Judge on the D.C. Circuit Court of Appeals. On the same circuit, is Thomas Griffith, who has the highest eigenvector centrality value. Interestingly, the D.C. Circuit Court of Appeals is home to the ten most central judges. The political party affiliation of these judges is evenly split.

Table 3: Ten Most Central Attorneys with Attributes and Network Statistics

Attorney Name	Title	Organization	Political Party	Strength	Eigenvector Centrality
Luther Strange	Attorney General	Alabama	Republican	43.96	1.00
Derek Schmidt	Attorney General	Kansas	Republican	43.04	0.97
Alan Wilson	Attorney General	South Carolina	Republican	39.64	0.87
Michael Dewine	Attorney General	Ohio	Republican	38.36	0.88
Scott Pruitt	Attorney General	Oklahoma	Republican	36.21	0.84
Wayne Stenehjem	Attorney General	North Dakota	Republican	35.64	0.69
Patrick Morrisey	Attorney General	West Virginia	Republican	34.79	0.83
Leslie Rutledge	Attorney General	Arkansas	Republican	34.11	0.78
Jeffrey Chanay	Chief Deputy Attorney General	Kansas	Republican	33.96	0.82
James Smith	Deputy Solicitor General	South Carolina	Republican	32.71	0.78

Table 3 displays the ten most central attorneys by degree centrality. The attorney with the greatest number of weighted connections is Luther Strange, Attorney General from Alabama, who also has the highest eigenvector centrality. These top ten attorneys feature soley individuals from the Republican party. In the list of the top ten attorneys, we find the titles of attorney general most prevalent, along with General Chief Deputy, and Deputy Solicitor General.

Figure 5: Degree Centrality for Attorney and Solicitor

After considering the networks of plaintiff states, we show the network of the litigation targets. These are the federal agencies or individuals whom the claims have been filed against. The network represented in Figure 6 depicts the connections among the litigation targets engaged in multistate litigation. This network exhibits several distinct clusters.

Figure 6: Network of the Litigation Targets

The interaction between plaintiff states and the litigation targets is visualized in Figure 7. The network consists of two distinct types of nodes, defendants and plaintiffs, engaged together in legal cases. This network graph displays the interaction of states arranged by the political party of their attorney general at the time of joining the litigation case with the separate federal agencies and targets.

Figure 7: A Network Graph of the Interactions between States and Federal Agencies. Edges are weighted by the number of connections between nodes.

Next, we address the timing of the number of attorney generals that participate on a litigation case relative to the filing date. Figure 8 shows the relationship of the filing date and the number of attorney generals. Perhaps, it would be interesting to correlate this with other world events or the presidential Gallup poll, except that I assume these cases take awhile and states join at different times.

Figure 8: Relationship between Filing Date and Number of Attorney Generals