The October 2022 term (OT2022) of the U.S. Supreme Court produced elements of a 3-3-3 division, with three reliable liberal justices, a Roberts-Kavanaugh-Barrett coalition and a slightly less cohesive conservative group of Gorsuch-Thomas-Alito.
Analysis here is based on the rates of agreement in judgment presented by EmpiricalSCOTUS. Their work is much appreciated.
Figure 1 shows the non-metric spatial scaling of the justices in two dimensions. The horizontal x-axis shows a clear left-right ordering of the justices based on their agreement in judgment.
Note that the left-right dimension reflects relative location, not an absolute measure of ideology. The three justices near zero are “in the middle” but that does not mean their positions are “moderate” in some objective sense. Further, the detailed content of opinions is not reflected by these agreement measures, which only tell us the voting coalitions, not the legal consequences of the rulings.
By this measure, Kagen, Sotomayor and Jackson are closely packed on the left of the Court. Roberts, Kavanaugh and Barett cluster in the middle, with Roberts slightly less conservative and Barrett slightly more conservative than Kavanaugh. On the right, Gorsuch is less to the right than Thomas, with Alito anchoring the right end of the Court.
The vertical, y-axis, dimension does not fit the notion of an “institutionalist” dimension that seemed to appear in 2021. Vertically, Jackson and Gorsuch are similar, as are Sotomayor and Alito, Kagan and Roberts, and Barrett and Thomas, with Kavanaugh in the middle of all. Without giving a label to this dimension, it is notable that the three most conservative justices are more spread out across this 2nd dimension than the three liberals or the three middle justices. In this we see a reflection of the lower agreement among Gorsuch, Thomas and Alito than among either of the other two clusters.
Figure 2 shows the agreement among justice as a heat map. The ordering here (which combines both dimensions) is slightly different than the left-right order in Figure 1, though the groupings are the same. Light shades indicate stronger agreement and dark shades indicate greater disagreement between each pair of justices.
The branches in the left and top margins show which clusters form in order of internal similarity and external difference. The first branch is clearly the 3 liberals vs the 6 conservatives. Next is the distinction between the 3 middle justices and the 3 most conservative ones. There are then finer, and more modest, distinctions within each of the 3 clusters, splitting Kagan, Barrett and Thomas slightly away from the other two justices in their clusters.
The grouping in OT2022 clearly argues for two primary clusters (the 3-6 Court) and for three secondary clusters (the 3-3-3 Court).
The similarly light shading of the Kagan-Sotomayor-Jackson group and the Barrett-Kavanaugh-Roberts group shows these two clusters were similar in high inter-agreement within the group. The three most conservative justices cluster in a somewhat less cohesive group, compared to the other two groupings. This reflects their relative disagreement on the 2nd dimention in Figure 1.
The table of agreement in judgement measure is shown in Table 1, with the three clusters highlighted.
Agreement among the three liberal justices range from 89%-95%, and among the three center justices also ranges from 89%-95%. In the cluster of the most conservative justices agreement is a bit lower, from 76%-87%. This less cohesive group is reflected in the spread in the 2nd dimension in Figure 1.
The liberal cluster has remained well defined with the addition of Jackson replacing Breyer in OT2022. In OT2021, Roberts and Kavanaugh agreed 100% of the time, defining the middle but not closely joined by Barrett who was more often aligned with the right cluster in both OT2020 and OT2021.