The Supreme Court's Long Shadow: Judicial Behavior on the Shadow Docket Between the 2010 and 2025 Terms

Using a dataset of decisions on the shadow docket with dissents between 2010 and 2025, this article presents of picture of coalitions and power distribution outside of the Court’s merits docket.

For most of the Court’s modern history, the orders list lived in the background—dense, technical, and largely ignored outside the bar. Over the last decade, that changed as emergency orders began steering headline disputes on compressed timelines. Legal academics have pressed the Court on process and transparency while empirical researchers have tried to measure what, exactly, is happening. William Baude’s early account supplied both the name—“shadow docket”—and a caution: the Court’s non-merits orders don’t always match the “high standards of procedural regularity” associated with argued cases, and a clearer view of the orders list should make readers more skeptical of reform debates that fixate only on merits procedure. Building on that framing, Steve Vladeck has emphasized the stakes in practical terms: from immigration and elections to abortion, the death penalty, religious liberty, and administrative power, the Court has—with increasing frequency—stepped into fraught disputes through decisions that are often unseen, unsigned, and almost always unexplained. Meanwhile, work by scholars including Jonathan Kastellec and Anthony Taboni along with that of Taraleigh Davis push the conversation toward measurement—treating the orders list as data to be described and compared, not just debated.

Following Kastellec and Taboni, I use a simple, workmanlike definition: the shadow docket comprises all decisions the Court makes outside the argued-and-signed merits docket. That category includes emergency stays and injunctions pending appeal, administrative stays, summary dispositions, and other orders that move or halt lower-court rulings—sometimes late at night, often with immediate real-world consequences.

A recent example shows why attention has spiked. On July 14, 2025, in McMahon v. New York, the Court granted the government’s emergency request to stay a district-court injunction that had blocked steps to wind down the Department of Education. The order itself was brief and unsigned; three Justices—Sotomayor, joined by Kagan and Jackson—issued a lengthy dissent, casting the dispute as a separation-of-powers fight and warning that allowing mass terminations to proceed risked letting the Executive “repeal statutes by firing all those necessary to carry them out.” Whatever one’s view of the merits, the mechanics matter for empirical analysis: a high-impact national policy shifted through an emergency order in which the majority’s reasoning appears only in the bottom line and timetable, with legal argument primarily visible in dissent.

Where This Analysis Is Headed

The data for this article goes from October 2010 through May 2025. According to docket counts organized by Kastellec and Taboni, during this period, the Supreme Court provided orders or decisions in 93,832 dockets outside of its merits docket. Within this set of decisions, a small subset of 387 dockets incurred dissenting votes.

The goal of this analysis is to understand judicial behavior in the shadow docket. Since decisions in this docket come without majority opinions, coalitions can be deduced in two ways. The dissenting coalitions are noted by their votes. There is also a practice of “would grant” or “would deny” which essentially functions as a dissent, but due to the different nature of these disagreements, I track these differently. In the last analysis in this article, I treat non-noted votes in the shadow docket as forming unique groupings akin to majority coalitions for these cases. Although scholars lack consensus about whether a non-described vote means a vote with the majority, treatment them as such at least creates a way to draw distinctions between justices within the same decisions.

Data

Of the 387 dockets with dissents, the greatest number (54) came during the 2018 term.

With this baseline in mind, here are the percentage of total decisions on the shadow docket with dissents.

What the data show

In this section I tell the story of the shadow docket through two complementary lenses. First, I use a dissent lens (Xs in the data) to show who breaks from the Court and with whom. Second, I use an overall voting lens that still treats dissents as the strongest signal but also folds in would-grant/deny signals (Ws in the data) and ordinary non-noted (majority) votes, so I can see who generally travels together For each single term and for non-overlapping three-term windows, I group the justices into two or three coalitions and choose the split that most clearly keeps similar voting patterns together. Isolating each justice individually we see the following:

Who anchors the Court’s poles

Across the period, the poles of the Court are easy to recognize because the same names reappear as anchors. On the left, Sonia Sotomayor is the most consistent dissent anchor, with Elena Kagan as her most frequent partner. Earlier terms often show Ruth Bader Ginsburg alongside them; in later terms Ketanji Brown Jackson steps into that role. On the right, Clarence Thomas and Samuel Alito anchor the conservative pole, with Neil Gorsuch regularly aligned and Brett Kavanaugh and Amy Coney Barrett typically rounding out that bloc as the membership changes. John Roberts most often sits with the conservative side when I look at overall and majority behavior, but he is also the likeliest conservative to peel off in particular shadow-docket disputes. On the left, Stephen Breyer (earlier) and Elena Kagan sometimes appear nearer the center once ordinary majority voting is back in view. These anchor roles are visible not only in totals but also in the pairs of justices who repeatedly dissent together.

Dissent coalitions, term by term

I decided the coalitions by testing two possible splits—two groups or three—for each term and each non-overlapping three-term window. For each option, I measured how “clean” the split was using a standard quality check called a silhouette score (I also used these scores to evaluate merits coalitions in a prior article), which, in plain terms, asks for every justice: do I look more like the people in my group than the people in the next-closest group? I averaged those justice-level scores to judge the split, and I picked the number of coalitions (2 or 3) that produced the higher average. When the two options were essentially tied, I chose the simpler split (two groups). In all cases, every justice who was on the Court in that period was included in the chosen coalition map.

When I focus exclusively on dissent, I am asking a simple question: who objects together when the Court acts? Most years resolve into two clear blocs. The liberal side typically features Sotomayor and Kagan together, with Ginsburg playing that part in earlier years and Jackson in later ones. The conservative side regularly centers on Thomas and Alito, often with Gorsuch, while Kavanaugh and Barrett tend to join depending on the term; Roberts is more variable in this dissent-only view. Some years split three ways. That usually happens when a hardline conservative cluster (for example, Thomas/Alito/Gorsuch) separates from a more institutional conservative cluster (Roberts/Kavanaugh/Barrett), or when the liberal side divides between a sharper dissent and a narrower one. This dissent lens gives the sharpest picture of ideological separation because it is built entirely on disagreement with the Court’s action.

Dissent coalitions, three-term windows

Looking at non-overlapping three-term windows smooths year-to-year noise and shows which alignments persist rather than popping up once. The same anchors recur across adjacent windows: Sotomayor/Kagan (with Ginsburg earlier or Jackson later) on the left and Thomas/Alito (often with Gorsuch) on the right. Membership at the edges turns over with appointments and the case mix, but the poles remain clear. This medium-run view confirms that the dissent groupings are not artifacts of a single term.

Coalitions based on majority and dissenting votes

What I count and how I weight it

Below I look at all shadow-docket choices the justices make together in cases with dissents—dissents, would-grant/deny signals, and inferred ordinary majority votes—and I give them different strengths. When two justices make the same choice in dissent, I treat that as the strongest sign they’re traveling together and count it three times. When they make the same would-grant/deny choice, I count it twice. When they make the same ordinary majority choice, I count it once. Then I group the Court into two or three coalitions for each term and for non-overlapping three-term windows, picking the split that most cleanly keeps the “often vote together” pairs in the same group.

What the groupings show.

With this weighting, the familiar poles remain visible. The left-leaning coalition is anchored by Sotomayor with Kagan alongside her; earlier windows often include Ginsburg, and later windows add Jackson. The right-leaning coalition centers on Thomas and Alito, frequently joined by Gorsuch, with Kavanaugh and Barrett rounding out that side in most recent terms. Roberts most often lands with the conservative coalition here, but the weighting allows me to see when he moves closer to the center in years where his ordinary majority voting aligns with conservatives while his occasional case-specific signals pull slightly away.

Why the weights matter.

Because dissent is a stronger statement than going along with the Court, giving it triple weight keeps the sharp ideological splits front and center. At the same time, counting would-grant/deny at double and ordinary majorities at single strength stops the map from being driven by rare flare-ups alone. The result is a coalition picture that reflects both who breaks together and who generally travels together across the run of shadow-docket actions.

How this differs from the other lenses.

Compared to the dissent-only view, this weighted map draws the poles a bit closer, especially for justices like Roberts (and at times Kavanaugh and Barrett) on the right and Kagan (earlier Breyer) on the left, who often rejoin broader company in ordinary orders. Compared to the majority-only view, the weighting preserves the edge cases where alliances harden, so I don’t lose sight of the sharper ideological breaks that define each bloc’s identity.

Single Terms

Three Year Windows

Where the left–right line bends

Finally, I call out the cases where at least one conservative and one liberal dissent together. These moments are rare but revealing. They flag issue-specific fractures—places where the legal frame cut across the usual alignment on the shadow docket. In the discussion that follows the figures, I briefly name the standout terms and a couple of illustrative matters to show how and why the axis bent in those instances.

These 16 cross-ideological dissents make up about 4% of the total dissents in this dataset. This next graph breaks the counts of cross-ideological dissents down by Term.

The story in short

Across lenses and years, I see a stable two-pole Court in the shadow docket with occasional three-way splits. In dissent, the map is sharpest: Sotomayor/Kagan(/Ginsburg or Jackson) on one side and Thomas/Alito(/Gorsuch) on the other, with Kavanaugh/Barrett usually extending the conservative pole and Roberts sometimes peeling off in specific disputes. When I add in would-grant/deny and ordinary majority votes, those poles persist but draw a bit closer — especially for Roberts and sometimes Kavanaugh/Barrett, and for Kagan (earlier Breyer) on the center-left. In majority-only views, the governing coalition sits firmly on the conservative side. The handful of cross-ideological dissents punctuate the narrative and deserve short, case-focused callouts.

What makes this analysis novel is that it maps how the Court actually behaves on the shadow docket rather than relying only on votes in argued merits cases. On the shadow docket, the justices act under time pressure and with limited briefing, so their patterns of dissent, would-grant/deny signals, and quiet majorities reveal who moves together when it matters most. The coalitions I identify show a durable left–right structure, but they also surface subtler instincts—who is more willing to intervene, who defers to lower courts, and who occasionally crosses the aisle on emergency posture. This gives a practical picture of the Court’s operational majority and its fault lines that merits decisions can blur with full opinions and months of deliberation. For those interested in the justices’ behavior, these coalitions help explain—and often predict—how emergency applications and other shadow-docket actions will break, offering insight you can’t get from merits votes alone.

---

To share, subscribe, or comments