This is the sixth piece in a series on what I believe the court can, and should, do better.
***
On the Supreme Court, there are two primary numerical voting rules. Under the “Rule of Four,” it takes four votes to grant a petition for a writ of certiorari. Thus, less than a majority is needed to accept a new case. And under the “Rule of Five,” it takes five votes to form a majority opinion, stay or enjoin a lower court ruling, or enter a judgment. As Justice William Brennan may have remarked, with five votes, anything can be done at the Supreme Court.
I will propose numerical rules based on votes of three and votes of six. This sixth installment of this series will talk about the Rules of Three. The next installment will talk about the Rules of Six. A minority of three justices could force certain procedural matters, while a supermajority of six can force certain substantive matters.
The chief justice can propose these rules, and will play an important part in enforcing them, but these rules should be adopted by the court unanimously. There must be complete buy-in, similar to the recent ethical rules adopted. The minority would have to value their procedural advantages, at the cost of giving the majority their substantive advantages. At the same time the majority would have to value their substantive advantages, at the cost of giving the minority their procedural advantages.
The Rule of 4
Under the so-called Rule of 4, it takes four votes to grant a cert petition. Yet every term, there are a host of cases where one, two, or three justices dissent from the denial of certiorari. On the left, Justices Elena Kagan, Sonia Sotomayor, and/or Ketanji Brown Jackson will sometimes dissent when the court leaves standing a conservative decision. And on the right, Justices Clarence Thomas, Samuel Alito, and/or Neil Gorsuch will sometimes dissent when the court refuses to review a liberal decision. (Justice Brett Kavanaugh is increasingly likely to dissent.) These dissents must be considered against the backdrop of the Supreme Court’s shrinking docket. Recently, the court has decided about 60 cases per term. By contrast, at the start of Thomas’ tenure, that number was closer to 100. And when Chief Justice John Roberts clerked, that number was above 150.
Roberts himself seems content with the number of cert grants. He has not dissented from the denial of certiorari in nearly a decade, and only a handful of cases before that. Justice Amy Coney Barrett seems to be in a similar position. She has only ever dissented from the denial of certiorari once. Indeed, Barrett defends the low number of cases that the court grants. At a book talk in September 2025, Barrett relayed that she asks her clerks every year whether there are more petitions that actually warrant review but are not granted. Barrett said that her clerks cannot identify any such missed cases. (In doing so, Barrett casts doubt on the cases where three of her colleagues dissent but Barrett declines to be the fourth vote.)
With the current composition of the court, the number of grants is therefore not likely to increase. But an important change in protocol could achieve that goal.
The Join-3
At their private conferences, the justices vote in the order of seniority, with the chief justice speaking first. Each justice can vote to grant or deny a petition. Justices can also ask to reschedule or relist a petition. These steps can be taken to provide further time for consideration, or to potentially convince others to grant cert.
There is another option: the Join-3. In short, if three other justices are willing to grant a case, the Join-3 provides a fourth vote for certiorari, known as a “courtesy fourth.” The Join-3 is not a vote to grant and is given without regard to the merits. Rather, the Join-3 is primarily an act of collegiality. Scholars observed that the Join-3 “originated during Chief Justice Burger’s ascension to the Court” and “became firmly established by the mid 1970s.” For example, in the landmark case of Chevron v. National Resources Defense Council on court deference to agencies, Justice Sandra Day O’Connor provided a “join 3” to grant review. In recent years, as the docket has shrunk and the number of dissents from denial have increased, it seems like the Join-3 has fallen by the wayside. As Sarah Isgur observed in Last Branch Standing, “courtesy fourths” are no longer even provided for “ideological allies.”
Unlike most changes that require four or five votes, a single justice can change course. To start, I propose that the chief justice should offer a guaranteed Join-3. In any case where three other members vote to grant certiorari, the chief would push the case over the cert-threshold. And since the chief speaks first at conference, he would give the Join-3 without knowing how his colleagues will vote. This move would promote collegiality. As Judge Harry Edwards has explained, “collegiality plays an important part in mitigating the role of partisan politics and personal ideology.” Moreover, the court’s minority would have greater control over the docket.
The Join-3 should also reduce the trend of never-ending dissents from denial of certiorari (known as a “dissental”). Trios of justices could no longer complain that the court is unwilling to grant an important case. This approach would also save a lot of time and effort. Often, justices may take months trying to scrounge up a fourth vote for cert. There is a persistent cycle of reschedules and relists that can stretch up to two years. I call this period “docket purgatory.” And when that process fails, even more time is taken to prepare a dissent from denial of certiorari. The Join-3 cuts through most of those delays and gets the docket moving. The Join-3 also will limit strategic denials, where justices band together to try to block a case from the docket. Finally, since this reform is unilateral, there is no expectation that other justices must comply.
There is, of course, an obvious admonition: be careful what you wish for. Cases granted through a Join-3 are far less likely to have five votes to reverse. The most likely outcome is an affirmance. The dissenters would have to choose between having an unfavorable circuit precedent and establishing an unfavorable SCOTUS precedent. At some level, when the justices dissent from the denial of certiorari, they may not actually want the court to grant review but instead want to grouse. Some such dissents amount to performative virtue signaling. The Join-3 will call the minority’s bluff.
Pre-3
If the automatic Join-3 catches on, there may be a way to expedite the process further: the Pre-3. Under the longstanding practice, the justices will only discuss cases on the discussion list. As Barrett explained in her book, Listening to the Law, “Any justice can list a petition to be discussed and voted on at conference.” This process could be tweaked. Instead of simply voting to add a case to the discussion list, a justice could pre-vote to grant certiorari. If there are three votes to grant, by operation of the Join-3, the case is added to the docket. No further deliberations at the conference are necessary.
DIG-3
As more cases are granted under the Join-3, the chances of a false positive increase – that is, a case that appeared to be certworthy but had issues because of procedural glitches or factual gaps that prevent a clean resolution of the legal issue.
This problem exists under the current rules. Often, the petitioner, hopeful for a grant, presents the case in a limited fashion in the cert-stage briefs. However, once cert is granted, the petitioner changes the case to address other issues. When the justices see such a bait-and-switch, they will often dismiss the case as improvidently granted, or “DIG” it. It is something of an academic curiosity how many votes are needed to DIG a case. It takes four votes to grant, but it might take five votes to DIG. (In 2010’s Boyer v. Louisiana, four justices dissented from the DIG. This practice suggests it takes five votes to DIG a case.)
In any event, the DIG usually occurs after the case is argued. In recent terms, there have only been a handful of instances where a case was DIG’d after briefing concluded but before oral argument. In 2016’s Visa v. Osborn, for example, the court observed that “[a]fter ‘[h]aving persuaded us to grant certiorari” on this issue … petitioners ‘chose to rely on a different argument’ in their merits briefing,” and thus DIG’d the case.
There is a benefit to DIGing a case before oral argument. It saves time and resources on the court’s crowded schedule. I propose a method to make it even easier to DIG a case after the merits briefs are filed, but before oral argument. The DIG-3 would work something like a motion for reconsideration. In Congress, and in most parliamentary bodies, a motion to reconsider can only be made by someone who voted on the winning side in the initial vote. At the Supreme Court, if any three members of the court who had voted for certiorari decide, following briefing, that certiorari was improvidently granted, the case should be DIG’d. The chief justice could also provide an automatic fourth vote for a DIG by virtue of his Join-3, although I don’t think that would prove necessary.
The Record-3
With signed opinions, each justice signals how they voted. But with other matters, the justices may not signal their votes. For example, when certiorari is denied, it is usually impossible to figure out the exact vote count. Even where three justices dissent from the denial of certiorari and one or more justices concur in the denial of certiorari, there will usually be a justice whose votes cannot be determined with certainty. Likewise, on the emergency docket, it is often unclear exactly how many votes are granted for relief. To be sure, five votes are needed for an emergency stay or an injunction. But if there are only three noted dissents, it would be unclear which five justices joined the majority. Only where there are four published dissents can you count the members of the majority.
There might be a way for the justices to address this phenomenon. Here, I look to the Constitution’s journal clause. It provides, “the Yeas and Nays of the Members of either House on any question shall, at the Desire of one fifth of those Present, be entered on the Journal.” This clause makes it harder for members to cast votes in secret and perhaps evade accountability. A minority, only 20%, could demand that the yeas and nays are recorded on any matter.
The court could adopt a new rule, the Record-3: Three members could require the publication of any vote. If there is a cert denial, the minority could require the members of the majority to signal how they voted. And if there is an emergency application, the minority could force their colleagues to go on record. One might argue that this rule allows for naming and shaming. Then again, if a justice votes for or against a party, they should be accountable for that vote. But given how effective moral suasion can be, this remedy should be used carefully so as not to diminish collegiality.
The Conference-3
So far, all of the above rules concern the Supreme Court’s merits docket. Yet, increasingly, much of the court’s work occurs on the emergency docket, also known as the interim orders docket, also known as the “shadow” docket. Whatever you call it, all agree that procedural reforms for fast-moving litigation would be welcome.
A common criticism of the emergency docket is that complicated cases are decided without fulsome deliberations. The justices usually don’t hold oral argument on emergency cases. Indeed, they may not even hold private conferences to discuss the outcome. In 2025’s Department of State v. AIDS Vaccine Advocacy Coalition, Kagan dissented from an emergency docket order. She wrote, “Per usual on our emergency docket, we have had to consider this application on a short fuse—less than three weeks.” Kagan added, “We have done so with scant briefing, no oral argument, and no opportunity to deliberate in conference.”
At the time I asked why the court couldn’t schedule emergency docket conferences. I wrote that “Roberts could convene a conference on short notice,” but failed to do so, even as three of his colleagues demanded one. As things stand now, we can only surmise that the justices decided the entire matter over email or through the circulation of memos. Indeed, we learned from the leak of the “shadow papers” that the Supreme Court resolved the Clean Power Plan case in 2016 entirely through memos. This is no way to resolve questions affecting the national interest.
The Conference-3 rule could address this situation. When any emergency application is filed, or is even imminent, three justices could call for a conference to be held within 24 hours. That conference can be hosted virtually, or in person, depending on the justices’ schedules. The purpose of the conference would be to discuss the pending application. Critically, this conference will likely be held before the briefing even concludes. It can even be held as the case percolates through the lower court. The justices can be proactive rather than reactive, and discuss what the briefing schedule should be, whether an administrative stay should be issued, whether oral argument is appropriate, if cert before judgment should be granted, and what the ultimate disposition should be. If, at the end of the conference, the court splits 6-3 or 5-4, all of the justices will at least have been heard. I think this modest effort would go a long way to improving emergency docket collegiality.
This approach might have to be applied differently with capital cases. For example, in April 2019, the Supreme Court declined to stay the execution of Christopher Price in Alabama by a vote of 5-4. Justice Stephen Breyer dissented, and explained the chronology. He “requested that the Court take no action until” its regularly scheduled conference the following day. His conservative colleagues disagreed. They wouldn’t wait a few more hours. With the Conference-3, the minority could insist that a conference be held before the matter is resolved.
Death penalty cases, though, have a wrinkle. Death warrants usually expire at a certain time, typically at midnight on a particular date. Appeals are timed so that the case gets to the Supreme Court shortly before the warrant expires, forcing the justices to make a split decision at the 11th hour. Even a 24-hour delay to hold a conference could force the warrant to expire, and restart the execution process by days, weeks, months, or even years.
I would therefore propose that the Conference-3 can only be used in capital cases if the 24-hour delay for holding a conference does not extend beyond the length of the death warrant. More often than not, the Supreme Court’s death penalty clerk knows that emergency cases are coming and can alert the justices in advance. Then, the justices could hold a conference before the appeal is even noticed. Again, they can be proactive rather than reactive to avoid the rush of late-night litigation.
The Argue-3
Relatedly, only some emergency docket cases will trigger oral argument. In response to this trend, Kavanaugh has urged the court to hold oral argument on an expedited basis. In recent years, there have been several prominent examples. This term, in Oct. 2025, the court scheduled oral argument in Trump v. Cook for January 2026. Last term, in April 2024, the court scheduled oral argument in Trump v. CASA for May 2025. And in December 2023, the court scheduled oral argument for Ohio v. EPA in February 2024. When the court wants to move quickly on holding oral argument for an emergency application, it can.
Right now, it is unclear how many votes are required to hold oral argument in an emergency docket case. I would propose an Argue-3 rule. With any pending emergency application, three justices could set a date for oral argument. The Argue-3 would ensure that even in a fast-moving case, the court has at least some time to ask questions of the parties. Based on past practice, these arguments can be scheduled within a month. Lower courts sometimes schedule emergency arguments in the span of days. I think the Supreme Court could move with similar speed.
This argument could be held-in person, if it occurs during a regularly scheduled sitting. Or the argument can be held virtually. I have proposed that the court hold oral arguments year-round to permit emergency arguments in person, but until that proposal is adopted, virtual arguments are a viable option. The Supreme Court successfully experimented with remote oral arguments during the pandemic. The lower courts routinely hold virtual oral arguments for emergency cases. It can be done.
However, the Argue-3 comes with a catch. It takes five votes to grant emergency relief. But with this proposal, three votes could keep the case on hold for a month or longer. To ensure the Argue-3 is not used for dilatory purposes, automatic administrative relief would be granted. For example, if the lower court issued an injunction, and Argue-3 is invoked, an administrative stay is automatically denied. In other words, if only three justices invoke the power to schedule the case for oral argument, those same three justices will have to be content with the lower-court judgment going into effect. After the oral argument, the justices can then vote on whether to grant emergency relief or take some other action.
These reforms would empower the Supreme Court’s minority on certain procedural matters. In the next installment, I will discuss how the Supreme Court’s supermajority could have greater authority over some substantive matters.
