

The Court let stand an injunction against the USPS implementing the rule in the 2026 midterms.
As I noted when the Supreme Court ruled in late August in favor of Donald Trump’s executive order (EO) on mail-in ballots and threw out a premature injunction against the U.S. Postal Service’s even writing a rule to comply with the EO, things would be different once the Court was presented with an actual rule from the U.S. Postal Service (USPS). When the case, now captioned United States Postal Service v. California, came back to the Court in September, the USPS noticeably failed to include in its proposed rule anything that would address the actual point of the EO: specifically, a prohibition (unauthorized by Congress) on delivering ballots that the federal government believed to be cast by noncitizens. So what’s the point?
This evening, the Court let stand an injunction against the USPS implementing its rule in the 2026 midterms. While Judge Indira Talwani of the District of Massachusetts has engaged in her usual ridiculous overreaches in this case, the emergency order refusing to stay her injunction got four votes from justices who had joined the opinion in favor of the EO: Chief Justice John Roberts and Justices Neil Gorsuch, Brett Kavanaugh, and Amy Coney Barrett. The order explained none of its reasoning, but Kavanaugh wrote a short concurring opinion explaining his view that the effort to implement the rule at the last minute, given the chaos that this will entail for states trying to comply with new USPS standards for ballot envelopes at the last minute: “There is at least a fair prospect that the final rule falls within the Postal Service’s statutory authority. . . . But applying the rule in the 2026 elections would be arbitrary and capricious in violation of the Administrative Procedure Act because state and local election officials do not have sufficient time to reasonably implement the rule before the elections.” Unstated is that the administration never made a case for why it was so urgent to change the envelope-design rules if it wasn’t actually going to use them to stop noncitizens from voting.
Even given the limited proper judicial role in second-guessing the rationales of agencies under the APA, Kavanaugh is on solid ground saying that the USPS insistence on rushing this rule into effect for the midterms — as opposed to proceeding with proper care to get the process right ahead of 2028 — would struggle to meet the APA’s standards given the short time entailed, the administrative complexity of the issue, and the lack of any expressed reason for urgency by the administration. But that just leads Kavanaugh and the majority into another jurisdictional thicket. As Justice Samuel Alito’s dissent (joined only by Justice Clarence Thomas) observed, Postal Service rules are exempted in many (but not all) cases from APA review. There are two key statutes: 39 U.S.C. § 410, in the statutory section on the general powers and liabilities of the Postal Service, exempts it in most cases from the APA unless some other statute applies; but 39 U.S.C. § 3001(m), within the statute on the power of the USPS to declare some things “nonmailable matter” (i.e., to exclude things from the mails), states that “proceedings concerning the mailability of matter . . . shall be conducted in accordance with” the APA.
Is this suit about mailability? Not openly, because the USPS isn’t purporting to do what Trump ordered it to do: exclude noncitizen mail-in ballots from the mails. But kinda, because any ballot that doesn’t comply with the ballot-envelope standards won’t be delivered. Alito may have the better of the argument that this isn’t really a mailability standard under Title 39, because it’s not saying that the contents of the envelope rather than its form are the reason that any ballot is nonmailable. As Alito argued:
It is also not clear that the Postal Service created a new class of nonmailable matter, let alone that it is specifically prohibited from doing so. . . . The rule’s requirements appear more similar to other threshold requirements for mailing — such as the basic requirement to list the recipient’s address on one side of the package or stamping requirements — that do not render noncompliant mail “nonmailable material.” Reading the nonmailable material provisions to prohibit the rule’s conditions on mailing would allow niche provisions to swallow the Postal Service’s otherwise broad power to regulate the delivery of mail.
Alito also has a point that there’s a separate problem here with primary jurisdiction, an administrative-law doctrine that some complaints about administrative action can’t be brought to court until they have first been raised with the agency, which has primary jurisdiction over the issue. Here, if this is a dispute over mailability, that agency would be the Postal Regulatory Commission. The law has been fairly favorable to the USPS over the years in arguing (I know this from having represented the USPS in such a case) that some complaints need to go to the PRC first.
On the other hand, Alito is grasping at straws when he says,
The Government has a strong interest in enforcing the rule, and implementing it will also enhance the visibility of Federal Ballot Mail in order to better detect election fraud. . . . The plaintiff States invoke the practical effects of implementing the rule close to the midterm elections. I take that problem very seriously, but it is not enough to convince me to deny the application” (quotation and citation omitted).
The practical challenges are immense, and by stripping the rule of any teeth, the government has eviscerated its own claims of any strong interest here. That said, Alito rightly called out Talwani, the plaintiff blue states, and (implicitly) the liberal justices who dissented from the original order for playing a major role in pushing this dispute so close to the election that the USPS couldn’t practically implement its rule:
The plaintiff States and the courts bear a substantial share of the blame for the rule’s timing. In April, the plaintiff States sued to challenge the President’s internal directive to subordinates mandating a rulemaking. For months, this litigation cast a cloud over the Government’s ability to issue a final rule. The District Court entered several injunctions months later prohibiting the Government from undertaking that rulemaking. Not only was the Government subject to months of premature litigation and a series of injunctions that this Court later determined were likely issued without jurisdiction, the Government had to wait nearly another full month for this Court to issue its decision so holding. Indeed, this Court spent from late July to the end of August drafting nearly 40 pages of opinions, including two dissents, despite the Government’s public notice that it would not publish a final rule while subject to the injunctions . . . and its representation that it had a strong interest in issuing its rule in early- to mid-August to allow time to implement it. The plaintiff States cannot claim the timing of the rule tilts the equities in their favor when they, and courts hearing their claims, are responsible for so much of the delay (citation omitted).
All told, given that the EO was originally designed to create new mailability restrictions without statutory power to do so, and given that the final rule doesn’t even try to pursue any such important goal, it seems the more prudent result for the Court to avoid the emergency step of overturning Judge Talwani’s injunction. With more time to get things right, the USPS may well be able to get its rule in place for 2028. But once again, the collision between a recklessly overreaching Trump order and a recklessly overreaching and politically calculating rogue district judge has driven the Court into deep jurisprudential waters on the limits of its own legitimate authority that are poorly navigated on short notice.