Rep. Hakeem Jeffries has accused the Supreme Court of being a rubber stamp for President Trump, and he is wrong. He escalated the line last week into a demand for “dramatic Supreme Court reform,” branding the conservative majority a “subsidiary” of the MAGA project. The Marquette Law School poll released this month gives the lie to the slogan — but the structural critique Jeffries gestures at deserves better than a slogan too, and the audit of it is more honest than the sloganeering.
The Marquette Law School Poll, surveying 1,076 adults nationwide between August 4 and August 14, asked respondents to weigh in on six specific rulings, and on every one of them a majority of Americans approved of what the Justices did. Three were signature Trump losses. Learning Resources v. Trump, the “emergency” tariffs case, was favored 66 percent to 34 percent. Trump v. Barbara, the birthright citizenship ruling, drew 63 percent support against 37 percent. Trump v. Cook, the decision blocking Mr. Trump from firing a Federal Reserve governor, won the country by 70 percent to 30. Three signature Trump initiatives, three public victories for judicial independence. The other three rulings the poll tested cut differently — and the Court won there too. West Virginia v. B.P.J., upholding state laws to protect girls’ sports, drew 65 percent. U.S. v. Hemani, the Second Amendment ruling striking down the ban on marijuana users owning guns, was favored by 61 percent. Louisiana v. Callais, 608 U.S. ___ (Apr. 29, 2026), the race-neutral redistricting decision, pulled 53 percent. (Per-ruling question wording, methodology, and full cross-tabs are in the Marquette Law School Poll’s August 2026 release, published at law.marquette.edu/poll.) These are not the rulings of a captive bench. They are the rulings of a Court that has, repeatedly, told the elected branches — including the President — no.
The strongest version of the Jeffries critique deserves to be set out before it is set aside, because the Marquette poll does not reach it. The critique is not a tally of signature losses. It is a structural argument about the Court’s use of what Stephen Vladeck named the shadow docket — the emergency-docket procedures by which a majority can issue orders without merits briefing, oral argument, or signed opinion. The documented evidence behind the critique is real. The Brennan Center’s 2025 tally put the second Trump administration’s emergency-application win rate at roughly 80 percent. The first Trump administration’s rate was 28 percent. Across the Biden, second-Bush, and Obama years combined, it was a fraction of either. The Court’s emergency orders have kept immigration enforcement operating on terms the executive has set, have blessed third-country removals without notice, and have permitted the removal of agency heads Congress designed to be independent. The cert-docket denials compound the picture: the Court has declined the merits petitions raising the very removals the shadow docket has blessed, has signaled it does not wish to confront the state-court redistricting questions on which it granted review in Callais only to resolve them narrowly, and has let Voting Rights Act enforcement actions go begging. The architecture across the term, on this reading, is not random: the Justices have used the procedures that do not require accountability to deliver outcomes the merits docket would have produced differently. That is the critique Jeffries is making. The Marquette numbers do not reach it, and a “rubber stamp” charge overstates it.
The audit, against the primary record. The 80 percent figure is not in serious dispute; it is the documented asymmetry of the emergency docket. What the audit does not show is that the merits docket has followed. The same term that produced the emergency-docket grants also produced Learning Resources v. Trump, Trump v. Barbara, and Trump v. Cook — three signed opinions, joined in places by Justices who would have had every institutional reason to defer, holding against the President on statutory grounds the administration had framed as core executive authority. The Federal Reserve firing question in Trump v. Cook cut hardest: the Court’s intervention preserved the independence of a statutorily designed agency, against an executive determined to remove its governor. The seventy-percent public approval reflects a country that read the merits ruling and saw what it saw. The shadow-docket critique therefore does not survive translation into the broader claim. The emergency procedures have done work the merits docket would not have done, but the merits docket has not been captured. The two are not the same thing, and the column that conflates them sells the same slogan Jeffries is selling.
The poll asked one more question worth reporting. When asked how often the Justices are ruling in Mr. Trump’s favor in his second term, 12 percent of respondents said “almost always” and 43 percent said “most of the time” — 55 percent combined. That perception is the only thing the Democratic complaint has going for it, and the source of that perception is the complaint itself. When your side’s loudest voices keep telling voters the Court is a Trump tool, you should not be surprised that voters believe the Court is a Trump tool. You built that. The Marquette poll also captured the partisan lens through which the public reads the Court: 74 percent of Republicans approve of the Justices’ job performance, compared with 22 percent of Democrats. That polarization is real, and the conservative majority has set itself against the practice of treating public approval as a relevant input. Justices see their job as following the law, not the vox populi.
Most Americans do not follow the Court closely. One-third of Marquette’s respondents said they had “never heard of” Chief Justice John Roberts. That is the public-information problem, not the legitimacy problem. Survey results always have to be taken for what they are worth. What they are worth here is clear: three Trump losses on the merits that the public read and approved, a shadow-docket pattern that is real but does not add up to capture, and a Democratic minority too invested in a slogan to tell the difference between a tally and an audit.
The next time a politician demands “dramatic reform” of the Supreme Court, the appropriate response is to hand him a copy of the Marquette poll and ask whether he has read the merits docket. The country, when it actually hears what the Court did, approves. The Court is not anyone’s subsidiary. It is the one branch still trying to do its job, and a slogan is no substitute for an audit.