1
Do Republican appointees on the Supreme Court tend to vote with President Trump? Do Democratic appointees tend to vote against him? Are their votes in particular cases affected by the political party of the current president?
These are my topics here. My guess is that some readers have strong priors on the right answers.
2
One way into these topics is to ask: What if we had a simple algorithm, suggesting that Republican appointees vote for President Trump, and Democratic appointees vote against him. Call it the Simple Partisan Heuristic. How well would it do?
I suspect that it would do very well. I have not done that research (yet), but I asked Claude, which said that in 2025 and 2026, it did very well indeed - scarily well.
The heuristic appears to do less well for some justices than for others. According to Claude, it did pretty well for all of them, mind you, and it did very well for most of them.
I am not reporting Claude’s actual numbers, because I have not checked them myself. So let’s take them as suggestive only - an invitation to get the actual numbers.
We also have to introduce a cautionary note. Suppose that the objectively right vote, in these cases, is that Trump loses almost all the time. Or suppose that the objectively right vote, in these cases, is that Trump wins almost all the time.
If either of these statements were true, our evaluation of the performance of the algorithm would be affected accordingly. In other words, it is hard to know what to say about the accuracy of the Simple Partisan Heuristic without knowing the objectively right answers to the legal questions (supposing that they have such answers).
So we have to be careful about conclusions. I am going to try to say a bit about what we know and what we do not know.
3
Members of the Supreme Court disagree about many things. Some of the justices are originalists; some are not. Some of the justices have long been committed to the idea of the unitary executive; some have not.
Some of the justices have long had a strict view of “standing”; some have not. Some of the justices have much liked the “major questions doctrine”; some have not.
All of these differences transcend narrowly partisan disagreements. Someone who is committed to the unitary executive should rule for presidential removal authority; it should not matter whether the president is named Biden or Trump. Someone who has a strict view of standing should insist on that view regardless of whether the president is named Obama or Bush.
4
A number of years ago, Tom Miles and I did a study of the role of politics and partisanship in administrative law cases. See https://chicagounbound.uchicago.edu/cgi/viewcontent.cgi?article=5371&context=uclrev
(It’s riveting reading, which is one reason that it’s become a Netflix series starring Jessica Chastain and Tom Hiddleston, with a soundtrack by Phoebe Bridgers.)
As shown in Table 1 on p. 832, we did not find much in the way of political partisanship on the Supreme Court. True, Justices Ginsburg and Stevens were a bit more likely to vote to validate agency decisions under Democratic than Republican presidents, and Justices Scalia and Thomas showed the opposite pattern. But in all four cases, the differences were not statistically significant!
The basic finding is: The identity of the president did not matter much to how the justices voted.
True, you could complicate the story in various ways. Here’s one: What if you code the agency decisions as (1) liberal, and challenged by (say) a corporation, or (2) conservative, and challenged by (say) a public interest group?
We tested for that - and indeed found significant political differences. Justice Thomas was significantly more likely to vote to strike down liberal agency decisions, and Justice Ginsburg was significantly more likely to vote to strike down conservative agency decisions.
Still, you have to be careful. Maybe Republican presidents oversaw far more (objectively) unlawful decisions than Democratic presidents did, or vice-versa. Or maybe liberal agency decisions are far more likely to be (objectively) unlawful than are conservative agency decisions.
History is only run once, which makes the analysis of partisanship hard. If a justice rules against Democratic administrations a whopping 90% of the time, and against Republican administrations just 10% of the time, we cannot know that the justice is partisan.
(Still, I say that we can, and should, suspect so.)
5
The Court’s decision of August 31, 2026, allowing the Trump administration to go forward with the East Wing project, was 5-4. Roberts, Kagan, Sotomayor, and Jackson were in dissent. Gorsuch, Alito, Barrett, Thomas, and Kavanaugh were in the majority.
I bet that you are not shocked. My own reaction is: That particular split is not good at all. (Or more precisely: Yuck.)
What should we do with that split? Notwithstanding the yuck, there is a good argument that the right answer is: Not much. The issue was one of standing, after all. The observed split is easily explained as a split on standing, and not on anything else.
Is that argument clearly correct? No. We would need to run history another time, with a Democratic president (say, AOC) doing exactly what President Trump did, and then ask: How does the standing issue come out then?
Maybe the answer is: The same 5-4 lineup. Or maybe not.
6
I confess that I much prefer to think that the divisions on the Court are not much affected by the political party of the current president. I want to think that. It makes me sad and upset not to think that.
Still:
In the presidential immunity case, the division was 6-3 in favor of immunity for President Trump. Is it horrible to ask whether we would have seen the same line-up if the case had involved President Biden?
In the tariffs case, the division was 6-3 against President Trump, with Alito, Thomas, and Kavanaugh in dissent. Is it horrible to ask whether we would have seen the same line-up if the case had involved President Biden?
In a series of cases, the Court ruled against President Biden on “major questions” grounds, sometimes with 6-3 divisions, with Kagan, Sotomayor, and Jackson in dissent. Is it horrible to ask whether we would have seen the same line-up if the cases had involved President Trump?
Recall: In a large number of cases in 2025 and 2026, Trump administration decisions were upheld (or stuck down), and if you predicted judicial votes with a simple algorithm that said “Republicans appointees vote for the Trump Administration” and “Democratic appointees vote against the Trump Administration,” you would do alarmingly well. As noted, Claude ran the numbers and found that the algorithm would do very well indeed - scarily well.
With a view to the last point in particular, a plausible reaction is: Uh oh.
Still, we have to be careful. A lot of those cases involved domains such as immigration, where we might expect to see the observed differences independent of the political party of the current president. (And we could say the same thing about the independent agency cases.)
7
Ok then.
The following proposition is clearly correct: Many of the differences between Democratic and Republican appointees have nothing at all to do with the political party of the current president. That is, the justices have views on standing, on ripeness, and on executive power, and their votes on those issues are unaffected, or very little affected, by the political party of the appointing president.
The following proposition seems right: For both Democratic and Republican appointees there is an overlap, which is hardly a coincidence, between political judgments and constitutional judgments; consider gun rights, abortion rights, and affirmative action. I realize that explaining the overlap, and the phrase “hardly a coincidence,” is not a simple matter. Still, these are not partisan differences, and the political party of the current president does not drive votes in the relevant cases.
I believe the following, deeply (annoyingly?) cautious claim to be true: In some cases, the votes of Democratic and Republican appointees are affected by the political party of the current president.
But I do not know how to prove that claim, and I do not know the precise meaning of the terms “some cases” and “are affected.” I am hopeful that careful empirical work could help.
Part of the test should be whether justices depart from their normal jurisprudence in such cases. Some examples stand out in this regard. The votes of some conservative justices to support Trump's position in the birthright citizenship and tariff cases seem fundamentally at odds with core conservative principles--textualism, originalism, and the major questions doctrine in the tariff case. Also, while not Trump-specific, all conservatives justices voted to create presidential criminal immunity out of essentially nothing in the Constitution and in contradiction of what relevant text and background the Constitution provides on this subject. On the other hand, yesterday's decision by five conservatives in the ballroom case based on lack of standing is right out of the conservative playbook, although Roberts' dissent is noteworthy.
The question is what motivates them to vote along party lines. Is it really adherence to the Democratic or Republican label ? Or would it be more accurate to say conservative justices tend to vote for the more conservative result and liberal justices for the more liberal result? My perception is that ideological orientation is far more important than political affiliation, but that today, unlike previous political eras, party and political philosophy tend to overlap almost completely.