First, some personal news: I sent off my completed book manuscript to my publisher last week(!).

I’ve been writing non-stop for two years and collecting data for the past decade, so sending it off feels a bit disorienting (but also great). As most readers of this newsletter already know, the book is focused on free speech and democratic backsliding worldwide. I’ll have more to say on that topic in the coming weeks, but it’s early July, which means U.S. Supreme Court roundup time. (Late June/early July typically marks the release of the last of the Court’s fully argued merits opinions, though as Steve Vladeck regularly reminds us, it doesn’t mark the literal end of the Court’s annual term.)

I recently had a great conversation with Matt Grossman about the Court and democratic backsliding. Our starting point was my 2024 paper on the topic, but a lot has happened since that paper was written, so I wanted to use this post to update my thoughts, by way of a roundup of the Court’s recent term through the lens of democratic backsliding. It’s my longest post to date; if your email reader clips it, click through to read it online.

In the 2024 paper, I draw on the comparative literature on courts and democratic backsliding and the American political development literature on democratic crises to assess whether, when, and to what extent the U.S. Supreme Court has served as an effective guardrail against autocratization.

The short answer is that the Court has not reliably filled this role. In most prior crisis periods that featured efforts by the governing regime to entrench their hold on power, the Court did not meaningfully push back. As to our current crisis, the paper concluded that the Roberts Court’s record was mixed to date. This assessment was perhaps overly generous even then, but the Court’s record has gotten significantly worse since the paper was published. In particular, I finished drafting the paper in early 2024, before the Court’s rulings in three cases related to January 6 that came down later that year.

Those three cases had nothing to do with one another legally—one turned on the original constitutional structure of federalism, one on the meaning of a key provision of the Reconstruction-era Fourteenth Amendment, and one on the scope of the key congressional statute enacted in response to the post-2008 financial crisis. It is probably not surprising to any of my readers that the Court drew on these three disparate legal sources for pro-Trump results across the board.

The most recent term featured more of the same. The Court ruled against President Trump on tariffs and birthright citizenship, but on issues directly related to Republican electoral fortunes, the Court has remained a reliable partner—most notably the Court’s continued gutting of the 1965 Voting Rights Act (VRA) in Louisiana v. Callais (2026) and its continued evisceration of campaign finance regulation in National Republican Senatorial Committee v. FEC (2026). On my reading, these election law cases qualify as examples of what Rosalind Dixon and David Landau call “abusive judicial review”—namely, the use of judicial review to undermine core components of electoral democracy.

On Dixon and Landau’s account, abusive judicial review comes in two variants. Its weak form occurs when courts fail to check actions by other power holders that undermine core norms and institutions of electoral democracy. The Court’s repeated failure to impose any meaningful checks on partisan gerrymandering is perhaps the clearest example here. (See in particular Rucho v. Common Cause (2019).)

In its strong form, abusive judicial review occurs when courts themselves undermine core norms and institutions of electoral democracy. This typically occurs when other power holders take actions meant to shore up and strengthen democracy, and the Court invalidates those actions in turn. The Callais and NRSC cases from the recent term are good examples here.

In addition to the activism/deference dimension highlighted by Dixon and Landau’s two varieties of abusive judicial review, I want to highlight an additional dimension on which these decisions vary. Some of the Court’s decisions undermining democratic health are best understood as the conservative justices fabricating legal arguments more or less out of whole cloth, in an effort to advance Trump’s agenda. But a second and more common dynamic involves the conservative justices’ stubborn adherence to longstanding commitments in the face of proliferating evidence that those commitments are enabling Trump’s autocratization.

With this frame in mind, consider some of this term’s decisions.

The conservative arguments in Trump v. Barbara, the birthright citizenship case, were fabricated out of whole cloth. Egalitarian citizenship irrespective of race or national origin is the central, defining commitment of the post-Civil War Constitution. This commitment is uncommonly clear in the text of the Fourteenth Amendment:

The commitment is also reflected in more than 100 years of SCOTUS precedent, as well as federal immigration law since the 1940s. Chief Justice John Roberts’s opinion for the Court in Barbara reaffirms this commitment without caveat. And Justice Ketanji Brown Jackson’s concurring opinion adds substantial additional historical detail.

Indeed, if you read just one thing from the Court’s recent term, I recommend this opinion from KBJ. It's got a lengthy defense of the anti-caste reading of the Fourteenth Amendment (with extensive citation to Isabel Wilkerson's Caste). It roots this reading in the constitutional claims of 19th century African Americans themselves. It offers a remarkable treatment of the Reconstruction framers’ discussion of Chinese & Roma immigrants (most of which I was unaware of). And it concludes with this passage that reminds me of Ta-Nehisi Coates's most recent piece:

All of that is good news. But from a democratic backsliding perspective, what seems most significant about Trump v. Barbara is that the Court recognized this central constitutional commitment by a bare 5-4 vote.

Until 18 months ago, there was no reputable Fourteenth Amendment scholarship suggesting that the President could withdraw U.S. citizenship from the U.S.-born children of undocumented immigrants. Thomas, Alito, Gorsuch, and Kavanaugh all wrote that the Fourteenth Amendment allows this policy, though Kavanaugh held that Trump would need congressional approval to make this change. In so holding, these four conservative justices cited multiple scholarly treatments of the Fourteenth Amendment, but all of them were published in the past year. Here’s Thomas:

One of Thomas’s sources is a Wall Street Journal op-ed. Another is a blog post. A third is a trade book (i.e., published for a public rather than scholarly audience). The other three are scholarly papers published in law journals. All six were published in 2026.

Thomas cites a handful of older sources elsewhere in the opinion, but the bulk of his scholarly support was crafted in response to Trump’s birthright citizenship policy, rather than providing preexisting support for the constitutional legitimacy of that policy. Recognizing this much, Chief Justice Roberts observed in the majority opinion that “there is scant evidence for this dramatically revisionist view.” Nonetheless, Trump came within one judicial vote of dramatically undermining the central commitment of the post-Civil War Constitution.

The clearest parallel I can think of is Bush v. Gore (2000), the case that decided the outcome of the 2000 presidential election, where this dynamic played out on an even more compressed timeline. The Bush campaign was desperate to stop the ongoing Florida recount, and the campaign’s legal team (which included John Roberts, Amy Coney Barrett, and Brett Kavanaugh) came up with a heretofore unheard of reading of the Constitution’s Article II elections clause in support. Three of the conservative justices adopted this argument, and together with two colleagues who adopted an altogether different argument, handed the election to George W. Bush.

To sum up so far, one recurring dynamic in Trump-era constitutional law is the conservative justices’ willingness to more or less invent legal arguments as necessary to suit the Trump regime’s agenda.

A second and more common dynamic is the conservative justices’ stubborn adherence to longstanding conservative doctrines despite proliferating evidence that these doctrines are enabling and facilitating Trump’s autocratization. In other words, the Roberts Court sometimes acts in pursuit of goals of the conservative legal movement that long predate the Trump era—and in this sense, the justices are not just “making stuff up” as in the birthright citizenship case.

Some of these conservative goals have been profoundly hostile to political equality or other democratic norms and institutions, and scholars have criticized them on those grounds for a long time. Voter suppression was bad before Trump came down the escalator, and it will still be bad long after he’s gone. But it seems to me even worse—in the sense of more harmful to our constitutional system—for the conservative justices to continue to pursue their voter suppression agenda when it now enables not just GOP partisan entrenchment, but active and ongoing efforts by the Trump coalition to turn the U.S. government into a Hungarian-style illiberal democracy. That is definitely worse, right?

From the Court’s recent term, this second dynamic is visible in the election law cases; in the decisions on presidential appointment and removal powers, rooted in so-called unitary executive theory; and in the LGBT rights cases.

Over the past week, folks have understandably been focused on the Court’s end-of-term decisions, but one of the most significant decisions issued earlier in the year was Louisiana v. Callais. Here, the Court upheld a Louisiana congressional map that included only one Black majority district (out of six), despite clear findings by the lower court that the map diluted Black voting power.

This decision was the latest in a string of decisions effectively gutting the Voting Rights Act. Here’s Justice Kagan on the recent history:

In fact, the conservative effort to gut the VRA dates back even further than Shelby County v. Holder (2013). In 1980, the Court held that sec. 2 of the Act applies only to intentionally discriminatory districting changes. Congress disagreed, revising the law to specify that it covered racially discriminatory effects, regardless of intent. Here’s a slide from my Constitutional Law course with the statutory language.

A young Reagan administration lawyer named John Roberts opposed these changes at the time, but President Reagan signed them into law, and the Court came around. Until Brnovich and Callais, when it again held that sec. 2 required evidence of discriminatory intent.

In short, the conservative legal movement has been gunning for the VRA for a long time. But it is now doing so in a context where Republican-controlled state legislatures are heeding President Trump’s explicit calls for aggressive gerrymandering in order to entrench his hold on power. This sort of partisan entrenchment is normatively bad even in the best of times; when the governing regime is actively undermining democratic norms and institutions, it is even worse. It is also a recurring feature of 21st century democratic backsliding worldwide.

A similar dynamic is visible in the Court’s campaign finance case law.

Since Watergate, Congress has periodically sought to rein in wealthy and corporate spending in federal elections. Congressional attention to this problem has been far from consistent, but during the Nixon era and again during the George W. Bush era, Congress tried to impose some limits.

Beginning with Buckley Valeo (1976), and most famously in Citizens United v. FEC (2010), the Court’s conservative justices have repeatedly held that campaign finance regulations violate the First Amendment. This conservative judicial campaign to undermine campaign finance law is even longer running than the campaign against the VRA. Most recently (just last week), the Court held that while Congress is still permitted (at least for now) to limit the size of donations to individual candidates’ campaigns, it may not restrict so-called coordinated expenditures by political parties, in which parties spend their own resources on behalf of candidates and coordinate with those candidates’ campaigns about when/where/how best to do so. The Court rejected Congress’s longstanding argument (reflected in the Court’s own precedent) that eliminating this rule will enable easy circumvention of the contributions limits to candidates.

On Justice Kavanaugh’s telling, the Court’s First Amendment stance enacted a kind of political equality:

Throughout this line of cases, the conservative justices have been obsessed with this sort of formal neutrality. But it is simply not true that an absence of state restrictions is always the surest path to equality. And Kavanaugh’s claim was immediately undermined when GOP officials started crowing about the huge structural advantage the NRSC decision gave them in this year’s midterm elections.

As with the VRA decisions, the Court’s campaign finance decisions have always had troubling impacts for political equality, but they now have the significant additional implication of enabling oligarchs like Elon Musk to bankroll the electoral campaigns of an autocratic coalition.

Another longstanding goal of the conservative legal movement reflected in last week’s decisions is rooted in unitary executive theory.

Drawing on the first sentence of Article II, conservative lawyers have been arguing since the Reagan era that the key feature of executive power under the U.S. Constitution is that it is vested in a single person.

On their telling, this provision means that any state action properly considered executive in nature is subject to the president’s discretion alone, and neither Congress nor the federal courts can impose limits on his exercise of that discretion. Or as Trump himself sometimes puts it:

The president obviously cannot literally conduct the operations of the national government by himself, and Article II authorizes him to appoint government officers, from Cabinet Secretaries on down. But a major tenet of unitary executive theory is that those officers are beholden to the president and hence removable by him at will.

Article II famously provides that ambassadors, public ministers, and all other officers of the United States are appointed by the president, “by and with the advice and consent of the Senate.” The text doesn’t mention how such officers are to be removed, and both the early republic and the Civil War/Reconstruction era featured extensive debate on that front.

But advocates of unitary executive theory think the question is clear, and the answer is the president’s discretion alone. Here’s Justice Scalia in a 1988 dissent, arguing that the post-Watergate independent counsel law, which provided a mechanism for appointing an independent prosecutor to investigate crimes by the president or other high-level executive branch officials, was unconstitutional:

The independent counsel law was unconstitutional, in Scalia’s view, because it empowered a federal prosecutor who could not be fired by the president. It did so for the obvious reason that Congress was trying to avoid a repeat of the “Saturday night massacre,” when President Nixon ordered the Attorney General to fire the Watergate special prosecutor. There are lots of federal statutes that empower federal officers who cannot be directly fired at the president’s discretion. Congress writes these laws for the obvious reason that some aspects of public administration should turn on expertise and independent judgment rather than partisan spoils.

One such statute is the one governing the Federal Trade Commission, which provides that the FTC shall consist of “five Commissioners, each of whom serves for seven years and may be removed by the President only ‘for inefficiency, neglect of duty, or malfeasance in office.’” (I’m quoting Roberts’s opinion for the Court in Trump v. Slaughter (2026), which quotes the relevant statute in turn.)

In early 2025, Trump fired the two sitting Democratic Commissioners, expressly declining to accuse them of any of the statutory prerequisites and instead asserting his Article II authority to do so without constraint. Rebecca Slaughter challenged her dismissal. The lower courts ruled in her favor, relying on the authority of Humphrey’s Executor v. United States (1935), in which FDR had done the same thing as Trump, and the Court had ruled against him. The Roberts Court reversed, holding that FTC Commissioners serve at the president’s discretion and can be fired at will.

As Justice Gorsuch noted in a concurring opinion, this decision has the effect of immediately placing the vast administrative powers of the federal government entirely in the hands of President Trump. (Except for the Federal Reserve, which the Court inexplicably but predictably subjected to a different set of rules in another decision issued the same day.) These administrative powers (and the agencies that exercise them) exist only because Congress has created them, and yet somehow Congress’s only options are to forego public administration altogether or to place it solely in the president’s hands.

Partisan hiring and firing of government officers has long posed risks to constitutional government. (If you’re not convinced, watch Death by Lightning on Netflix.) But in the context of twenty-first century democratic backsliding, it is particularly associated with efforts by autocratic leaders to bend election commissions, media councils, and other independent bureaucracies toward anti-democratic ends. (If you’re not convinced, read Tom Ginsburg & Aziz Huq.) In short, and as my colleague David Driesen has argued at some length, unitary executive theory has always been inconsistent with the Madisonian system of checks & balances, but it is now indisputably clear that it also facilitates autocratization.

As with election law and executive power, so also with LGBT rights. Conservative lawyers have long argued against constitutional protection for such rights. The Roberts Court has repeatedly enacted such arguments into law. And it has done so despite proliferating evidence that attacks on LGBT rights are a key recurring component of 21st century autocratization.

The two key cases from the recent term are Chiles v. Salazar (2026) and West Virginia v. B.P.J. (2026).

In Chiles, the Court struck down a Colorado law prohibiting the practice of LGBT conversion therapy—i.e., efforts by mental health providers to help LGBT youth change their sexual orientation or gender identity to match societal norms. In the Court’s framing, this is a free speech case. Indeed, together with the campaign finance case discussed above, Chiles signals the Roberts Court’s continued efforts to remake the First Amendment as a tool on behalf of conservative interests.

But Chiles also signals the Court’s increasing hostility towards LGBT persons, especially when read together with B.P.J., the Court’s June 30 decision upholding state legislative bans on the participation of trans students in sports.

In B.P.J., the Court upheld West Virginia and Idaho laws barring transgender girls from youth athletics. Twenty-seven states have enacted such laws in recent years, and Justice Kavanaugh held for the Court that the Constitution leaves them free to do so. Kavanaugh’s primary authority was the Court’s decision last year in United States v. Skrmetti (2025), which rejected an equal protection challenge to Tennessee’s ban on gender-affirming healthcare for minors. But the arguments have a longer history.

Here’s Justice Scalia in Romer v. Evans (1996), voting to uphold a state constitutional amendment in Colorado that explicitly sought to authorize anti-gay discrimination:

Here’s Scalia again in Lawrence v. Texas (2003), voting to uphold Texas’s criminal ban on consensual sodomy:

In short, conservative justices have been arguing for a long time that the Fourteenth Amendment extends virtually no constitutional protections to LGBT persons. The recent decisions are worrisome not just because the conservative justices now have a reliable majority for these arguments, but because they have indicated no awareness that crackdowns on LGBT rights are a recurring feature of twenty-first century autocratization worldwide. (See, for example, this piece on Brazil.)

Justices Jackson, Kagan, and Sotomayor have objected to all of these lines of conservative doctrine, but they have by and large done so within the confines of ordinary doctrinal disputes, rather than articulating any special warning about the consequences of the Court’s decisions for Trump-era autocratization.

For example, dissenting alone in Chiles, Jackson warned that the decision would undermine state medical licensing regimes, which routinely feature standards of care that incidentally burden providers’ speech. But she didn’t say anything about the Court’s decision undermining antidiscrimination protections for a vulnerable group that has been a recurring target of 21st century autocratizing governments.

Dissenting in Callais, Kagan called attention to the decision’s likely impact on minority representation, but didn’t say much about the added urgency of resisting voter suppression under conditions of democratic backsliding:

Dissenting in NRSC (the campaign finance case), Kagan came a bit closer. She opened her dissent by objecting that her colleagues have “jettison[ed] a rule needed to protect our democracy's integrity,” and closed with this warning:

in between, Kagan illustrated the underlying purpose of our campaign finance laws by recounting a Watergate-era example in which President Nixon shook down the dairy industry for a $2 million contribution in exchange for increased government milk subsidies. Given President Trump’s well-documented record of corruption, this passage reads as at least a veiled reference to contemporary challenges.

Dissenting in Slaughter, Sotomayor offered perhaps the term’s clearest alarm bell about Trump-era autocratization:

I close with one final case from the recent term. In Mullin v. Doe (2026), the Court allowed President Trump to unilaterally revoke Temporary Protected Status for Haitians and Syrians living in the United States, even though Trump had done so without following the requisite procedures in federal immigration law. One issue in the case was the Haitians’ assertion that Trump’s decision had been racially motivated and hence violated the Fourteenth Amendment. The immigrant rights advocates cited repeated racist comments about Haitian Americans by Trump and former Homeland Security Secretary Kristi Noem, but Justice Alito, writing for the conservative majority, dismissed this evidence:

Justice Kagan responded like this:

A healthy democracy cannot long feature a national executive who regularly makes overtly racist comments about vulnerable segments of the nation’s population, nor a national apex court that willfully ignores those comments. Yet here we are.