Font Size: A A A


The Supreme Court Picks Its Battles

David Cole
June ended with far-reaching upsets for civil rights, but a closer look at the Supreme Court’s decisions across the 2022–2023 term suggests the justices are not immune to public scrutiny.
A woman holds a cloth banner in front of the Supreme Court that says SCOTUS IS ILLEGITIMATE in paint

Celal Gunes/Anadolu Agency/Getty Images

A demonstrator holding a sign in protest of the US Supreme Court's decisions to strike down President Biden's student loan forgiveness plan and affirmative action, Washington, D.C., June 30, 2023

For those who care about civil rights and civil liberties, the Supreme Court’s 2022–2023 term, which concluded on June 30, was a case of waiting for the other shoe to drop. In the previous term the Court’s conservatives, exercising their newfound majority power, had overturned the right to abortion, struck down a century-old law limiting the carrying of guns in New York, rewritten religious freedom jurisprudence, and announced a newly crafted and ill-defined “major questions doctrine” to bar the Environmental Protection Agency from requiring electrical power producers to shift to cleaner sources of energy. Virtually every one of those decisions was 6–3, with the Republican- and Democrat-appointed justices dividing along party lines.

If the Court could do all of this in a single term, the conventional wisdom went, there was no limit to what it could do. Precedent apparently meant very little; in the abortion decision, Dobbs v. Jackson Women’s Health Organization, five of the Court’s conservatives, including all three of President Trump’s appointees, were willing to reverse a precedent that had stood for half a century and that protected a right enjoyed by half the nation. With so much apparently up for reexamination, people began to ask how law professors could continue to teach constitutional law.

As this term began, it looked to be another disaster waiting to happen. The Court agreed to hear multiple cases urging it to overturn even more precedents, this time especially focused on issues of equality. The Court agreed to consider the constitutionality of affirmative action, the Indian Child Welfare Act, the Voting Rights Act, state limits on partisan gerrymandering, and the application of a “public accommodation” law to a website designer who opposed same-sex marriage and asserted a right to turn away gay couples.

The other shoe didn’t really drop until the term’s final two days. But then it dropped like a heavy boot. In a pair of 6–3 decisions along partisan lines, the Court ended affirmative action and granted the web designer a First Amendment right to discriminate. Both were major blows to equality—and precedent. And invoking its newly minted “major questions doctrine,” the Court invalidated President Biden’s student loan forgiveness program. Evidently the Court remains willing, at least in some cases, to impose its worldview on the nation, regardless of precedent and heedless of the consequences.

Before those final two days, there had been a few other disappointing decisions. These included one drastically narrowing the definition of “wetlands” protected by the Clean Water Act and another denying a man imprisoned for conduct that the Court itself had ruled is not a crime any opportunity to raise that claim in the courts.

But far more often, the Court surprised observers by reaching liberal results, as two or more conservative justices voted with their three liberal counterparts. By a vote of 7–2, with Justice Amy Coney Barrett writing the majority decision, the Court rejected a multipronged constitutional challenge to the Indian Child Welfare Act, a law designed to preserve the integrity of Native American families and tribes by preferentially placing children for adoption within their tribe. Only Justices Samuel Alito and Clarence Thomas dissented. By the same 7–2 vote, the Court upheld the right of private victims of discrimination to sue for damages under statutes that impose civil rights obligations as conditions on government contracts. The challengers had argued that because the victims were not signatories to the contracts, they had no right to sue. This case was considered such a loser that many in the civil rights community, myself included, urged the parties to settle to avoid Supreme Court review.

The Court rejected a challenge by Texas and Louisiana to a Biden administration guidance document that prioritized the deportation of certain noncitizens, such as suspected terrorists and dangerous criminals, and deprioritized enforcement against others. This time only Justice Alito dissented. In a pair of major cases involving free speech online, the Court ruled that victims of terrorism could not sue Twitter, Google, and other social media platforms for merely failing to take down posts that were said to radicalize followers of ISIS. The Court ruled that the government can prosecute people for threatening others only when the speaker intended the threat or recklessly disregarded the risk that his speech would be perceived as a threat. It unanimously ruled that states cannot seize a person’s house for back taxes, sell the house, and then keep the entirety of the proceeds, even if they exceed the amount owed. And in a case that many in the labor movement feared would cripple the right to strike by authorizing lawsuits for economic injuries caused by the strike, the Court resolved the dispute on such narrow, fact-specific grounds relating to this particular strike that the decision will have little impact on most strikes; only Justice Ketanji Brown Jackson dissented. In all, the Court’s three liberal justices were in the majority in non-unanimous cases more often than its most conservative justices, Alito and Thomas. 


One measure of how much changed between the 2021–2022 term and the 2022–2023 term is how the Court ruled in cases in which the ACLU filed a brief. In the 2021–2022 term, the Court ruled for the side we favored only five times, all in fairly obscure cases, while ruling against our side thirteen times. In the 2022–2023 term, by contrast, the Court sided with the ACLU eleven times and against us in seven cases, and many of the victories were in the Court’s most consequential cases.


Two of this term’s most important wins for civil rights involved legal challenges to congressional maps enacted by Republican state legislatures after the 2020 Census. In Allen v. Milligan, Alabama urged the Court to declare unconstitutional, or radically reinterpret, a central provision of the Voting Rights Act. In Moore v. Harper, Republican legislators in North Carolina invoked the so-called independent state legislature theory in asking the Court to override a state constitutional limit on partisan gerrymandering. In both cases, the challengers appealed to views that several conservative justices had previously expressed in dissent, and that many thought would now garner a majority on the new Trump-inflected Court. But in both cases, Chief Justice John Roberts wrote decisions reaffirming precedent.

Allen v. Milligan arose under Section 2 of the Voting Rights Act, which requires states to draw districts that afford minority voters a fair opportunity to elect candidates of their choice where there is starkly racially polarized voting and other evidence of discrimination in the political system. To prevail, plaintiffs have to show, among other things, that the state could have drawn a map with additional “majority-minority” districts, that is, districts in which a minority group is sufficiently numerous to forge a majority. Black Alabamians are 27 percent of the state’s voting-age population, but the Alabama legislature drew a map in which only one of seven (or 14 percent) of the congressional districts were majority Black. The plaintiffs showed it was possible to draw a map that contains two majority-Black districts, and a three-judge court unanimously concluded that in light of the evidence of racial polarization and discrimination, the VRA required Alabama to create a second such district.

Alabama appealed, arguing that this approach to the VRA was unconstitutional because it permitted a map to be drawn according to racial demographics, and that in any event the VRA was not violated unless the state’s map was intentionally designed to disadvantage minority voters. The argument appealed to the same “colorblind” mindset that prevailed in the affirmative action cases. But in Allen, the Court rejected that contention. It explained that race consciousness is inevitable in districting, and that Alabama’s approach would have essentially nullified VRA reforms from 1982 designed to require states to ensure that minority voters have an equal opportunity to elect voters of their choice. The result means that Alabama’s Black voters will now have a second district in which they can elect representatives of their choice. And several other challenges to other states’ redistricting following the 2020 Census are likely to bring similar results.

Moore v. Harper involved partisan gerrymandering. After the North Carolina state legislature enacted on party lines a map that favored congressional Republicans far beyond their statewide share of the vote, the North Carolina Supreme Court rejected it as an “extreme partisan gerrymander” in violation of the state constitution. Republican legislators appealed to the US Supreme Court, maintaining that because the US Constitution tasks state legislatures with devising rules for congressional elections, the state’s highest court could not hold the legislature accountable to its own constitution. The Supreme Court rejected that argument, 6–3, with Chief Justice Roberts and Justices Barrett and Brett Kavanaugh joining the liberal justices. As Roberts explained, “Legislatures, the Framers recognized, ‘are the mere creatures of their State Constitutions and cannot be greater than their creators.’”

The decision is critically important, in part because, after the Supreme Court’s 2019 decision in Rucho v. Common Cause that federal courts cannot hear partisan gerrymandering challenges, state courts are one of the only remaining checks on this pernicious practice. More broadly, in this era of politically motivated challenges to election results, it means that partisan legislatures are not free to run roughshod over the will of the voters, because they will have to answer to their state’s highest court if they do.


To be sure, the civil rights losses this term were also substantial. In Students for Fair Admissions v. President and Fellows of Harvard College, the Court overturned forty-five years of precedent upholding the limited use of race in college admissions. Beginning with its 1978 decision in Regents of Univ. of Cal. v. Bakke, the Court had established that race could be considered as one factor in a broad assessment of applicants, to serve the states’ compelling interest in assembling a diverse student body. That view had always drawn detractors, including Justices Alito, Thomas, and Roberts. But they had always been in dissent. With the addition of the Trump-appointed justices, that changed, and in a pair of cases involving Harvard and the University of North Carolina the Court declared affirmative action unconstitutional. 


Chief Justice Roberts maintained that “diversity” is too imprecise a standard to justify the use of race, because courts are not able to assess when a student class is sufficiently diverse. But many interests the Court has deemed compelling in other settings are at least as difficult to quantify, among them the interest in national security or the integrity of the courts. Roberts also reasoned that the use of race as a positive factor for some applicants means that race will be a negative factor for others, because college admissions is a “zero-sum” game. True enough, but minority applicants were on the losing side of that game for more than a century, and using race to include underrepresented groups is surely different than using it to exclude them. And he noted that the Court had always insisted that affirmative action should be only a temporary measure and had specifically stated in a 2003 case that it expected the practice to end in twenty-five years, by 2028. (In a remarkable example of “math for lawyers,” the Court sought to elide the fact that it is only 2023 by noting that this year’s applicants will be accepted into the “class of 2028,” as if that somehow turns twenty years into twenty-five.)

The result will likely be substantially less opportunity for underrepresented students to attend the nation’s best colleges. But in closing, Roberts mitigated the damage by cautioning that “nothing in this opinion should be construed as prohibiting universities from considering an applicant’s discussion of how race affected his or her life, be it through discrimination, inspiration, or otherwise.” As long as the applicant links his race to a characteristic or quality that makes him a good candidate for admission, he explained, the college can consider race in this way. While Justice Sonia Sotomayor in her dissent dismissed this as no more than “an attempt to put lipstick on a pig,” it’s an important caveat that race can still be considered. Other justices noted the continuing propriety of race-neutral alternatives, such as programs giving preferences to low-income and first-generation applicants, and those who speak more than one language. Much will depend on the continuing commitment of colleges and universities to ensuring opportunity for marginalized groups.

The real division between the majority and the dissenters in the affirmative action cases lies in their approach to the Fourteenth Amendment’s Equal Protection Clause. The majority reads it to require colorblindness and therefore to call into question any consideration of race, even those designed to ameliorate inequality. The dissent reads it as directed toward achieving justice and equality for newly freed slaves, and not as insisting on colorblindness, especially where formal equality only reinforces existing inequality. As Justice Jackson put it in her dissent, “Those who demand that no one think about race (a classic pink-elephant paradox) refuse to see, much less solve for, the elephant in the room—the race-linked disparities that continue to impede achievement of our great Nation’s full potential.”


In 303 Creative v. Elenis, the Court for the first time in its history upheld the right of a business to violate a public accommodations law by invoking the First Amendment. In a 6–3 decision written by Justice Neil Gorsuch, the Court remarkably likened a Colorado law that merely requires all businesses that choose to serve the public not to discriminate in doing so to the World War II–era requirement that students recite the pledge of allegiance even if it violated their religious scruples. The case involved a website designer who claimed she wanted to design wedding websites, but not for same-sex weddings. Colorado law said that if she offered website design services to different-sex couples, she couldn’t turn away same-sex couples seeking the same service.

The Court’s reasoning was straightforward: the website design business involves customized expression, and Colorado’s law, by requiring the business to treat same-sex and different-six couples the same, would compel the business owner to express a message—celebrating same-sex marriage—with which she disagreed. Justice Gorsuch wrote, “When a state public accommodation and the Constitution collide, there can be no question which must prevail.”

But it’s not so simple. The fact that a business provides a customized “expressive” service doesn’t mean states can’t require it to treat all customers equally. In fact, the Court had previously rejected arguments that the First Amendment permitted a restaurant to refuse to serve Black customers, a university to refuse to admit applicants engaged in interracial dating, and a law firm to deny promotion to a woman, even though each could be said to provide customized expressive services. Surely architecture firms, no less than hardware stores, can’t put up signs saying, “No Blacks served here” even if their owners object to the message serving Black customers sends.

When the state regulates conduct that is expressive, the First Amendment does not provide an automatic license to ignore the law. Instead, the Court asks whether the state’s interest in enacting the regulation is related to the suppression of ideas or is content-neutral. If the law regulates conduct regardless of its content, the Court applies relaxed scrutiny. If, by contrast, the law regulates conduct because of what it communicates, the law is treated as a content-based regulation, subjected to “strict scrutiny,” and virtually always declared unconstitutional. Because public accommodations laws prohibit discrimination in sales by all businesses open to the public, whether hardware stores or architecture firms, and the state’s interest is in ensuring equal access to the economic marketplace, such laws should properly be subjected only to intermediate scrutiny and should survive.

Justice Gorsuch failed even to engage with this argument. Instead he ignored precedent and erected straw men, claiming that if the website designer could be made to serve a gay couple, then nothing would stop states from requiring a Muslim film director to direct a film with a Zionist message or a gay website designer to create websites for a group advocating against same-sex marriage, as long as these speakers would accept commissions from the public with different messages. But that doesn’t follow. Public accommodations laws apply only to businesses open to the public, as Colorado explained, and most artists and writers are freelancers not serving the public generally, so the Muslim film director would not be bound by the law. By contrast, a director who opened a school to provide filmmaking lessons to the public at large could not put up a sign saying “No Zionists need apply.”  The law governs only those who affirmatively choose to serve the public, as 303 Creative did here.

Even when a business chooses to serve the public, moreover, Colorado’s law does not dictate what it may sell. It leaves that decision entirely to the business, and merely requires that whatever it sells, it does so on equal terms to all. A T-shirt designer can choose to design only white supremacist T-shirts, and as long as he sells such shirts to all, he complies with public accommodations law. He need not make “Black Lives Matter” or “We Shall Overcome” T-shirts if he would not make them for anyone.

In the majority opinion, Gorsuch at times seems to treat this case as involving only the refusal to provide a website design that 303 Creative would not sell to anyone. He cites a stipulation that the business owner “will not produce content that ‘contradicts biblical truth’ regardless of who orders it.” But as Colorado made clear, its law does not require her to sell any particular service, so if she simply did not want to make websites celebrating same-sex marriage for anyone, she was free to do so.  She sought a broader ruling, however, allowing her to refuse service to same-sex couples regardless of the content of the wedding site they sought. Even if a gay couple’s site was indistinguishable from a site announcing a different-sex marriage, and merely identified the time and place of the wedding, she refused to provide the service. Indeed, as Justice Sotomayor pointed out, the website design contained in the record, the only evidence of what 303 Creative sought to provide, was just such an anodyne site, communicating little more than wedding logistics. Because the case was filed before the business had provided a website design to anyone, the exact scope of the claim was unclear. Gorsuch’s opinion does not clarify matters.

The implications of the Court’s ruling will consequently be litigated for years to come. Does it mean that a business owner who provides application essay assistance to the public at large, a customized expressive service, could refuse to serve minority students because he thinks the nation’s elite colleges should be reserved for white students? Does it mean a corporate photography studio could refuse to take portraits of women if it objects to women working outside the home? Could a wedding website designer put up a sign announcing “No interracial couples please” if she objects to such marriages? Justice Gorsuch did not say.


So the Court’s record this term was definitely mixed. Some will see the glass half full, others half (or more) empty. But what is indisputable is that the Court delivered far more surprises, far more bipartisan opinions, and far more protection for civil rights and civil liberties than anyone expected. And while last year seemed to herald a new “originalist” method for interpreting the Constitution, almost none of the Court’s decisions this term were decided on originalist grounds. How are we to account for this shift?

One possibility is that protest matters. In response to last year’s term, especially to the decision to strip women of the right to choose whether to have a child, many warned that the Court risks losing its legitimacy when it falls out of step with where the American people stand on fundamental issues, divides along partisan lines, and appears to place justices’ personal policy preferences over precedent. The Court’s disapproval ratings rose to 58 percent, the worst they have been in the twenty-first century. The recent attention to the Court’s ethics breaches only underscores the point; Justice Thomas has failed to report gifts for years, but only now, as people worry about the Court’s role in our democracy, have those failures become a public controversy.

Meanwhile, every time the issue of abortion has appeared directly or indirectly on the ballot since the Dobbs decision, the right to abortion has prevailed. Sixty-nine percent of Americans think that abortion should generally be legal in the first three months of pregnancy. Political scientists studying the Court find that its decisions rarely part company with popular opinion, and when they do, a correction generally follows.

The Court is of course not directly responsive to popular will, nor should it be. The justices do not consult polls and act in accordance with public opinion, as their counterparts in Congress often do. We insulate them from the political pressure of having to run for reelection because we ask the Court to protect those who lack the political power to protect their own interests: the poor, the vulnerable, the marginalized, the unpopular.

But at the same time, the Court is a powerful institution in our democracy, and therefore must in some larger sense be answerable to the people. At the end of the day, its power resides in its legitimacy, and the justices are not immune from criticism. Had Americans accepted last term’s radical decisions without protest or criticism, it seems likely that this term’s cases would not have come out as favorably as they did. So for those of us who care deeply about our democracy and our rights, it is essential that we both praise the Court when it does the right thing and continue to raise our voices when it does not.

New York Review + Paris Review covers

Save $168 on an inspired pairing!

Get both The New York Review and The Paris Review at one low price.

Already a subscriber? Sign in

© 1963-2024 NYREV, Inc. All rights reserved.