As the Supreme Court term winds down, there is discussion whether the Court is in some way drifting to the left. That narrative may take on additional steam after Thursday’s decisions—King v. Burwell, negating a far-right challenge that might have destroyed the Affordable Care Act, and Texas Department of Housing and Community Affairs v. Inclusive Communities Project, an attempt by the state of Texas to radically scale back the scope of the federal Fair Housing Act.

Both challenges failed Thursday. But it would be a mistake to read Inclusive Communities as a “liberal” decision. Although four justices clearly wanted to radically cut back the Act, the majority—Justices Anthony Kennedy, Ruth Bader Ginsburg, Stephen Breyer, Sonia Sotomayor, and Elena Kagan—signed on to an opinion narrowing it in crucial ways.

This was no ringing victory for civil rights; it was a near-death experience that may produce health problems for the Act down the road.

The issue in Inclusive Communities was whether the Act allows plaintiffs only if they can show that a government agency or private actor in renting, selling, or otherwise making housing available on the basis of race, intended to discriminate on the basis of race, or whether the Act also forbids acts that have a “disparate impact” on housing opportunities. Disparate impact means that some policy, adopted for a non-racial reason, might end up burdening minority-housing opportunities more heavily than those provided to whites; if so, the policymaker would be required to justify the policy to a court on grounds other than race.

In this case, for example, the plaintiffs produced evidence that the Texas Housing Department awarded tax credits to more developments in poor, minority areas than in areas with majority-white population. The effect, they argued, was to lock minorities into certain parts of the Dallas metro area, perpetuating and extending the area’s segregated housing patterns. The department argued that its decisions followed a complicated set of criteria generated in conjunction with the U.S. Department of Housing and Urban Development and including questions like the cost of land and construction (lower in low-income areas) and “community revitalization,” meaning the economic boon that development can represent to poorer areas. The federal district court held that the agency’s non-racial justifications were not sufficient. The Court of Appeals reversed, arguing that disparate-impact claims are permitted but that these plaintiffs had not made their case.

Before that case could be reheard, however, the state of Texas asked the Supreme Court to take a radical step by reinterpreting the FHA to bar disparate-impact claims altogether. Every circuit court of appeals that had considered the issue had held that the Act permitted disparate-impact claims. But the high court granted review. The Court had granted review in two earlier cases, but both had settled at the last minute. The Texas case finally teed the issue up squarely.

From a distance, the result in Inclusive Communities looks like a win. Writing for himself and the four moderate-liberals, Justice Kennedy explained that the disparate-impact interpretation had a lot going for it: it tracks two other Court precedents concerning the employment-discrimination provisions of the Civil Rights Act of 1964 and the Age Discrimination in Employment Act; it has been upheld by every court of appeals to consider the issue; Congress readopted the Act in 1988 with language that seems to recognize disparate-impact liability in all but a few categories of cases; and it has become a part of the landscape of urban planning, such that many large cities—including San Francisco, New York, Boston, and Baltimore--submitted a brief asking the Court to leave the Act alone. Eliminating disparate-impact claims would thus destabilize not only other areas of civil-rights law, but also a great deal of city planning. “The Court acknowledges the Fair Housing Act’s continuing role in moving the Nation toward a more integrated society,” Kennedy concluded.

But the majority opinion is less a ringing reaffirmation than a stern warning—claims like those brought by the plaintiffs in this case, Kennedy wrote, actually might raise “serious constitutional questions.” That is, the use of statistics, no matter how persuasive, to show disparate impact without additional evidence creates a danger of “abusive disparate impact claims” that may hobble local governments and developers. Without strict safeguards, the opinion said, “disparate-impact liability might cause race to be used and considered in a pervasive way and ‘would almost inexorably lead’ government or private entities to use ‘numerical quotas.’”

Kennedy concluded that “we must remain wary of policies that reduce homeowners to nothing more than their race.” And the implication is that anything outside the “heartland” of disparate-impact liability—that is, “zoning laws and other housing restrictions that function unfairly to exclude minorities from certain neighborhoods without any suffi­cient justification”—would be dangerous territory.

These particular plaintiffs, the opinion made clear, almost certainly must lose on remand. Disparate impact lives on. But the lower courts have plenty of ammunition in this opinion to use against any novel use of the FHA.

If the result seems “liberal,” it is only in contrast to the dissenting opinions, which are truly radical—contemptuous both of judicial precedent and the history of executive enforcement of civil-rights laws over the past half-century. Justice Clarence Thomas, writing for himself, urged the court to begin overturning all those precedents—especially the case that originated the theory of disparate-impact liability, Griggs v. Duke Power, which used it to invalidate a neutral-seeming employment policy that had the effect of trapping black workers in laborers’ positions. The Civil Rights Act, Thomas said, didn’t justify the decision; the Court had relied on a deceitful bunch of bureaucrats at the Equal Employment Opportunity Commission, who schemed to enlarge the scope of the law—and their own power—by hoodwinking the Court. It is a striking argument, especially when made by the former head of the EEOC—one who, during his tenure, had tried to reorient the Commission away from the role it had played in the ‘60s and ‘70s.

Thomas concluded with a strange set of musings about racial disparities in general. Disparity is the way of the world. He quoted conservative economist Thomas Sowell to the effect that some minority groups end up running the economies of entire nations: “the Chinese in Malaysia, the Lebanese in West Africa, Greeks in the Ottoman Empire, Britons in Argentina, Belgians in Russia, Jews in Poland, and Spaniards in Chile—among many others.” Besides, he said, “over 70 percent” of players in the NBA are black.

The principal dissent, by Justice Samuel Alito writing for Chief Justice John Roberts and Justice Antonin Scalia, is, a bit less eccentric but equally radical. The precedents are not worthy of respect, Alito argued. Congress may have re-enacted the Act in 1988, but its changes didn’t mean everyone supported DI; the Reagan administration had said the Act didn’t allow DI claims. HUD, which issued regulations supporting DI, was actually trying to manipulate the Court rather than expressing its “’fair and considered judgment.’”

Like Thomas, Alito pointed out that disparities are everywhere—not only in the National Football League but in the Office of the Solicitor General, which mostly sends young lawyers to argue in front of the Court.

With one more vote, these two screeds would have created a gaping hole in the fabric of civil-rights law, with malign effect far beyond housing. That Kennedy rejected that course is an occasion for relief. But if you think that the great danger facing the United States is too much separation rather than too much equality, Kennedy’s opinion is no cause for celebration. It is a slow and measured step to the right, rather than a radical one. But its direction is clear.