Pro-choice advocates rally outside of the Supreme Court on March...

Pro-choice advocates rally outside of the Supreme Court on March 2, 2016 in Washington, DC. Credit: Getty Images / Drew Angerer

What’s an undue burden?

That question was at the heart of Wednesday’s oral argument at the U.S. Supreme Court in the Texas abortion case of Whole Woman’s Health v. Hellerstedt.

In particular, the conversation focused on whether the court needs to do a cost-benefit comparison to determine an undue burden — and if it does, what statistical evidence is needed to do it properly.

As expected, the oral argument reinforced the sense that the outcome of the case depends on Justice Anthony Kennedy. The four liberal justices made it pretty clear that the Texas law, which requires abortion clinics to operate more like hospitals, should be struck down. The three conservatives, sorely missing the support of Justice Antonin Scalia, will surely vote to uphold it, although Justice Clarence Thomas kept silent on Wednesday. What matters, therefore, is how Kennedy is thinking about the undue-burden problem.

Justice Sonia Sotomayor wanted to help. She pressed counsel for both sides on the question of what benefits, if any, the law might have. Under her careful guidance, Stephanie Toti, the lawyer for the reproductive-rights group, said there was no medical benefit to be had from taking abortion-inducing drugs in a clinic rather than at home, as the law requires.

And Toti said there was no medical advantage to performing the procedure known as D&C (dilation and curettage) during an abortion in an expensively equipped ambulatory surgical center given that an identical procedure is often performed in other cases in a doctor’s office.

Then Sotomayor asked Scott Keller, the Texas solicitor general, “Are you taking into account in the undue-burden analysis the value of the need being imposed?”

Sotomayor’s point was that determining whether the law created an undue burden would require comparing the burden imposed with the benefits created by the law.

Keller tried to deflect, by saying that the court should assume that a law has a rational basis unless proved otherwise. The reason for Texas to avoid a weighing of costs and benefits is pretty clear: The real purpose of the law is to reduce the total number of abortions.

The state legislature almost certainly considered that to be the law’s benefit, but it couldn’t say so when writing and passing the law, because that would seem to violate the very idea of a constitutional right to abortion. Instead, the law was presented as protecting women’s health. On that dimension, the law does little or nothing.

Kennedy was listening. After Sotomayor’s foray he asked the Texas lawyer whether the premise or effect of the law was “to increase surgical abortions as distinct from medical abortions.” He added that “my reading indicated that medical abortions are up nationwide but down significantly in Texas.” Furthermore, Kennedy said, “this may not be medically wise.”

Abortion-rights activists will be heartened to hear Kennedy’s final suggestion. But it would be a mistake to assume that he fully bought the argument that the law has no medical benefits.

If the court’s method of analysis were to balance cost against benefits, then it would be relevant that more women seem to be having surgical abortions in Texas, thus incurring greater medical risk than they would if they were having the drug-induced procedure. But if undue burden is just a question of individual women being able to get an abortion of some kind, then a different sort of evidence would be needed.

On this point, Kennedy was also animated. He suggested that it might be necessary to send the case back to the lower courts to ascertain whether, as a matter of practice, women in Texas have sufficient access to abortions going forward.

His thought seemed to be that new ambulatory surgical centers might come into existence to satisfy the need created by the closure of existing abortion clinics. He even asked, somewhat fancifully, whether the federal district court would have the authority to freeze the operation of the law for several years to see if such new centers came into existence. After all, he said, “district judges often think they can do anything.”

Kennedy got a laugh for this quip; district judges, who, statistically speaking, are rarely reversed on appeal, may be among the most powerful government officials for that reason and may seem omnipotent. Yet the serious implication of Kennedy’s question was that if the Texas law is to be analyzed based on abortion accessibility, rather than costs and benefits, the court might need more empirical facts.

The upshot is that it seems relatively unlikely that Kennedy would vote with the conservatives, creating a 4-4 split that would uphold the law without creating a national precedent. But it remains uncertain — possibly even to Kennedy himself — whether he would vote to strike down the law now by a vote of 5-3 or to send the case back to the district court for more fact-finding.

The court’s outcome depends on whether Kennedy chooses the cost-benefit route or chooses to focus on access. Whichever he picks, he’ll be pointing the way for future jurisprudence in the area of abortion rights, even after he’s gone from the court.

Bloomberg View columnist Noah Feldman is a professor of constitutional and international law at Harvard University.