Abortion rights activists rejoice in front of the Supreme Court in Washington as the justices struck down a strict Texas anti-abortion restriction law, June 27, 2016. Credit: J. Scott Applewhite | AP

Abortion isn’t mentioned in the Constitution, but it is the main topic in the debate over replacing retiring Justice Anthony Kennedy on the Supreme Court. We are certainly going to hear a lot of polarized rhetoric about abortion over the next few months. But we should keep in mind three middle-of-the-road truths.

First: The public is ambivalent about abortion. People who favor legal abortion can point to a lot of polls that appear to show them in the majority. Because it’s a Supreme Court nomination we are going to be debating, right now they are especially eager to note that polls have pretty consistently shown strong support for Roe v. Wade. In the most recent Kaiser Family Foundation poll, two-thirds of Americans supported it.

But polls that ask other questions about abortion undercut that finding. Gallup has consistently found that a small majority of the public, currently 53 percent, believes that abortion should either be “illegal in all circumstances” or “legal only in a few circumstances.” Only 45 percent of Americans believe that abortion should be legal in the first trimester “when the woman does not want the child for any reason,” and only 28 percent think it should be generally legal in the second trimester.

Roe and its successor cases require abortion to be legal for any reason in the first and second trimesters. (They allow bans after fetal viability, so long as they include an exception when a doctor believes abortion is necessary for the physical or emotional health of the woman.)

In short: Most Americans do not want a ban on abortion, and favor its availability especially early in pregnancy and in cases of rape and incest, threats to the mother’s life, and severe fetal abnormality. But they also favor restrictions that are incompatible with current Supreme Court jurisprudence.

Second: Roe has force as a precedent but can still be overruled. Almost all legal thinkers accept that some past holdings have to be left in place even if they were mistaken, in order for the law to have stability. On the current Supreme Court, only Justice Clarence Thomas denies this concern should apply, and he does so only in constitutional cases. But nearly everyone also agrees that precedents must sometimes be abandoned.

One factor that is generally conceded to affect the force of a precedent is how wrong it was in the first place. Roe had very little constitutional basis, and has been assailed on that basis even by many legal scholars who approve the result that abortion is legal.

Another factor is the age of a precedent. Roe came down in 1973. But the court has abandoned old precedents before. Last week, in a case involving the rights of public-sector workers to refrain from paying dues to unions, it overruled a decision nearly as old as Roe. In another case last week, it condemned the even older decision that allowed Franklin Roosevelt to intern Japanese-American citizens during World War II.

The stability of the precedent — and the related question of how much reliance can be placed on it — should also be part of the analysis. Roe was reaffirmed, but also substantially modified, in the Casey decision of 1992, which upheld some restrictions. During the next major abortion case, in 2000, the authors of the Casey decision split about how to apply it to partial-birth abortion. In 2007, the Supreme Court essentially reversed its 2000 decision.

In both 1992 and 2007, the court had suggested that protecting human life is a legitimate governmental interest even before fetal viability. In 2007, it said that the government has an interest in preventing the coarsening of society with respect to human life. In 2016, another decision cast doubt on both points.

A state legislator trying to decide how to vote on an abortion bill cannot confidently tell what provisions will be held constitutional. The court could decide in some future case that it has been unable to use the Constitution to come up with a workable legal code on abortion — and get out of the business altogether.

Third: The confirmation of a new conservative justice would raise the probability that Roe v. Wade will be overturned but not guarantee it. You may have noticed that politically savvy opponents of abortion, who dearly want Roe overruled, are nonetheless emphasizing the no-guarantee point. That’s because they know that there are a lot of ambivalent voters who fear sweeping and drastic change in abortion policy.

But what they’re saying is also correct. While Chief Justice John Roberts and Justice Samuel Alito are conservatives who have voted to uphold abortion restrictions, we do not know that they are willing to overrule Roe. Roberts, in particular, has tended to prefer making narrow rulings — which in this case could mean continuing to uphold discrete restrictions on abortion while leaving Roe on the books. Of the sitting justices, only Clarence Thomas has said Roe should go.

The coming confirmation hearings are unlikely to give us a definitive answer about the nominee’s proclivities. President Donald Trump has said that he wants a justice who will overrule Roe. But the president says a lot of things. While his legal advisers will be looking for someone who has demonstrated evidence of being a strong legal conservative — something that correlates with opposition to Roe — there is no reason to think that they, or Trump himself, would ask the nominee a direct question about it. They will avoid it in part because it would be hard for the nominee to dodge senators’ questions about any such exchange.

The senators aren’t going to get a definitive answer either. The nominee will say, appropriately, that it would be improper to commit to ruling one way or another on a case that might come before the court. The nominee will probably also say, less defensibly but in keeping with a common convention of nominees from both parties, that it would be wrong to comment at all on the merits of disputed constitutional questions.

The nominee might allow that the Constitution protects a right to privacy, as when it restricts the government from quartering troops in your house, but refuse to say whether that right extends to abortion. The nominee will almost certainly take a moderate line on the power of precedent: It exists, but it’s not absolute, and it would be wrong to make pre-commitments about how it would apply in specific cases.

We don’t know for sure how many justices now on the court would vote on Roe, and we won’t know for sure how the new justice will vote either. But we can expect that Roe will be on shakier ground if a new justice is confirmed.

People for whom that’s an unacceptable risk will, as they should, oppose confirmation of the president’s nominee. The problem for them is that all of that public ambivalence on abortion policy leaves them in the minority. The upshot is that legislatures are likely to get more power to restrict abortion — and maybe a lot more.

Ramesh Ponnuru is a Bloomberg Opinion columnist. He is a senior editor of National Review and the author of “The Party of Death: The Democrats, the Media, the Courts, and the Disregard for Human Life.”

Follow BDN Editorial & Opinion on Facebook for the latest opinions on the issues of the day in Maine.