Sherif Girgis on the Supreme Court's Options in the Next Abortion Case

There are "Two Obstacles to (Merely) Chipping Away at Roe in Dobbs," he writes

On a few recent episodes of Divided Argument, co-host Dan Epps and I talked about the Supreme Court's recent grant of certiorari in Dobbs v. Jackson Women's Health Organization, and what the Court might do when it decides the merits. One of our listeners, Sherif Girgis (who will be a starting law professor at Notre Dame Law School this fall) wrote to me with some extended thoughts disagreeing with some of what I said. I'm not entirely persuaded by them, but I thought they were interesting enough to be worth sharing with readers:

Two Obstacles to (Merely) Chipping Away at Roe in Dobbs

I'm surprised by commentary on the Supreme Court's decision to take Dobbs v. Jackson Women's Health Organization.  At issue is Mississippi's prohibition of abortions after 15 weeks (with exceptions for maternal health emergencies or severe fetal abnormality).  Given the law's reach, the Court's makeup, and the timing, some observers say it is "unlikely that there will ever be a more opportune vehicle" to revisit Roe v. Wade.  That much seems plausible.  But many on both sides also assume first, that even if Dobbs pared back Roe, the Justices would stop short of eliminating Roe's constitutional right to elective abortions; and second, that this would be part of a "long game" of chipping away at that right until the Court could deal the now-crumbling right a final blow.

I make no predictions.  But reflecting on the Roberts Court, the history of abortion cases, and the details of this case, I have doubts about both assumptions.  I think the conventional wisdom vastly underestimates how hard it would be for the Dobbs Court to (1) stake out a middle ground, or (2) position itself to fully reverse Roe down the line.  (Here "fully reverse Roe" means: eliminate any constitutional entitlement to elective abortions—those not needed for maternal health—such that states could prohibit or allow those, at any stage, subject to rational-basis review.)

On (1):  I see how the Court could use standard judicial reasoning (sound or not) to justify an all-or-nothing outcome.  At one end, it could invalidate the 15-week law based on Roe and its progeny.  At the other end, the Court could scrap those precedents and uphold the law based on textual or historical arguments.  (The objections to each path are familiar.)  It is much harder to imagine the opinion from this Court that upholds the Mississippi law (or remands) without fully reversing Roe.

On (2):  For reasons peculiar to Dobbs and given below, a "moderate" ruling couldn't simply narrow the existing abortion right.  It would have to double down on any new version of the right it created, making it hard for the same Justices to then abolish that residual right later on.  If Dobbs was meant to advance a long game against Roe, it was an odd choice.  For better or worse, a moderate Dobbs might cut the game short.  The reasoning needed to go moderate here differs in kind from that used by the Roberts Court in cases that became stepping-stones to eliminating a disfavored doctrine.

To begin to see why on both (1) and (2), step back for a second to consider the arc of our abortion jurisprudence.  The right declared in Roe is not expressly set out in the Constitution's text, and it lacks a history of well-defined sub-constitutional protection.  So there are few legal footholds on which a court can rest any specific doctrines meant to implement this right.  There's nothing specific to cite in text or history, but only in precedents that arose since the right was announced, plus any bright-line ethical or policy lines a court can find.  (Bright-line, so that the reasoning feels judicial, not legislative.)  But in some cases, there may be no bright lines.  Or there may be too many bright lines to choose from, which again makes any choice among them look legislative, not judicial.  In those cases, only the vaguest implementing doctrine won't seem plucked out of thin air.  (That's why it sounds more plausible to say that the Constitution contains an unwritten "no undue burdens on abortion" rule, than to say that it contains an unwritten "no bans after 13.5 weeks" rule.)  So there will be a tradeoff between a doctrine's determinacy and its apparent legal authority.  More determinate doctrines will seem made up, and more authoritative-sounding ones will be vague and manipulable.

When the Court declared a right to elective abortions in Roe, it fashioned quite specific legal standards to implement the right.  Roe made regulation of abortion hardest in the first trimester, easier in the second, and easiest in the third.  And sure enough, it lacked the feel of judicial reasoning.  It was widely panned for that.  John Hart Ely, who favored legal abortion, said the decision was "not constitutional law and [gave] almost no sense of an obligation to try to be."  Still, once Roe was on the books, the Court could cite it as precedent the next time it conducted a wholesale reexamination of the legal test for abortion laws.  That happened in Planned Parenthood v. Casey (1992).  But even in Casey, the controlling plurality—Justices O'Connor, Kennedy, and Souter—was embarrassed by the specificity of Roe's elaborate rules.  (While "legislatures may draw lines which appear arbitrary without … offering a justification," it said, "courts may not.")  So Casey whittled down Roe's complex rules to a simpler-looking test meant to capture Roe's core.  Since then, the Court has tried only to apply or tweak (not replace) Casey's test, always to laws more modest than Mississippi's 15-week ban.

It can't do that in Dobbs.  To rule for Mississippi (or remand), without abolishing Roe, the Court would have to devise a test that allows prohibitions starting sometime after conception but before 15 weeks (or tell the lower courts to do so, based on a new doctrine that Dobbs would provide).  But not only will Dobbs be unable to cite text or history to support any test.  Unlike Casey, it can't cite precedent (or even repurpose precedent, as shown below).  That's because the Court's precedents uniformly reject laws as sweeping as Mississippi's.  The latter bans abortions well before viability, which Casey forbids.  Indeed, the question presented in Dobbs is "whether all pre-viability prohibitions on elective abortions are unconstitutional," and Casey answers that yes-or-no question with "yes":  "[A] State may not prohibit any woman from making the ultimate decision to terminate her pregnancy before viability."  Thus, to do anything but block Mississippi's law, Dobbs must reverse Casey (as the cert petition asks it to do if needed).  And if the Court does that while still keeping some right to abort, it will have to create a test for abortion laws from scratch, for the first time since Roe.

But this isn't the freewheeling Burger Court of Roe, ready to constitutionalize "legislative code[s]," as then-Judge Ruth Bader Ginsburg lamented.  Nor is this the Casey Court, controlled by decided non-formalists O'Connor and Kennedy.  It's the Roberts Court.  Most of its members—and certainly the perceived swing Justices, Chief Justice Roberts and Justice Kavanaugh—are more lawyerly.  They will want at least a bright-line ethical principle (if not history or precedent) on which to rest their decision in Dobbs.  And as suggested below, there's no cogent candidate (for a moderate ruling).  While they could always fall back on creating a vague standard, any that saved Mississippi's law would have to be vaguer than Casey's.  And a ruling based on sharp but arbitrary lines would be too obviously ungrounded in legal sources.  In any scenario, an in-between approach to Mississippi's law would require Roberts or Kavanaugh to be less formalist than O'Connor, Kennedy, and Souter.

Similar dynamics make it hard to use Dobbs as a stepping-stone to overturning Roe.  Again, the law is so restrictive that the Court can't uphold it just by refusing to extend abortion precedents, or by reading them narrowly.  The Court would have to scuttle its abortion precedents.  So if it didn't eliminate with them any constitutional right to elective abortions, Dobbs would have to create a new and narrower right, based on a new and more permissive test.  This, as seen below, would require the Justices to justify the new right and test in their own voice (if only based on reliance interests).  And that would make it hard for them to say in a future case that there is, after all, no such right.  For then they would be rejecting not only Roe and Casey, but their own affirmative reasoning in Dobbs.

Some might respond that with abortion, the Court isn't bound by legal niceties—that it will use any means to advance its goals while limiting the blowback.  But in this case, the incentives to be lawyerly and to limit any headache for the Court might align.  Recall the tradeoff between determinacy and legal authority.  Again (and as explored below), here the in-between options would be un-lawyerly in the sense of (1) having an apparently less principled basis than any of the Court's abortion rulings since Roe, or (2) requiring a new constitutional test for abortion laws that is less determinate than any previous test.  But a less-principled-looking basis would harm legitimacy, by making it hard for the Court to say its hands were tied to reach its "conservative"-ish result.  And the second problem, a malleable new test, might further politicize the Court by ensnarling it in endless litigation of ever more stringent abortion laws upheld by emboldened lower courts.  That is something the Roberts Court may be as keen to avoid as decisions that seem unprincipled.

Likewise, if the goal is to chip away at Roe now, to soften the blowback from a future reversal, using Dobbs as the means might be self-defeating.  It might worsen the blowback, by forcing the Justices to flip-flop so visibly.

In Dobbs, then, legitimacy and de-politicization may point in the same direction as lawyerly self-respect.  They may push the Roberts Court to go all or nothing now.  To see why the alternatives may be incoherent, or too apparently lacking in authority, or so vague that they spawned more litigation than they resolved, or patently inconsistent with reversing Roe later, take them in turn.

0. Just say "no" and remand? Some wonder if a Roe-skeptical but gun-shy Court might simply answer the question in Dobbs with "no, not all pre-viability prohibitions are unconstitutional," and remand for the lower courts to figure out if this one is.  It can't.  The first part of that ruling—"no, not all"—would scrap Casey, the source of the current legal standard.  So what test would the lower courts apply to Mississippi's law on remand?  Dobbs would have to create one.  (And for a remand to make sense, the new test's implications for 15-week bans must be debatable.)

What new test might Dobbs create?  I'll quickly review the options, in three buckets.

1. Repurpose Casey's "undue burden" test, much as Casey revised Roe? Mississippi's attempt to square its law with Casey, though it fails, might point to one set of options.  Casey forbade laws imposing an "undue burden" on abortion before viability (~24 weeks).  Mississippi argued that its law complies with Casey by leaving pregnant women ample time—up to the 15th week—to decide to abort a pre-viable fetus.  But this is a flat misreading.  For one thing, Casey's undue burden test is relevant only to laws that make it harder to abort pre-viability.  It says they shouldn't make that too ("unduly") hard.  Since prohibitions like Mississippi's make such an abortion impossible, they're always "undue"; no test needed.  In fact, "undue burden" is just Casey's way of telling when a regulation is so severe that it effectively prohibits pre-viability abortion, as Casey rules out.

More important, Casey forbids undue burdens (and prohibitions), not on the ability to abort at some point before viability, but on the ability to abort at any given point until then.  And it does so on the absolutist ground that "[b]efore viability, the State's interests are not strong enough to support a prohibition of abortion."  (This wasn't dictum.  Repeated in various terms, it was the heart of Casey's rationale for the viability-based test that Casey then applied to a range of pre-viability regulations, precisely to see if any amounted to an unlawful "prohibition of abortion" "before viability.")  That rules out Mississippi's prohibition throughout a solid two months before viability, from week 15 to 24.

But what if Dobbs turned Mississippi's misreading of Casey into the new rule?  For legal authority of a sort, Dobbs could appeal to the example of Casey itself.  Casey rejected Roe's trimester framework as "rigid."  Dobbs might say that of Casey's viability rule.  To replace Casey's "don't preclude abortion before viability" rule, maybe Dobbs could say, "don't 'unduly' narrow the window for aborting."  But I doubt this would sit well with the Roberts Court, for several reasons.

For one thing, this approach would layer vagueness upon vagueness.  Casey was ridiculed for using the vague "undue burden" concept to tell when a regulation makes abortion too hard.  (Since "unduly" is synonymous with "too…", saying that a burden goes too far if it's undue is like saying that it goes too far if it goes too far.  Casey's other phrase, "substantial obstacle," was little help, though I've elsewhere tried to reconstruct that concept and its cognates in other bodies of law.)  But at least Casey had a crisp rule for when abortion shouldn't be "unduly burdened":  pre-viability.  By contrast, the test we're imagining would deploy the vacuous "undue" concept for that, too.  So it would forbid states to impose (1) unduly burdensome (hard-to-clear) obstacles, to abortions obtained past an (2) unduly burdensome (early) cutoff point in the pregnancy.  If one "undue-ness" test was bad, two may be unacceptable to Justices open to Mississippi but keen to avoid endless re-litigation.

The Court could avoid this indeterminacy by drawing an arbitrary sharp line:

"To honor the spirit but not the letter of Casey, which honored the spirit but not the letter of Roe, which sought to honor the spirit if not the letter of the Fourteenth Amendment, we must guarantee women a fair opportunity to abort.  And we must otherwise leave states free to protect life.  Yet we would rather not have the spirits of Roe and Casey haunt our docket indefinitely, through endless litigation of a vague 'fair opportunity' test.  So we declare that states may prohibit abortion starting [three? four? six?] weeks after a woman has learned or should have learned of her pregnancy.  The Fourteenth Amendment does not fix that number.  But we choose it in order to give pregnant women and states clear notice of their rights, and the lower courts clear guidance."

Setting aside whether card-carrying originalists like Kavanaugh could put their names to this, the Roberts Court might find it unappetizing for several reasons previewed above.

First, any version of this approach would either be unadministrable or feel arbitrary.The state laws permitted by the version above would be hard to administer, requiring officials to figure out when a woman knew of her pregnancy.  To avoid that, the Court could simply allow prohibitions after a fixed point in pregnancy.  But then it would have to permit emergency (very emergency) as-applied challenges brought by women who couldn't learn of their pregnancy in time to get the fair opportunity to which they'd now be constitutionally entitled.  And either of these numerically specific "fair opportunity" rules would purchase determinacy at the price of seeming legislative.  The Court could mask the arbitrariness by refusing to say how long is needed for a fair opportunity.  (It could just say that 15 weeks is enough.)  But that minimalism wouldn't last long.  States are pushing the envelope.  Does Missouri's eight-week ban, now under review by the Eighth Circuit, leave a fair opportunity?  If the Eighth Circuit (with 10 of 11 active judges appointed by Republicans) says yes, the Court would have to disagree, or agree, or telegraph agreement by not taking the case.

So the minimalist version of a "fair opportunity" decision might be a Pyrrhic victory for de-politicization.  For one moderate (if contrived) ruling now, it would ensure that the hard issues came back in a string of cases, each more contentious than the last, as the stakes kept growing with more aggressive laws.  Nor could the Court look apolitical by refusing to hear later challenges.  Its refusal to review rulings that upheld sweeping bans may prove nearly as controversial as outright approval.

Second, any "fair opportunity" ruling would make it hard for the majority to later abolish any constitutional abortion entitlement, if that is indeed their goal.  For the ruling's first part, crucial to its holding, would not only limit the existing right, but double down on a new, narrower right.  ("To honor the spirit of Casey… we must guarantee women a fair opportunity to abort.")  Having rejected Casey, as Mississippi's petition asked them to do if needed, the Justices would own Casey's replacement.

Put otherwise, it's one thing to pointedly refuse to extend a precedent today, and scrap it tomorrow.  The Roberts Court has done that more than once.  It's another thing to reject a precedent's main holding, and so have to justify a replacement, before coming back to reverse even that replacement.  The latter requires manifest self-contradiction.  The Court did the first thing with (to take just one example) Abood—which it cast aspersions on in Harris before overruling in Janus.  But the Janus majority could say it wasn't contradicting its earlier position in Harris, which had merely "refused to extend Abood to situations where it does not squarely control … while leaving for another day the question whether Abood should be overruled."  Dobbs couldn't do this with Casey.

Someone might say it could—that Dobbs could arrive at a "fair opportunity" rule by rejecting part of Casey while just bracketing the rest (not endorsing it).  On this view, Dobbs could scrap Casey's rule against pre-viability bans, while leaving undisturbed (as undisputed in this case) Casey's protection of a woman's "ultimate control over her destiny and her body."  (NB:  Dobbs couldn't leave intact Casey's similar-sounding "ultimate decision" language, since context shows that this always refers to a right to abort at any stage pre-viability, which is disputed in this case.)  But Dobbs couldn't actually rest there.  First, having rejected a key part of Casey, Dobbs couldn't leave this "ultimate control" part in place merely on Casey's authority as precedent.  The Justices would have to say why the "ultimate control" part warranted more respect than the rest of Casey.  And whatever they said would necessarily be in their own voice.  Second, the "ultimate control" principle isn't self-executing, as the Casey plurality recognized.  Thus, to confirm that Mississippi's law complied with it, the Court would have to give this principle effect with a new rule along the lines sketched above.  It would then own that rule, just as the Casey plurality owned the rule it adopted to give effect to the same principle.  Dobbs would thus become the new Casey—and Casey was no stepping-stone to reversing Roe.

Finally, this approach might seem (to the key Justices) to contradict Roberts' own reason for reversing a different precedent in full, not in part:  "It should go without saying … that we cannot embrace a narrow ground of decision simply because it is narrow; it must also be right."  There's a reason it's rare to see what Justice Scalia called Casey's "[un]heard of" "keep what you want and throw away the rest version" of stare decisis, which this tack in Dobbs would also require:  For obvious reasons, that version is systematically likelier to lack a coherent foundation (as Scalia thought Casey did).

And it might well lack one here—as might any of the above efforts to move the "undue burden" test back from viability, but not too far back.  Each rejects Casey's straightforward rationale that "[b]efore viability, the State's interests are not strong enough to support a prohibition of abortion."  Each presupposes instead that the state's interests begin to trump sometime before viability.  Yet the most obvious asserted state interest that exists anytime pre-viability, is the interest in protecting the individual life at hand.  And that interest, if legitimate at all, is constant throughout the pre-viability phase (and beyond).  It's imperiled as much by a law that leaves ten or 15 weeks to decide to end the life, as by a law that leaves 22 or 24.  (Indeed, making sure that pregnant women could override this supposed interest would be the express point of any "fair opportunity" ruling.)  Conversely, if Dobbs continues to forbid states to pursue an interest in saving the life at issue, what interest is there to justify bans on abortions shortly before viability?  There's no consistent case for a "fair opportunity" test.  Unless Dobbs can find an interest that does get strong enough only right before viability…

2. Base a new test on the Blackmun-Stevens sliding scale of state interests? In a case decided just before Casey, Justice Harry Blackmun (Roe's author) quoted an earlier opinion by Justice John Paul Stevens for just such a view.  Both thought the state interests that matter are those that strengthen over time, as the fetus becomes more like a born human being.  (The risk of complications for the woman also grows.)  If so, then there may come a point at which the interest in aborting is outweighed by the interest in banning.

Blackmun's and Stevens' opinions provide some legal pedigree Dobbs could cite for this rationale.  But again, Dobbs would still have to translate this rationale into a test that allowed a 15-week ban and not too much more.  And ideally, given the Roberts Court's concerns for lawyerliness and legitimacy, the test would have its own basis in some legal source.  (I'll relax this demand in the final section.)  That would help overcome the arbitrariness plaguing this approach in picking (a) which interests the state may act on, and (b) when those are outweighed by the woman's interests.  What basis for resolving these questions would find roots in history or precedent

A. First trimester? Drawing the line at the first trimester (12–13 weeks) would save the 15-week ban and might seem to have a basis in precedent, in Roe.  But Roe let states regulate in the second trimester only to guard the mother's health.  It offers no support for the idea that the interest in protecting life gets strong enough.  Besides, Roe's trimester latticework of rules was so ungrounded that Casey refused to retain it even on stare decisis grounds.

B. Quickening? Could the Justices draw the line at quickening (when a woman begins to feel fetal movements), based on its historical relevance to some common law protections of pre-natal life?  No.  The quickening argument wasn't briefed, much less passed on below, perhaps because quickening may happen too late to save Mississippi's ban.  More to the point, quickening was used historically only for some forms of legal protection before birth, not all—and only for a now-obsolete reason:  the evidentiary difficulties of establishing a pregnancy earlier.  Beyond that, as a metaphysical argument, the old folk belief that at "quickening" an entity is "animated"—or a "brute animal" becomes human—is discredited by modern science.  Both sides grant that an embryo or fetus is a living member of Homo sapiens—whatever its moral status—from day one, not week 16.

Trimesters and quickening aside, there's no line between zero and 15 weeks with a historical or precedential (or other legal-sounding) basis for deciding when fetal life may trump.  The alternative is for the Justices to do the moral math themselves.  Imagine:

"Attached to this opinion is an appendix containing a graph with two curves (derived from the Fourteenth Amendment's Due Process Clause).  The upward sloping line measures the state's growing interest in protecting fetal life over time; the horizontal line reflects a woman's constant interest in choosing whether to abort.  The lines cross at 13.5 weeks of gestation, so this 15-week law passes muster."

The Chief, who may be the keenest to avoid an all-or-nothing decision, wrote just last year that in abortion cases, "the benefits and burdens are incommensurable."  As for balancing them, he quipped, "they do not teach such things in law school."  No Justice ruling for Mississippi would disagree.

3. Draw a novel developmental bright line? The Court isn't strictly bound by its own precedents, and sometimes devises tests without a legal basis.  It could pick a novel switchover point between zero and 15 weeks.  But to make up for its lack of legal pedigree, this approach would at least have to rest on a bright-line, qualitative fetal development.  I see two options.

A. Heartbeat? The emergence of a heartbeat seems morally relevant to some.  But it occurs at six weeks, before many women realize they're pregnant.  As a basis to allow prohibitions, it would cut Roe down to a constitutional right to take the morning-after pill.

B. Pain? The fetus's capacity for pain might seem to implicate a new state interest, and so justify greater restrictions.  But implementing this approach would require the Court to draw still more bright lines (along at least two spectrums) without any legal authority to cite for support.  It wouldn't easily justify a prohibition as early as Mississippi's anyway.  Most problematic, as seen below, a pain-prevention rationale would be incoherent, contradicting its own premise that abortion is a fundamental right.  And recasting the test to get around that would require unconstrained philosophizing, and invite more in future cases.

a. The practical case for a pain-based standard. Some think the fetus's attainment of a capacity for pain grounds a new reason to regulate:  pain-prevention.  But to begin, gaining sentience is a gradual process—actually, a series of gradual developments—as explained by the medical sources cited by Mississippi.  The landmarks stretch from the earliest weeks, when the "neural circuitry" needed for "primitive" responses to pain develops, to 20 weeks, when hormonal stress responses occur, and on from there.  Sooner or later, the Court would have to make a philosophical choice about which kind of pain-sensitivity to declare legally sufficient (but on what publicly compelling basis?).  And to save Mississippi's law, the Court would have to justify a type of sentience that kicks in before 15 weeks—even though the more commonly targeted sort of pain-response arises at 20 weeks.  (More than a third of the states have attempted 20-week bans.)  Then the Court would have to decide how risk-averse states may be in trying to prevent that pain, in the face of inevitable uncertainties about what is or isn't going on in the psychology of a fetus of a given age.  On none of these issues could the Court rest its answer on history or precedent or any other apparently legal foothold; there is none.

Finally, this approach is incoherent.  It is hornbook law that states may not prohibit exercises of a fundamental right for the sake of interests they could achieve by less restrictive means.  Yet the ruling we're imagining now, while presupposing that abortion is a fundamental right, would let states ban abortion to prevent pain they could just as well prevent without a ban:  They could require abortion providers to anesthetize the fetus (absent a heightened risk of complications for the woman).  Some jurisdictions already do so, and anesthesia is routinely used for fetal surgeries.

Of course, requiring this extra step would itself burden women's right to abort, and the merits of such a policy would be intensely disputed.  But the narrow, doctrinally relevant point is that requiring anesthesia for later abortions would be far less restrictive than a flat ban on the same abortions.  That's why Dobbs couldn't bless such bans to prevent pain, and still treat abortion as a fundamental right.

The Court might manage to avoid addressing this objection for now.  And it might remand for the lower courts to take a first pass at the line-drawing problems for any pain-based test.  But it couldn't leave these (entirely foreseeable) questions to trial and circuit courts for long.  As under the cagey version of a "fair opportunity" ruling, explored above, states would keep pushing the envelope.  So would friendly lower courts.  Refusal to countermand them would itself look political.  And so on.

b. The philosophical case for a pain-based standard. Some might say that the capacity for pain matters not because it implicates a new state interest in pain-prevention, but because it imbues the fetus with a higher moral status.  But once preventing actual pain isn't the point—once the focus is on capacity—it'll be hard for the Court to say why only the capacity to immediately feel pain should count.  Getting there is just one stage in the long and gradual development toward a functioning nervous system.  In that respect, a late fetus's "capacity" and an embryo's differ only in degree.  To say that the one bestows moral status but not the other would require a nakedly philosophical choice.  (Pause and try to imagine how Roberts or Kavanaugh would write a paragraph supporting this choice without any legal citations.)  And it would open the Court up to purely philosophical arguments in future cases about what else might bestow moral status.  So this tack would undermine legitimacy and further politicize (indeed, philosophize) the Court's docket.  Who on this Court would welcome that?

Is there any other qualitative and relevant change between the emergence of a heartbeat at six weeks, and Mississippi's cutoff at 15?  Even Casey would admit not.  Indeed, across a whole pregnancy, the plurality there plausibly confessed, "no line other than viability … is more workable."  (And even for that line, Casey's own rationale was unilluminating:  When the fetus can live outside the womb, it said, "the independent existence of the second life can, in reason and all fairness, be the object of state protection that now overrides the rights of the woman."  In other words, a viable life can justify banning abortion because … viability can fairly justify banning abortion.  As John Hart Ely wrote of Roe's identical case for viability, this "seems to mistake a definition for a syllogism.")  But if Casey strained to justify a line at viability, and thought others clearly too arbitrary, no Justice inclined to move the line back for Mississippi in Dobbs will likely find another better.  If the Roberts Court is no less formalist than the Casey Court, it may have only two live options in Dobbs.