What Justice Powell's Papers on His Opinion in Marks Tell Us About the Marks Rule

Yesterday, I got the kind archivists at Washington and Lee Law School to digitize Justice Powell's case file in Marks v. United States, which he wrote.  The entirety of his rather slender file on Marks can be viewed here.  Marks was a fairly unnegotiated opinion that began its life as a per curiam, and Justice Powell's papers don't tell us a great deal about what he understood his famous narrowest-grounds rule to mean.  (Of course, even if they did, it's doubtful that evidence about what Justice Powell understood Marks to mean would tell us anything about what Marks means; cf., as Scalia was fond of pointing out in this connection, Justice Stevens's woeful misunderstandings of Chevron, which he wrote.)  However, they do show that Justice Powell was at least inclined to reject predictive, fifth-vote approaches to Marks that attempt to identify what Judge Kavanaugh has creatively called the "Marks middle ground" of a fractured opinion, as opposed to the narrowest grounds Marks actually says matter, by determining which Justices' views would occupy the outcome-determinative center of gravity in future cases on the same subject matter. 

A.  An historical prolegomenon (which you can skip if you already know all about Warren/Burger Court obscenity law and the background of Marks).

To begin with, if you want to understand Marks you should take some time to understand why it was that almost 200 years into its history, the Court finally got around to announcing a rule about the precedential effect of its plurality opinions, which is something that people who work with Marks rarely bother to do.  The answer has to do with the circular evolution in the Warren and Burger Courts' obscenity law, which eventually put the Burger Court in the position of working out the jury instructions that pornographers ought to receive for pre-Burger-Court-precedent conduct under fractured Warren Court precedent.  

In the beginning of the Warren Court's evolution on obscenity, Justice Brennan, who was then a moderate on the subject, wrote a 1957 opinion for the Court in Roth that held material was obscene and unprotected if "to the average person, applying contemporary community standards, the dominant theme of the material taken as a whole appeals to prurient interest."  By 1966, however, Brennan's views had liberalized and he was prepared to hold, in Memoirs v. Massachusetts, a case about suppressing the publication of a pornographic 18th-century novel, that material wasn't obscene unless it appealed to the prurient interest and was "utterly without redeeming social value." 

Unfortunately (speaking only from the perspective of predictability, not the merits of Brennan's position), Brennan was unable to get five votes for that view.  Warren, who wrote a confusing separate opinion in Roth that had voiced some inchoate concern about applying obscenity law to great art, joined his plurality opinion, as did Justice Fortas, who joined the Court after Roth.  But Justice Stewart would not join; he instead would have protected the Memoirs on the ground that they weren't "hard-core pornography."  It's unclear whether that test was more or less demanding than Brennan's (it's possible the answer is both); in an attempt to "prevent any possible misunderstanding," Stewart "set out in the margin a description, borrowed from the Solicitor General's brief, of the kind of thing to which I have reference."  That way, one senses Stewart thinking, if you didn't want to read the SG's scandalous descriptions of hard-core pornography, you could just read the body of his opinion.  The footnote makes amusing reading but doesn't clarify much.  Then there was Justice Black, who concurred in the judgment on the ground that the First Amendment banned all obscenity prosecutions, complaining that the plurality's "utterly without redeeming social value" test was "as uncertain, if not even more uncertain, than is the unknown substance of the Milky Way."  Finally, Justice Douglas also concurred in the judgment, also reasoning that the First Amendment banned all obscenity prosecutions, on the basis of a sort of self-congratulatory anti-majoritarian snobbishness:

Every time an obscenity case is to be argued here, my office is flooded with letters and postal cards urging me to protect the community or the Nation by striking down the publication. The messages are often identical even down to commas and semicolons. The inference is irresistible that they were all copied from a school or church blackboard. Dozens of postal cards often are mailed from the same precinct. The drives are incessant and the pressures are great. Happily we do not bow to them. I mention them only to emphasize the lack of popular understanding of our constitutional system. Publications and utterances were made immune from majoritarian control by the First Amendment, applicable to the States by reason of the Fourteenth...

That's Wild Bill Douglas for you.  (On the same day, he wrote an opinion defending "the masochistic yearning that is probably present in everyone and dominant in some"—too much information!—on the ground that "some like Chopin, others like 'rock and roll.'")  Anyway, by 1966 and until 1973, a majority of the Court couldn't agree on a test for obscenity; in the interim, the Court famously retreated into its film room, watched the pornographic films for which petitioners had been convicted, and then issued unexplained summary affirmances or reversals on the basis of their film-room judgments.  

By 1973, however, Justice Fortas had been forced off the Court, Warren had retired, Black had died, each had been replaced by conservative Nixon nominees (including Blackmun, at the time), and Brennan no longer had the votes against a looser approach to obscenity, nor was willing any longer to defend his own barely non-absolutist approach.  So it was that the Court more or less circled back to the obscenity test it had announced in Roth and held in Miller v. California that works which, taken as a whole, appealed to the prurient interest and lacked serious literary, artistic, political or social value were obscene.  At a time when obscenity was still vigorously prosecuted, this was a big deal; while Miller was pending, Burger and Nixon discussed the case at length in a recorded exchange where neither comes off very well.  (Burger on Memoirs: "This means if they, uh, if they have one of these outrageous orgies, then if they mention Vietnam or the condition of the, uh, ghettos, it redeems the whole thing!")

B.  Then Marks

In the aftermath of Miller, the courts of appeals were initially unanimous that Miller's obscenity standard didn't apply retroactively to pre-Miller conduct.  If Miller had substantively changed obscenity law, they reasoned, it couldn't apply retroactively, and while Miller wasn't much different from the prior Roth standard, it was meaningfully different, as the Miller Court emphasized, from that of the Memoirs plurality.  The only question was whether the Memoirs plurality's standard was the law after Memoirs and before Miller, and the courts of appeals had little difficulty concluding it was; if the crucial third, fourth, and fifth votes in Memoirs protected the titular Memoirs on the ground that the Memoirs weren't utterly without redeeming social value, lower courts were bound to protect materials that weren't utterly without redeeming social value.  However, when Stanley Marks was federally prosecuted for a barely pre-Miller screening of an adult film, a Kentucky district judge instructed the jury under Miller, and the Sixth Circuit affirmed on somewhat obscure grounds, setting the stage for the Court to decide what the law was in the pre-Miller period and what the precedential effect of its fractured opinions was generally.

When the Court granted certiorari, the Solicitor General, who was then Robert Bork, decided to confess error.  (It's not at all clear that Bork did so at the cert stage.)  Weeks before Bork filed his Frank Easterbrook co-authored brief, the Court decided Gregg v. Georgia, a case that also turned on interpreting fractured precedent.  Gregg and a cluster of consolidated cases grouped under it involved  constitutional challenges to the death penalty; several years before, the Court had decided Furman v. Georgia, a case in which a coalition of abolitionist Justices and death-penalty moderates struck down death-penalty statutes that the moderates deemed excessively arbitrary.  Gregg would itself lack a majority opinion, but Justice Stewart would write for a three-Justice plurality in a footnote that "[s]ince five Justices wrote separately in support of the judgments in Furman, the holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds— MR. JUSTICE STEWART and MR. JUSTICE WHITE," and generally endeavor to apply Justices Stewart's and White's views to the statutes before the Court in Gregg.**  The members of the Gregg plurality apparently took Furman's stare-decisis effect fairly seriously; Justice Powell, who dissented in Furman but joined Justice Stewart's plurality in Gregg, would tell the conference after argument that he "accept[ed] Furman as precedent" and note that the states before the Court had "endeavored to meet the views of Stewart + White," though he was unsure if the two states whose statutes the plurality ultimately invalidated satisfied Furman.

** By the way, the fact that Furman headlined a trio of consolidated cases with multiple judgments explains why the Marks formulation, which is just a quote of this footnote, awkwardly talks about "Members who concurred in the judgments on the narrowest grounds."

Bork had almost certainly decided to confess error before Gregg; he would report to the Court at oral argument that three years prior he had "communicated" the government's position that pre-Miller conduct should be prosecuted under the Memoirs plurality standard, but that somehow that communication hadn't reached the U.S. Attorneys.  But the Gregg footnote almost certainly modified Bork and Easterbrook's analysis, or rather, awkwardly interrupted it.  In the middle of the relevant section of their brief, which is accessible on Westlaw, Bork and Easterbrook plopped this acknowledgement of the Gregg footnote:

The Court has recognized that when no position commands a a majority of the Justices, the rule of the case is expressed by the most narrow view of a Justice or group of Justices concurring in the disposition.  See Gregg v. Georgia, No. 74-6257, decided July 2, 1976, slip op. 12 n.15, 38-39 n.47 (opinion of Stewart, Powell and Stevens, JJ.); Roberts v. Louisiana, No. 75-5844, decided July 2, 1976, slip op. 9-11 (White, J. dissenting).  The views of the plurality in Memoirs consequently became the prevailing rule.

Everything they said before and after that paragraph made a different argument altogether, namely that the Memoirs plurality was the law before Miller because "[b]etween 1966 and 1973 . . . no conviction could obtain the five votes necessary for its upholding unless it satisfied the tests laid down by the Memoirs plurality."  Bork and Easterbrook's argument for this wasn't really that Memoirs was a precedent, but that it was no precedent at all; they claimed that in the post-Memoirs, pre-Miller period, the Court didn't follow whatever Memoirs' holding was, but "reverse[d] obscenity convictions based on the views of each individual Justice" (a fact which the Court didn't make public until it decided Miller, though hints at this decisional approach were contained in the Court's 1967 opinion in Redrup).  Since the Justices on the Memoirs Court continued to vote their individual Memoirs positions as if Memoirs had never happened, lower courts, they reasoned, were obliged to apply the Memoirs plurality view because no conviction could ultimately be sustained by the Court in these years without satisfying the Memoirs plurality.  And people in Marks's position would have expected their conduct to be judged by the Memoirs plurality's standard, as only satisfaction of that standard would satisfy the Court, should it choose to review their convictions.

This frankly retrospective-predictive approach to figuring out what the public would have understood obscenity law to be from 1966–73 has an obvious defect (besides the total obscurity to the public of what the Court was doing from 1966–73); by 1970, two of the three members of the Memoirs plurality were gone, replaced by members of the Miller majority.  It shouldn't have been at all obvious to Marks that a pre-Miller Court sitting in instant judgment of his February 1973 conduct would have held it protected unless it satisfied the by-then defunct Memoirs plurality; after all, just four months later the same nine Justices would decide Miller.  Indeed, on December 14, 1971, eight days after Justice Powell replaced Justice Black and four days after Justice Rehnquist replaced Justice Harlan, completing what would become the Miller majority, the Court issued the last of its unexplained obscenity summary reversals.  The next time it spoke on the subject before Miller, it issued a reasoned per curiam that reversed a conviction under the Roth standard, not mentioning Memoirs at all.

These flaws in Bork's reasoning notwithstanding, when he got around to arguing Marks, he found an extremely receptive audience for his confession of error; four out of the five members of the Miller majority were happy to deny Miller retroactive application and apply to pre-Miller conduct a test that they had complained in Miller was "virtually impossible to [satisfy] under our criminal standards of proof" and doubted "had any meaning at all."  In fact, the Court was so receptive to Bork's confession that it didn't even bother to appoint an amicus to defend the Sixth Circuit's judgment.  This is in some ways rather remarkable given that after Memoirs itself the Court never publicly applied the Memoirs plurality test, and that the three senior members of the Miller majority regularly dissented from the Court's summary reversals tacitly applying Memoirs.  Only Justice Rehnquist gave Bork a hard time for declining to defend what Rehnquist thought an eminently defensible position.  And at conference, only Rehnquist voted to affirm the Sixth Circuit, "tentative[ly]" opining that before Miller there was "no firm precedent in this Court."

It is here where things get a little interesting.  After getting the assignment, and probably some time before the circulation of the first draft that we have in the file, though it may have been in response to a subsequent request (discussed below) by Justice Rehnquist to add a footnote to the circulated draft, Justice Powell suggested to his clerk David Martin, now a Professor Emeritus at Virginia Law, that he add a footnote citing some of the pre-Miller cases that followed the Memoirs plurality.  He explained this change in the following way:

I would make a change along the foregoing lines because changes in the personnel of the Court weaken the "five Justices" argument [i.e., the argument pushed by Bork and Easterbrook that before Miller, a conviction would have to satisfy the Memoirs plurality to be upheld by five Justices because the Justices continued to vote their respective Memoirs views after Memoirs].  The point is that the view of the Memoirs plurality was the holding of the Court and followed as such.

I believe this change was made at what would ultimately be the opinion's footnote 8, which string-cited the circuit cases following the Memoirs plurality after the opinion's statement that "[t]he view of the Memoirs plurality therefore constituted the holding of the Court and provided the governing standards. Indeed, every Court of Appeals that considered the question between Memoirs and Miller so read our decisions."  Indeed, it's possible that Martin, in addition to adding Powell's requested footnote, closely paraphrased Powell's note in the sentences I just quoted, though it's also possible that Powell was paraphrasing Martin's draft.  We can't tell because we don't have it.

This note perhaps isn't everything that those of us who detest fifth-vote approaches to Marks could want from Powell's papers, but it's pretty good.  We could imagine, I suppose, a theory of precedent as deference to smart people.  Such a theory could conceive of a fractured opinion of the Court as a sort of opinion poll of our best and brightest and most well-briefed judges.  On this theory, we would follow the "fifth vote" in a fractured opinion, the one "that occupies the middle ground between (i) the broader opinion supporting the judgment and (ii) the dissenting opinion," not because it would predict future results, but because the median view of the Supreme Court is likely to be the best.  If that were our rationale for the fifth-vote approach, it wouldn't matter that the approach has a predictive shelf life of the next retirement.

The Court, however, has never been arrogant enough to suggest this view of its precedents, such a view couldn't possibly explain the categorical obeisance we pay to them, and fifth-vote advocates don't argue for their approach in these terms.  Rather, they seem to believe, as Bork and Easterbrook argued, that their approach can predict what the Court would actually do with a given case before the lower courts.  As Judge Kavanaugh has put it, by identifying a "Marks middle ground," "lower courts will decide cases consistently with the opinions of a majority of the Supreme Court," whether that hypothetical majority is comprised of the swing vote plus a plurality, or a swing vote plus dissenters.  In fact, Judge Kavanaugh rather remarkably refuses to apply Marks to opinions in which the dissent doesn't address the issues addressed by the fractured majority, because then he can't tell who occupies the true middle ground and therefore can't tell what the Court would do with the case before him.

The most fatal of many flaws in that rationale for fifth-voteism is the one Powell pithily identified: that "changes in the personnel of the Court" rapidly denude the fifth-vote approach of any genuinely predictive power.  Instead of theorizing the narrowest-grounds rule he announced in Marks as a prediction about what the Court might have done before Miller with Marks' case had the Justices continued to vote their respective views and not follow Memoirs, Powell seems to have genuinely believed that the Memoirs plurality's view was the "holding of the Court" and was to be "followed as such" until overruled, even if there were no longer any reason to predict that the Court would continue to adhere to it.  It's also noteworthy that the Bork/Easterbrook "five Justices" argument is completely absent from Powell's opinion itself.  While Powell urged Martin to more or less parrot Bork's brief on other points, Bork and Easterbrook's extended argument for why the Memoirs plurality was the law before Miller never shows up in the opinion; only their almost parenthetical citation to the Gregg "narrowest grounds" footnote does.  Nor did anyone on the Court ever suggest that the Bork/Easterbrook argument make its way into the opinion.

It is also interesting to learn that the one Justice on the Court who initially found "no firm precedent" in Memoirs would ultimately join Powell's opinion because he became convinced by the "five Justices" argument that Powell internally rejected.  After (probably) Powell's note to Martin, Martin produced and Powell circulated a polished first draft that reads almost identically to the opinion we have today (with the exception of its having been drafted as a per curiam), complete with its citation to the delphic "narrowest grounds" formulation in the Gregg plurality's footnote.  Join memos began to pile in—Burger, who whatever his other faults, wasn't a humorless man, wrote to say that he thought the case important enough to merit a signed opinion after having assigned it as a per curiam, and that "[t]o show my bona fides, I would volunteer to sign it if you declined to do so!"—and a slightly modified draft went out, again unsigned.  Then Rehnquist wrote on January 14, 1977 to join on the condition that Powell add some language distinguishing a fair-warning case he had recently written.  Explaining why he now could "subscribe" to the view that Memoirs was the law pre-Miller, he initially praised the opinion for "writ[ing] up more persuasively than I thought could be done the arguments for reversal," but then went out of his way to give an argument for the result that was identical to Bork and Easterbrook's and was nowhere in the draft opinion:

[The obscenity statute's] broad language was necessarily confined by the decisions of this Court determining what is, and what is not, obscenity.  Although the formulation of that test in Memoirs never attracted a majority of the Court, a process of vote counting makes clear that after that decision and before Miller this Court would not affirm a conviction which did not satisfy the test stated by the Memoirs plurality.

It is possible that Powell's note was a reaction to this letter; the note is immediately preceded in Powell's file, or at least his archivist's digitization of that file, by a typescript of Powell's first stab at meeting Rehnquist's concern about distinguishing his opinion, which came in the form of a footnote, as did Powell's suggestion in his note to Martin.  I believe, however, that the two are unconnected and that the footnote Powell describes in his note to Martin and his proposed footnote addressing Rehnquist's problem are entirely different footnotes.  Either way, the only evidence we have that anyone on the Court joined Marks on predictive grounds involves Marks' least committed joiner.

Powell's note does not conclusively resolve what Powell would have thought of the disagreement between me and Professor Williams discussed in my last post.  To recap, Professor Williams thinks that under Marks, lower courts are free to take, and should take, the following untraditional approach to fragmented opinions:  follow the Court when the Justices concurring in the judgment would agree on the result, and freely choose between the rivaling concurring/plurality opinions in cases where they would disagree.  By "narrowest grounds," he thinks Marks meant, or should be read to mean that set of cases described by the overlap of differing grounds.  For example, if three Justices want to apply strict scrutiny to some kind of statute and two Justices want to apply intermediate scrutiny to it, a lower court must uphold a statute of that kind that passes both, must invalidate a statute that fails both, and can choose between strict and intermediate scrutiny if it passes the latter but not the former.  I think that that is the correct approach to fractured precedent, subject to one potential caveat that I haven't worked out yet, but that it is absolutely foreclosed by Marks.  

In my favor, I have the point that Marks is quite clear that the Memoirs plurality stated the holding of the Court, not that lower courts had to follow it only in cases where the plurality and the concurring Justices would agree.  For example, Marks remanded for the district court to instruct the jury under the Memoirs plurality standard, not to choose whether to  dismiss the case, as Black and Douglas would have done, or instruct the jury under Memoirs.  Williams, however, reasonably argues in the comments to my prior post that that's just because, or could be just because, the concurring Justices' views were precedentially foreclosed by Roth, a majority opinion that held that there was an obscenity exception to the First Amendment.  Since Black and Douglas were just two Justices, they had no power to overrule Roth by concurring in a judgment on their Roth-inconsistent grounds.  Hence, even under Williams's rule, the Memoirs plurality, and just the Memoirs plurality, was the law pre-Miller.  I don't think that Powell's note intimates any view on whether he concluded the Memoirs plurality was "the holding of the Court" for that rather elaborate reason, or for the reason that, as his opinion actually said, the Memoirs plurality concurred in the judgment on the narrowest grounds relative to the judgment, i.e., the grounds that would upset the fewest obscenity prosecutions.  I do think that the statement in the Gregg footnote, which formed the basis for the Marks rule, that "the holding of the Court [in Furman] may be viewed as that position taken by . . . MR. JUSTICE STEWART and MR. JUSTICE WHITE," not some compound of their position and other concurring positions, cannot be as easily explained away by recourse to some prior opinion ruling out the other concurring Justices' positions, since no opinion like that exists.

However, it is interesting to learn that Justice Powell's law clerk reasoned his way to the Memoirs plurality being controlling in what at first looks like the way that Professor Williams does.  In his bench memo, he wrote:

Miller did emphasize that the Memoirs tests were accepted by only three Justices.  This may have made it easier for five Justices in Miller to change the formulation, but it certainly cannot obscure the fact that the Memoirs tests were very much alive in the intervening years.  They were operative because the other two Justices who made up the Memoirs majority did not believe that the First Amendment permitted suppression of obscene materials at all.  (Their position is never mentioned by CA6).  "[T]he holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds. . . ."  Gregg v. Georgia, at 12 n. 15 (Opinion of Stewart, Powell and Stevens).

Of course, Martin doesn't say why it follows from the sweep of Black and Douglas's view that the Memoirs plurality test was "operative."  He could mean, as Williams reasons, that the plurality was operative because Black and Douglas's view couldn't operate in light of Roth.  Or he could just mean that the plurality was operative because it was narrower and Black and Douglas's view was broader.  The citation to the Gregg footnote suggests the latter, and so too does the opinion he produced, in which he only said of Black and Douglas that they "concurred on broader grounds" and "reiterated their well-known position that the First Amendment provides an absolute shield against governmental action aimed at suppressing obscenity"—not that their position was precedentially foreclosed.  Ultimately I think the Marks opinion itself affords only a smidgen of daylight to Williams, but that nothing in the papers shows that Powell and Martin didn't have his view in mind.  What the papers and the rejected approach taken in the Bork/Easterbrook brief do show is that Powell didn't understand his own opinion in terms of the predictive fifth-vote approach that several circuits use and for which other judges and scholars advocate.