Although Supreme Court cases often leave room for various interpretations, seldom does the Court take a case solely to instruct lower courts on the best way to read a prior opinion.  In Hughes v. United States, the Supreme Court will consider both how to read each opinion in the prior case of Freeman v. United States and how to determine which opinion is controlling.  By resolving how to handle fractured cases with no majority rationale, Hughes will also offer meta guidance on how to read cases more generally.

Litigants and scholars have offered various solutions for determining the precedential effect of non-majority opinions.  This post will detail those solutions, which entail the finest, purest legal logic in a fantastically meta case.  My current thinking is that the Court should apply the “logical subset” rule to govern fractured opinions in most cases, except if the logical subset rule produces absurd results.

 

The Facts of and Decision in Hughes

 

Hughes involves a problem that arises when no majority rationale governs a Supreme Court decision.  In those types of cases, Marks v. United States instructs lower courts that the holding of the Supreme Court is “the position taken by those Members who concurred in the judgments on the narrowest grounds.”  What that means has baffled courts, professors, and students for decades.

Case in point:  Hughes v. United States.  In Hughes, the defendant pleaded guilty to drug and firearm offenses and entered into what is called a C-Plea agreement, after the federal rule that allows this type of plea.  Subsequently, the Federal Sentencing Guidelines range was lowered for crack-cocaine offenses like the one with which Hughes was charged.  Hughes then sought a sentence reduction allowed for defendants “sentenced to a term of imprisonment based on a sentencing range that has subsequently been lowered by the Sentencing Commission.” (emphasis added.)  The key question is whether, even though Hughes had been sentenced pursuant to a C-Plea, his sentence was still “based on” the Guidelines sentencing range.

A prior case, Freeman v. United States had confronted this exact question, but did not receive a five-member majority for any rationale.  Justice Kennedy’s four-Justice plurality opinion held that a plea agreement is generally based on the relevant Guidelines sentencing range because judges use the Guidelines to determine if the sentence specified in the plea agreement is appropriate.  (To complicate matters, some lower courts have interpreted Justice Kennedy’s opinion more narrowly, to mean the judge has to evidence that he/she is using the Guidelines to gauge an appropriate sentence.)  Justice Sotomayor’s one-Justice concurrence reasoned that a plea agreement is “based on a sentencing range” if the plea agreement itself “call[s] for the defendant to be sentenced within a particular Guidelines sentencing range.”

Because Freeman’s plea agreement noted the sentencing range and there was evidence that the judge relied on the sentencing range, both the plurality and the concurrence formed a majority on the result, which held that Freeman could receive a sentencing reduction.  The dissent would have held that plea agreements are never based on the Guidelines range, because the appropriate sentence is suggested by the government, within the plea agreement.  Thus, according to the dissent, sentencing reductions are never permissible for plea agreements where the Guidelines range has subsequently been lowered.

Whether to apply the plurality’s or the concurrence’s rationale in Freeman has split the circuits.  In Eleventh Circuit, in Hughes, held that Justice Sotomayor’s concurrence is controlling because it is the “narrower grounds” under Marks.  A smaller subset of defendants would have their sentences reduced under Justice Sotomayor’s rationale (cases where the plea agreement mentions the Guidelines range) than under Justice Kennedy’s plurality (cases where, under one reading, every defendant’s sentence is based on the Guidelines range, and in another, a judge just has to in some way rely on the Guidelines range).

Most courts of appeals have also held Justice Sotomayor’s opinion is controlling as the narrowest grounds” upon which the Freeman opinion was based.  Two circuits have, however, found that Justice Sotomayor’s opinion is not a “logical subset” of Justice Kennedy’s plurality because there are some scenarios in which a defendant could not seek a reduction under Justice Kennedy’s rationale but could seek a reduction under Justice Sotomayor’s rationale.  If one reads Justice Kennedy’s plurality opinion to require that judges sentencing pursuant to a C-Plea agreement mention the Guidelines or indicate reliance in some way in order for a sentence to be reduced, then a plea agreement that mentions the Guidelines but where the sentencing judge does not note the Guidelines would be reduced under Justice Sotomayor’s concurrence but not under Justice Kennedy’s plurality.  Thus, there is only some overlap between a narrow reading of Justice Kennedy’s opinion and Justice Sotomayor’s opinion.

The Eleventh Circuit was not troubled by this.  In Hughes, the Eleventh Circuit held that the “logical subset” rule is not required by Marks.  Instead, the Eleventh Circuit reasoned that, for fractured Supreme Court opinions, Marks requires the lower courts to apply a rule that would lead to the result with which most Justices would agree.

The Different Ways of Reading Fractured Opinions

Scholars and courts have proposed a variety of solutions for determining what is meant by the “narrowest grounds” upon which to determine binding precedent.  One scholar has also proposed abandoning Marks altogether – arguing that precedent should be created only when a majority of the Court actually agrees on a rationale.  Below, I detail the proposals and their virtues and vices.

1. The “Logical Subset” and “Partial Overlap” Rules:

This approach is the one I think best captures whether a particular opinion is the “narrowest grounds” of a fractured Court.  If Justice Kennedy’s plurality would allow all plea agreements to be reduced as “based on” the Guidelines, then Justice Sotomayor’s concurrence is, logically, narrower than that ruling.  Thus, Justice Sotomayor’s concurrence should be the binding precedent.  However, if Justice Kennedy’s plurality is interpreted to allow sentence reduction only if the judge mentions the Guidelines, then there is no logical subset.

Given no logical subset, the partial overlap rule could be triggered.  In that case, a court could hold that a defendant is permitted a sentencing reduction only if both the plea agreement and the judge mention the Guidelines.  Or, a court could hold that if both the C-Plea and the sentencing judge mention the Guidelines, then a reduction is certainly permitted, but there is no binding precedent on what to do if only a judge or the plea agreement mention the Sentencing Guidelines (because, in those cases, there is no overlap between the fractured opinions).

Scholars have noted problems with the logical subset approach.  Professor Adam Steinman argues that the logical subset approach does not account for biconditional rules.  Biconditional rules involve holdings that specify both when a particular result should occur and also the opposite, when the result should not occur.  Thus, a rule would specify both that A leads to B and -A leads to -B

Hughes involves a biconditional rule.  Justice Sotomayor reasoned that a plea agreement is based on the Guidelines (and thus deserving of a reduction) when it mentions the Guidelines, and that a plea agreement is NOT based on the Guidelines when it does not so mention them.  Justice Kennedy, under some readings, reasoned that a plea agreement is based on the Guidelines if either the plea agreement mentions the Guidelines or the judge relies on the Guidelines in some other way, and not deserving of a reduction otherwise.  The problem is, for the affirmative framing of when a sentence can be reduced (A leads to B), Justice Sotomayor’s opinion is the logical subset.  However, for the negative, when a sentence cannot be reduced (-A leads to -B), Justice Kennedy’s opinion is the logical subset.  Justice Kennedy would hold that a sentence is not reduced if neither the judge nor the plea agreement mentions the Guidelines; that is a logical subset of Justice Sotomayor’s negative rule that the sentence is not reduced if the plea agreement does not mention the Guidelines.

According to Steinman, the presence of biconditional rules means that whether the rule is framed in the affirmative or the negative determines which opinion is the logical subset.  Although Steinman’s reasoning is incisive and provocative, I am not particularly troubled by this.  All a court must do is determine the default setting to decide whether to frame the rule in the affirmative or in the negative.  If the default is that a criminal sentence should not be retroactively reduced, then the rule should be framed in the affirmative, when a sentence can be reduced, to determine the narrower grounds.  If the default is that criminal sentences are retroactively reduced, then the negative articulation of the rule (when a sentence should not be reduced) should be judged based on the logical subset rule.

Thus, the Court should acknowledge biconditional rules.  Steinman is right to point out that they present particular problems.  However, I do not believe these problems not intractable.

2. Abandon Marks:

There is another problem with the logical subset rule.  Professor Richard Re, in an amicus brief, compellingly argues that the logical subset rule is based on a logical fallacy.  Just because a rule is a subset of a larger rule does not mean a majority of the Court would accept the narrower rule.  He thinks binding precedent should be created only for reasoning that captures a majority of the Court.

His example is illustrative.  Consider a case where three Justices hold that courts can never regulate protests, but two Justices hold that courts can regulate protests only if the subject of the protest is a particular political view, like Communism.  Four other Justices, in dissent, would hold that you can always regulate protests.  The two-Justice concurrence, which would allow viewpoint discrimination, might be even less desirable to the three-Justice plurality than a rule allowing for more regulation of protests.  Thus, as Re notes, “[w]hat is true of the whole is not necessarily true of its component parts. For example, salt is edible, but one of its components (chlorine) is toxic. Likewise, approval of a broad rule does not imply approval for any of its component rules or outcomes.”

I think the solution to this should be that the logical subset rule should be followed, but only if it does not lead to absurd results, such as in the First Amendment example.

Abandoning Marks, although an intriguing solution, would leave the lower courts utterly without guidance in cases where there is no majority opinion.

3. The “Results” Approach:

During oral argument, the Supreme Court seems to have preferred the results approach employed by the Eleventh Circuit.  Under this approach, determining the “narrowest grounds” for Marks purposes would simply entail deciding how a majority of the Supreme Court would decide a case, given the fractured opinions.  This seems, to me, contrary to rule of law principles.  A case’s decision should be based on reasoning that leads to results, not determining a preferred result and reverse-engineering reasoning.

Although the results approach allows the Justices to preserve a common-sense understanding of Marks, it lends no further clarity to determining what the narrowest grounds are.  The results approach is also nonsensical because Marks considers only the reasoning of the Justices who ultimately agreed with the result, but a pure prediction method would also need to count dissenting Justices in a fractured opinion.

 

I look forward to the Court’s opinion in Hughes and hope the Justices will provide clear guidance on a challenging meta-issue.