Argument analysis: Justices hint at categorical approach to correcting forfeited Sentencing Guidelines errors

Sometimes, an appellate court uses oral argument to help it decide who ought to win. Other times, the justices know who will win, and oral argument becomes an opportunity for the judges to use counsel as a sounding board as to how the opinion should be written. Wednesday’s Supreme Court oral argument in Rosales-Mireles v. United States had the earmarks of the latter.

The issue in this case is whether a plain Sentencing Guidelines error should normally be corrected, even when the defendant has failed to object in the district court. It appears that six or seven justices will vote in favor of petitioner Florencio Rosales-Mireles, convicted of illegal re-entry into the United States and sentenced to 78 months in prison. Based on the probation officer’s inadvertent double-counting of a misdemeanor assault conviction, the district court incorrectly believed the sentencing range to be 77-96 months, and sentenced Rosales-Mireles to what it believed was nearly the bottom of the range. In fact, the pronounced sentence fell near the middle of the correct range, 70-87 months. Rosales-Mireles’ lawyer did not detect the error.

A panel of the United States Court of Appeals for the 5th Circuit affirmed the sentence. Because Rosales-Mireles had not objected to the error, Federal Rule of Criminal Procedure 52(b) required him to demonstrate four things: (1) an error; (2) that was clear or obvious; (3) that affected his substantial rights; and (4) that seriously affects the fairness, integrity or public reputation of judicial proceedings. The 5th Circuit panel found that Rosales-Mireles had satisfied the first three criteria, but not the fourth, primarily because the actual sentence happened to fall within the correct range.

As I remarked in my argument preview, in light of the Supreme Court’s 6-2 decision for the petitioner in Molina-Martinez v. United States in 2016 — another 5th Circuit case with almost identical facts — it was always going to be difficult for the government to get a different result in Rosales-Mireles’ case. On Wednesday morning, many of the justices’ questions were aimed at eliciting a bolder, more categorical argument from Rosales-Mireles’ counsel than had been offered in the merits brief. There, Rosales-Mireles had seemed to argue for a presumption in favor of correcting forfeited Sentencing Guidelines errors, with a case-by-case inquiry into when that presumption had been sufficiently rebutted.

Assistant Federal Public Defender Kristin Davidson was not far into her presentation when Justice Ruth Bader Ginsburg encouraged her to take the further step of arguing for a categorical approach to Sentencing Guidelines errors. “You have agreed with me that in guidelines miscalculations, the error should be corrected routinely, not sparingly,” Ginsburg said.  “I thought you were cordoning off guidelines miscalculations from all other errors.”

“No. Let me clarify,” Davidson replied, suggesting that a majority of the circuits have not gone so far as to treat a forfeited guidelines error as categorically eligible for correction. However, she added, “there’s a particularly close nexus between the error and the outcome and how that outcome frustrates the purposes served by the sentencing guideline scheme.”

“Well, then your answer to Justice Ginsburg should be yes, and it should have been yes at the outset,” broke in Justice Anthony Kennedy, author of the majority opinion in Molina-Martinez. “We have separate rules for sentencing, in part because the costs of remand are much less than the cost of a new trial, there can be some complexities, and it seems to me that you just have to confront the consequences of that choice to say that, in the sentencing case, an ordinary error is very close to plain error, but you seem to resist that.”

This time, Davidson was not about to turn down the help. “No, I — let me clarify. I think that’s the correct formulation, Justice Kennedy.” Before she could get any farther, Justice Samuel Alito, a dissenter in Molina-Martinez, protested, “Here, we’re not even talking about something that’s mandatory,” an allusion to the Supreme Court’s decision in United States v. Booker holding that the federal Sentencing Guidelines are unconstitutional unless treated as advisory. “These guidelines exist in some kind of middle universe that I don’t understand, but that’s another — that’s another question. Why is this different?”

“Well, the analysis doesn’t turn on whether … the sentencing guidelines are mandatory versus advisory,” Davidson responded. “It’s looking at how close of a nexus exists between the error and how it affects the outcome.” Having answered Alito’s question, if in a conclusory manner, Davidson was now ready to roll into the heart of her rationale, broad enough to encompass the categorical approach being urged by Ginsburg and Kennedy.

“Because the sentencing guidelines are the starting point for every sentence and are … what a sentence becomes anchored to, we have empirical data which reflects their anchoring effect, that when there’s an erroneously high guideline range, there’s a serious risk that … a defendant’s sentence was also higher than it would have been had the district court not been improperly influenced by it,” Davidson said. Implicit in her argument was that the courts might assess this “serious risk” on a case-by-case basis, but they might instead treat the risk as so serious that it should trigger automatic remand to the district court for correction, much in the way that certain trial errors are regarded as “structural” and hence exempt from harmless-error analysis.

Later, during the argument of Assistant to the Solicitor General Jonathan Ellis, the notion of empirical data came up again. But it was not Ellis who raised it. Justice Stephen Breyer — one of the principal architects of the Sentencing Guidelines system when he worked for Senator Edward Kennedy — did.

The guidelines system was designed such that the United States Sentencing Commission and the courts “cooperate in gathering statistical information so that the Commission can see how that’s working,” Breyer explained. “Now, as soon as you have people who depart for incorrect reasons from what they’re supposed to [do], you muck up that statistical information. … [I]f you will read, as I hope you would do someday, the introduction to the initial version of the guidelines, which happens still to be there, you will see that the purpose of the Commission is first to create a set of guidelines and then, through the procedures I’m talking about, to see what judges actually do in administering the guidelines so that those can be improved and changed over time.”

Breyer’s remark could potentially form part of the basis for a majority decision to categorically remand all forfeited guidelines miscalculations for resentencing. In response to the government’s strong insistence that such errors should generally be allowed to stand because the defendant is ultimately at fault for failing to object in a timely manner, Breyer may be suggesting that ignoring such errors harms not just the defendant, but the entire system. When a defendant is sentenced to 78 months in a 70-87-month range, as in this case, it creates different data than when a defendant is sentenced to 78 months in a 77-96 month range. The integrity of the data set, which the commission then relies on for future adjustments, is compromised.

Not all the justices who intend to vote for Rosales-Mireles may be prepared to treat forfeited guidelines miscalculations as automatically triggering a full remand. In his Molina-Martinez majority opinion, Kennedy mentioned a case in which a panel of the U.S. Court of Appeals for the 7th Circuit had issued a “limited remand” simply to ascertain whether the district judge would have given a difference sentence under the correct range. On Wednesday morning, Kennedy asked Ellis whether there are some courts of appeals “that just write the district judge a letter and say would it make a difference?”

“You can ask that,” Ellis replied. “But if the judge says yes, there’s still the fourth prong and there’s still the full resentencing that follows. Post Molina-Martinez, we haven’t found any examples of courts utilizing that for a guidelines range error.” It would not be surprising for the majority opinion in Rosales-Mireles’ case to adopt such an approach explicitly.