Latest Supreme Court Opinions - Fourth Amendment, Plain Error, Padilla, Double Jeopardy

Florida v. Harris, No. 11-817 (Kagan, unanimous)
The government does not need to produce a dog’s training and certification records, in addition to other evidence, to demonstrate a drug detection dog’s reliability. Instead, "[t]he question—similar to every inquiry into probable cause—is whether all the facts surrounding a dog’s alert, viewed through the lens of common sense, would make a reasonably prudent person think that a search would reveal contraband or evidence of a crime. A sniff is up to snuff when it meets that test." In so holding, the Court reversed the Florida Supreme Court’s decision.
[The other dog sniff case, Florida v. Jardines (No. 11-564), which presents the question "whether a dog sniff at the front door of a suspected grow house by a trained narcotics detection dog is a Fourth Amendment search requiring probable cause," has not yet been decided.]

Bailey v. United States, No. 11-770 (Kennedy, Roberts, Scalia, Ginsburg, Sotomayor, and Kagan, majority; Breyer, Thomas, and Alito, dissent)The detention of an individual who had left the premises before the search began and was a mile away before police detained him was not lawful under Michigan v. Summers, 452 U.S. 692 (1981), which authorizes law enforcement officers "to detain the occupants of the premises while a proper search is conducted" without need for any level of suspicion. The Court added that "[i]n closer cases courts can consider a number of factors to determine whether an occupant was detained within the immediate vicinity of the premises to be searched, including the lawful limits of the premises, whether the occupant was within the line of sight of his dwelling, the ease of reentry from the occupant’s location, and other relevant factors." Lawful detention under an alternative avenue, such as a Terry stop, is of course still possible.
Henderson v. United States, No. 11-9307 (Breyer, Roberts, Kennedy, Ginsburg, Sotomayor, and Kagan, majority; Scalia, Thomas, and Alito, dissent)
Under Federal Rule of Criminal Procedure 52(b), error is plain if it is clear at the time of appeal.
　

Chaidez v. United States, No. 11-820 (Kagan, Roberts, Scalia, Kennedy, Breyer, and Alito, majority; Thomas, concurrence; Sotomayor and Ginsburg, dissent)
Padilla v. Kentucky, 559 U. S. ___ (2010), does not apply retroactively to cases already final on direct review. In Padilla, the Supreme Court held that the Sixth Amendment requires an attorney for a criminal defendant to provide advice about the risk of deportation arising from a guilty plea. The Court concludes in Chaidez that, "under the principles set out in Teague v. Lane, 489 U. S. 288 (1989), Padilla does not have retroactive effect."

Evans v. Michigan, No. 11-1327 (Sotomayor, Roberts, Scalia, Kennedy, Thomas, Ginsburg, Breyer, and Kagan, majority; Alito, dissent)
Retrial following a court-decreed acquittal is barred, even if the acquittal was based on an incorrect legal conclusion, as was the case here.

Where Circuit Had Not Yet Addressed Question at Time of Trial, Error Is Neither "Clear" Under Rule 35(a) Nor "Plain" Under Rule 52(b)

United States v. Henderson, No. 10-30571 (5th Cir. July 8, 2011) (Smith, Southwick, Graves)

Another reason, as if you needed one, to preserve, preserve, preserve error in the district court.

Henderson, who pleaded guilty to being a felon in possession of a firearm, received an upward departure "to ensure that [he] had an opportunity to enroll in the federal Bureau of Prisons drug treatment program[.]" He did not object to his sentence at that time. Eight days later he filed a Rule 35(a) motion to correct his sentence, arguing that the district court violated 18 U.S.C. § 3582(a)'s prohibition on using imprisonment as a means of promoting rehabilitation—the Tapia issue.

The court first addressed whether Henderson's Rule 35(a) motion preserved the error. As relevant here, Rule 35(a) only allows a court to correct an error that is clear. This error wasn't, because at the time a circuit split existed over the Tapia question and the Fifth Circuit hadn't addressed it:

Before Tapia, there was a circuit split on whether a district court can consider a defendant’s rehabilitative needs to lengthen a sentence. Tapia, 2011 WL 2369395, at *3 n.1. Moreover, we have not pronounced on the question. In that situation, when there is no binding precedent on a question on which there is a circuit split, an alleged error is not “clear.” If we had confronted the question, we might have gone either way, so the error would not “almost certainly result in a remand of the case.” The error was not correctable under rule 35(a), and Henderson’s motion failed to preserve the error. We must therefore review for plain error.

If, like me, you thought the Fives had in fact addressed the Tapia issue and held that rehabilitation is a permissible factor in deciding on a term of imprisonment, here's what Henderson says:

In United States v. Giddings, 37 F.3d 1091 (5th Cir. 1994), we held only that a court could consider a defendant’s rehabilitative needs when sentencing him to imprisonment upon revocation of supervised release. Our decision in United States v. Lara-Velasquez, 919 F.2d 946, 953-57 (5th Cir. 1990), held only that the court can consider rehabilitative potential as a mitigating factor within an appropriate range of punishment, but not necessarily as a reason for a sentencing enhancement.

Okay, let's move on to plain error review. This one should be easy, right? There was error, and in light of Tapia the error was plain at the plain at the time of appellate consideration.
 
Wrong:

Tapia established that it is error for a court to “impose or lengthen a prison sentence to enable an offender to complete a treatment program or otherwise to promote rehabilitation.” Tapia, 2011 WL 2369395, at *9. Henderson cannot show that the error in his case was plain, however, because an error is plain only if it “was clear under current law at the time of trial.” United States v. Jackson, 549 F.3d 963, 977 (5th Cir. 2008) (emphasis added). At the time of trial, the Supreme Court had not yet decided Tapia and, as we have just explained, we had not yet addressed the question. Where we have not previously addressed a question, any error cannot be plain.

(second emphasis added). That strikes me as an incorrect approach to the plainness prong, but the court cites authority for that proposition: "See United States v. Vega, 332 F.3d 849, 852 n.3 (5th Cir. 2003) (“We conclude that any error by the district court in this regard was not plain or obvious, as we have not previously addressed this issue.” (citing United States v. Calverley, 37 F.3d 160, 162-63 (5th Cir. 1994)(en banc)))."
 
Whether this approach is consistent with the Supreme Court's pronouncments on the plainness prong, or even Fifth Circuit precedent, perhaps the more important takeaway is: PRESERVE!

An Important Lesson On Error Preservation, An Open Miranda Question, and Knowledge of Interstate Movement in FIP Cases

United States v. Potts, No. 10-10257 (5th Cir. June 15, 2011) (Smith, DeMoss, Owen)

A general rule of preserving error at trial (which I'm shamelessly stealing from someone else) is to keep asking for things until the court says no. Failure to do so, as Potts illustrates, lands you in the dark realm of plain error review.  Not a good place to be when, as in Potts, the question of whether there was error at all concerns an issue that remains open in the Fifth Circuit. Oh, and one more thing: to be guilty of being a felon in possession of a firearm, need a defendant know that the firearm had moved in interstate commerce? Important issues all.

So what happened here? Potts was stopped by Officer James (for reasons that aren't at all clear to me from the opinion). Thence,

James instructed Potts to park the car and shut off the engine. Potts began to reach under his seat, at which point James ordered Potts to show his hands. Potts did not immediately comply with that command and instead continued to reach under the seat. James approached the vehicle and ordered Potts to exit the car. Potts complied, and James was able to see that a firearm was protruding from under Potts’s seat.

James examined the firearm, handcuffed Potts, and sat him down on the street curb. James then asked Potts whether the gun belonged to him, but Potts did not respond. A search was conducted of the car, which yielded two additional firearms and some ammunition. Potts was then arrested.

(emphasis added). "At trial, the prosecution elicited testimony from James regarding Potts’s silence after being asked who owned the pistol." Potts objected to the testimony as a Fifth Amendment violation. The court did not rule on the objection, but suggested that it could instruct the jury that Potts had no obligation to answer the officer's question and that it's not against the law to do so. Potts agreed to the instruction, which the court then gave. "Potts did not reassert his objection to the testimony, object to the instruction, or move for a mistrial." When the prosecutor brought up Potts' silence again in closing argument, "Potts objected, not on any Fifth Amendment ground but on the ground that the prosecution was attempting to shift the burden of proof." The jury found Potts guilty.
 
On appeal, Potts argued first "that the government violated his Fifth Amendment right against self-incrimination by referencing, during trial, his silence in response to police questioning[.]" The court reviewed for plain error, because an objection alone was not sufficient to preserve the issue.

In Salinas, 480 F.3d at 755, we held that plain-error review was appropriate for a Fifth Amendment claim even though defense counsel “timely objected to each of the prosecutor’s references to [the defendant’s] post-arrest silence,” because “the trial court sustained all of those objections, and the trial court’s instructions to the jury made it clear that the jury was not to consider any of the challenged remarks.” Furthermore, we remarked that the defense “never took exception to the district court’s handling of his objections and, significantly, . . . never requested that the district court declare a mistrial.” Id. at 755-56. Plain error review was appropriate, because the defendant “effectively received all of the relief that he requested from the district court.” Id. at 756.

Salinas is not directly controlling, because the court did not explicitly sustain Potts’s objections, but the principles of Salinas inform us. As with the defendant in Salinas, Potts never raised any concerns with how the district court chose to handle his objection, and Potts explicitly agreed to the court’s proffered curative statement.

(emphasis added). Potts argued that the district court had implicitly overruled his objection, but the court of appeals didn't buy it:

Potts objected to testimony regarding his silence. The district court then offered a curative instruction, which Potts accepted. By accepting the instruction, Potts failed to obtain a definitive ruling on his objection—there was no implicit overruling, but rather no ruling at all. Following that failure to obtain a ruling, Potts accepted the court’s curative instruction without objection, thus failing to preserve error.

And that was a problem. Potts couldn't satisfy the second prong of plain-error review—that any error be clear or obvious— because the Fifth Circuit "has yet to address conclusively whether the use of pre-Miranda silence as substantive evidence of guilt is a Fifth Amendment violation."
 
On to Potts' next argument, the travelling gun one:

Potts contends that the government was required to prove, but did not, that he knew the firearm had traveled in interstate commerce. Potts was convicted pursuant to 18 U.S.C. § 922(g)(1)and sentenced pursuant to 18 U.S.C. § 924(a)(2). He argues that the word “knowingly” in § 924(a)(2) requires the government to prove that every element in § 922(g)(1), including that the firearm had traveled in interstate commerce, was known by the defendant.

That argument, however, is foreclosed by United States v. Rose, 587 F.3d 695 (5th Cir. 2009). There, the defendant raised the same argument as Potts does here, but we rejected it because the defendant had not been sentenced under § 924(a)(2). Id. at 705-06. Rather, he was sentenced under § 924(e)(1), which does not contain a “knowingly” requirement. We went on, however, to state that “[e]ven assuming arguendo that the ‘knowingly’ requirement in § 924(a)(2) applied throughout that section, there would be no corresponding impact on the elements of a crime listed in § 922(g)(1).” Id. at 706 n.9. That statement in Rose was not mere dictum; rather, it was an alternate holding that carries the force of precedent. Thus, Potts’s claim fails.

TX Injury to Child by Act Not 2L1.2 16-Level COV; No Invited Error Where Counsel Misunderstood Relevant Precedent

United States v. Andino-Ortega, No. 09-40498 (5th Cir. June 8, 2010) (Barksdale, Garza, Dennis)

Two important holdings in this appeal of an illegal reentry sentence. But first, the facts:

Andino had a prior Texas conviction for injury to a child (Tex. Penal Code § 22.04), for which the PSR hit him with a 16-level crime-of-violence enhancement. At sentencing, he asked for a downward variance on the basis that he thought he was only facing a 4-level any-other-felony enhancement for that offense. Defense counsel stated:

The predicate offense in this case was injury to a child. . . . There is case law that says that injury to a child by omission is not a crime of violence and would only garner a 4-level enhancement.

However, there is case law—[p]articularly, I’m looking here at Perez-Munoz v. Keisler, 507 F.3d 357 [5th Cir. 2007], which finds that if there—the injury to a child stems from an intentional act, then it definitely is a crime of violence. Your Honor, . . . my client’s position throughout has been that [the injury] was the result of an accident.

Now, the fact remains that he pled to the offense as charged in the indictment which charged an intentional offense. . . .

Andino got a 60-month within-Guidelines sentence, and challenged the 16-level enhancement on appeal.

Which brings us to the First Important Thing: The Government argued that defense counsel's statements at sentencing concerning the enhancement constituted waiver of the issue. In United States v. Fernandez-Cusco, the Fifth Circuit "concluded that [similar] statements did not rise to the level of waiver, but 'arguably' constituted invited error. In an abundance of caution, the court nonetheless reviewed for plain error." Which is what the court did here, too.

Although we recognize the factual similarity here—Andino-Ortega’s counsel acknowledged that the 16-level enhancement was proper—he did so on the basis of a misunderstanding of this court’s precedent. The statements regarding the propriety of the crime-of-violence enhancement do not constitute a waiver because they do not evidence an intentional and knowing relinquishment of a right. Counsel’s failure to object below because he did not recognize the argument now being made on appeal is not a waiver. Accordingly, we review for plain error.

(citations omitted).

And plain error it was, which is Important Thing 2d. Defense counsel was mistaken about Perez-Munoz because that case involved the COV definition in 18 U.S.C. § 16(b), which is different from the definition applicable to guideline §2L1.2's 16-level enhancement. Section 16(b) reaches felony offenses that involve a "substantial risk that physical force against the person or property of another may be used in the course of committing the offense." Guideline §2L1.2's COV definition, by contrast, asks whether force is an element of the offense. Not the same question, so Perez-Munoz isn't controlling. (Yes, the §2L1.2 definition also lists specific offenses that are crimes of violence regardless of whether they have a force element, but injury to a child isn't one of them.)

Also not controlling is United States v. Gracia-Cantu, which held that Texas injury to a child by omission is not a COV under 18 U.S.C. § 16(a). Although § 16(a) is identical to the force clause of §2L1.2's COV definition, Andino was convicted of injury by act, not omission.

So what is controlling? United States v. Vargas-Duran and United States v. Calderon-Pena. They held that an offense must necessarily require physical force to fall within the force clause, and that if any set of facts would support a conviction without proof of force, then force isn't an element of the offense. "For instance, an offense under [the Texas injury-to-a-child statute] can be committed by intentional act without the use of physical force by putting poison or another harmful substance in a child’s food or drink." So it's not a 16-level COV.

Defendant's Admission That PSR Was Correct Costs Him On Plain Error Review of Otherwise Erroneous Sentence Enhancement

United States v. Velasquez-Torres, No. 09-40646 (5th Cir. June 18, 2010) (per curiam) (Jolly, Smith, Owen)

Here we have another object lesson in the perils of admitting—explicitly or implicitly—the accuracy of the PSR.

Recall that, under the illegal reentry statute, 8 U.S.C. § 1326, enhanced penalties apply if an alien was deported after having been convicted of a qualifying offense. Which brings us to Mr. Velasquez. He'd been deported twice before, in 2004 and 2008. He had also been convicted of felony domestic assault in-between the two deportations. At his rearraignment on the illegal reentry charge, he agreed with the prosecutor's statement of the factual basis for his guilty plea, which identified only the earlier of the two deportations. The PSR of course listed both of them, and hit Velasquez with an 8-level bump for the assault conviction. That also elevated the statutory maximum from the 2 years allowed under § 1326(a) to the 20 years authorized under § 1326(b)(2).

At his sentencing, Velasquez-Torrez affirmed that he had received and read the PSR. Through counsel, Velasquez-Torrez stated that there were no mistakes in the report, although he “remained silent on Paragraph 28,” which contained a narrative discussing the facts of his prior assault conviction.

Velasquez got 37 months, and appealed. He argued that his sentence could not exceed 2 years because the 2008 deportation was not proved beyond a reasonable doubt, just like in Rojas-Luna, where the defendant prevailed on this issue. And also like Rojas-Luna, Velasquez faced plain-error review. But unlike Rojas-Luna, Velasquez could not show plain error. Why the difference?

In Rojas-Luna, the defendant pleaded guilty to reentering the United States illegally. The factual basis for the charge was that Rojas-Luna had been deported in 1988 and had reentered the United States in 2006. No mention was made of a prior conviction. The subsequent PSR, however, noted that Rojas-Luna was convicted of aggravated assault in 2003 and was removed in 2006. Because Rojas-Luna had been convicted and subsequently removed, the district court enhanced his sentence under § 1326(b)(2). On appeal, we held that the district court’s reliance on the 2006 removal, when this fact had not been proven to a jury or admitted by Rojas-Luna, was plain error.

In Ramirez, we dealt with facts that differed slightly, but materially, and came to a different conclusion. Ramirez pleaded guilty to illegal reentry at his rearraignment. The indictment and factual basis for his plea did not specify the date of his removal. After his plea, the PSR reflected that Ramirez had been convicted of aggravated assault in 2003 and deported in 2005. Relying on the conviction and deportation mentioned in the PSR, the district court increased Ramirez’s sentence beyond the statutory maximum contained in § 1326(a). On appeal, Ramirez argued that the district court’s reliance on the 2005 removal was plain error under Rojas-Luna. We disagreed. We noted that, whereas in Rojas-Luna the defendant had not admitted to the relevant deportation, Ramirez and his counsel “affirmed specifically that they had read the PSR, reviewed it for legal and factual accuracy, and made no objections to it.” While reliance on a PSR alone to establish the fact of a defendant’s removal is improper, “reliance on a defendant’s admission of facts that are contained in the PSR is permissible.” Because Ramirez implicitly admitted the accuracy of the PSR, the district court did not plainly err in relying on the facts contained therein.

In this case, the district court relied on Velasquez-Torrez’s 2008 deportation to increase his sentence beyond the statutory maximum. While Velasquez-Torrez’s 2008 deportation was not proven to a jury, it was described in the PSR. At his sentencing, Velasquez-Torrez stated that he had received and read the PSR. Beyond “remaining silent” on a section of the PSR discussing the facts of his prior assault conviction, Velasquez-Torrez through counsel affirmed that the PSR contained no mistakes. Just as in Ramirez, Velasquez-Torrez agreed to the accuracy of the PSR. Having admitted the fact of his deportation, Velasquez-Torrez cannot now argue that the district court improperly relied on that deportation. Therefore, the district court did not commit error, much less plain error, when it enhanced Velasquez-Torrez’s sentence pursuant to § 1326(b)(2).

Now there's actually a little more to Ramirez than the implicit admission of the PSR's accuracy. In Ramirez,

[T]he PSR incorporated ICE records provided to the probation office by the government and set forth the details of Ramirez's three prior removals in 1998, 2002, and 2005. The ICE records were provided to Ramirez by the government. In addition, Ramirez and his counsel affirmed specifically that they had read the PSR, reviewed it for legal and factual accuracy, and made no objections to it.

Distinguishing that scenario from Rojas-Luna, Ramirez said,

In Rojas-Luna, this court reached the conclusion that there was plain error because “there is no evidence in the record that Rojas-Luna ever agreed to the accuracy of the PSR.” In addition the panel noted that the only evidence of the defendant’s prior removal was the unsupported statement in the PSR that he was removed in 2006. In this case, in contrast, there is evidence that Ramirez agreed to the accuracy of the PSR and the facts of Ramirez’s prior removals in the PSR were based on ICE documentation of Ramirez’s prior deportations, which documents were provided to the defendant.

So what to take away from all this? Be very careful about conceding the accuracy of the PSR, even implicitly. Because even if Ramirez rested on more than an implicit admission—and putting aside the question whether Ramirez's and his counsel's statements regarding the PSR really amounted to an admission, implicit or otherwise—the court appears to read the case as if the admission alone was enough to defeat plain error review.

No Presumption of Vindictiveness When Defendant Receives Higher Sentence On Remand From Different Judge Of the Same Court

United States v. Rodriguez, No. 09-20181 (5th Cir. Mar. 25, 2010) (King, Barksdale, Elrod)

From the be-careful-what-you-wish-for files: Rodriguez entered into plea agreement in exchange for 1) dismissal of some counts, and 2) the Government's promise not to seek offense level enhancements at sentencing. The Government then breached the agreement by defending the PSR's calculation of the Guidelines range, which included some of those enhancements. The district court agreed with only some of the PSR's recommendations, and imposed a low-end sentence of 37 months. Rodriguez appealed his conviction and sentence, arguing that the Government breached the plea agreement. The Fifth Circuit affirmed the conviction, but vacated the sentence and remanded for resentencing before a different judge. Judge 2 rejected Rodriguez's objections to the PSR, and imposed a low-end sentence of 47 months—10 months greater than Rodriguez received on the first go-round. Rodriguez appealed again, arguing that his 47-month sentence was either presumptively or actually vindictive.

First, a threshold issue: is Rodriguez's claim reviewable? The Government argued that the following statement by Rodriguez's counsel—after the sentence was imposed—amounted to either waiver or invited error. Responding to a question from the court as to whether he had "any argument on the mandate," counsel said:

Judge, he was ordered a re-sentencing. And from the looks of it, you have done, you have went through that file extensively. We reurged [prior counsel’s] objections, which were extensive; and you have just summed up exactly what you just did.

Said the court of appeals:

[T]hat reply, made after imposition of sentence, does not show counsel knew of the potential vindictiveness issue and chose intentionally to relinquish it (waiver). And, certainly, sentence having already been imposed, the reply did not induce the claimed error (invited error). Accordingly, the claimed error is reviewable[, albeit only for plain error].

One to keep in mind if you find yourself in the unfortunate position of having to refute a waiver/invited error argument by the Government.

On to the meat of the matter: "[North Carolina v.] Pearce held a presumption of vindictiveness arises when, after a new trial, a judge sentences a defendant to a harsher sentence. [cites omitted] At issue here is whether that presumption applies when a different judge of the same court imposes the new sentence."

Thus begins a lengthy exegesis of Fifth Circuit and Supreme Court case law on the subject. Long story short, the Fifth Circuit held that the presumption applies in these circumstances in United States v. Floyd, a case decided several years after Pearce. Floyd reached that conclusion, in part, because the resentencing-before-different-judge-on-remand scenario was what had happened in one of the two companion cases that comprised Pearce. But the Supreme Court has had a lot more to say about the presumption since Pearce, both before and after Floyd. In particular, there was Texas v. McCullough.

McCullough addressed whether the Pearce presumption applied where: a jury imposed the initial sentence; because of prosecutorial misconduct, however, the trial judge granted a new trial; upon retrial before the same judge, defendant asked that the judge, rather than the jury, impose sentence; and that judge imposed a greater sentence than had the jury for the first trial. The judge, on defendant’s motion, made findings of fact on why her sentence was greater.

(cites omitted). McCullough held that the presumption of vindictiveness does not apply in that situation. It acknowledged that Pearce itself involved a case with a different sentencer on remand, but noted that Pearce hadn't actually addressed this particular issue and—to the extent Pearce could be read otherwise—effectively overruled Pearce on this point. Therefore,

McCullough expressly rejected part of the basis for Floyd’s reasoning—there having been different sentencing judges in Pearce. Moreover, McCullough held having different sentencers was one of the primary reasons that the presumption was unnecessary. Restated, the possible same-court institutional and collegial pressures that were of concern to Floyd were held in McCullough not to require the presumption. Accordingly, we hold, as Rodriguez’ counsel conceded at oral argument, that Floyd was overruled by McCullough.

(You might be wondering why the court spent 11 pages analyzing this issue, given Rodriguez's concession. For starters, "[a]t oral argument, Rodriguez’ counsel conceded McCullough controls, but urged our criticizing McCullough as having strayed from the principles announced in Pearce." The court of course declined the invitation, but had to decide the issue for itself because it had never before considered whether Floyd survived McCullough.)

According to the court, seven other circuits "do not apply the presumption when different judges preside over the first and second sentencing." But some of them "hold the presumption is inapplicable in this different-sentencers situation only if the second states objective, non-vindictive reasons for imposing the greater sentence (added condition)." Although opining that "the evolved reasonable-likelihood-of-vindictiveness standard negates this added condition[,]" the court notes that it's good practice.

Neither party, however, has raised this added-condition issue. [cite]. In any event, as discussed infra, because the second judge stated more than sufficient reasons for the greater sentence, we do not need to reach the issue. In other words, even assuming the added condition is required, the record amply demonstrates that the second judge more than fulfilled that condition. Along that line, regardless of whether the added condition is required, the second judge should obviously state such reasons. They assist not only in our reviewing a vindictiveness challenge but also in reviewing other challenges to the sentence as well.

That's not the end of the matter, though. The court still had to "consider whether there was a reasonable likelihood that the increase in Rodriguez’ sentence was due to actual vindictiveness." And there wasn't:

Rodriguez contends the record shows the second judge acted vindictively because: he summarily overruled Rodriguez’ objections to the PSR; and, he did not allow the Government to contribute to the sentencing hearing. Neither matter even remotely suggests vindictiveness.

The record reflects that the second judge carefully considered Rodriguez’ objections. Moreover, Rodriguez has offered no authority requiring the second judge to state his reasons for denying them. [cite].

Rodriguez does, apparently, recognize the patent absurdity of his contention that the second judge’s stated intention not to engage the Government at resentencing somehow evinces vindictiveness. He admits in his reply brief that his contention creates a Catch-22 for the district court: according to Rodriguez, the second judge is vindictive for not having engaged the Government; but, had the judge done so, he would have caused the Government, once again, to breach its plea agreement. The record shows, as discussed infra, that the second judge was entitled to rely on other bases, negating any need to solicit input from the Government.

Again, there is simply no evidence of actual vindictiveness. Therefore, there was no error. In sum, our plain-error review for vindictiveness—presumed or actual—ends.

And so does this post.

Panel Breaks With Prior Panel Precedent; Holds Ex Post Facto Clause Does Not Apply to Guidelines Increases Post-Booker

United States v. Castillo-Estevez, No. 09-40096 (5th Cir. Feb. 9, 2010; rev'd Mar. 10, 2010) (Jones, Smith, Elrod)

See update below.

Here we have an instance of dueling panels. Castillo-Estevez holds that any error in the district court's application of the Sentencing Guidelines was not plain, despite the fact that there is controlling circuit precedent on the precise question at issue, and under that precedent the district court plainly erred.

The scenario: Castillo pleaded guilty to illegal reentry. He had a prior New York conviction for criminal sale of a controlled substance; that offense can be committed simply by an offer to sell; it does not require an actual sale. Under the 2007 Sentencing Guidelines in effect at the time Castillo committed the illegal reentry, an offer to sell a controlled substance was not a "drug trafficking offense" for purposes of the 12- or 16-level enhancements under §2L1.2. (In fact, it probably wasn't even an "aggravated felony" for purposes of the 8-level enhancement.) But under the 2008 Guidelines in effect at the time of Castillo's sentencing, an offer to sell was a §2L1.2 DTO (due to an expanded DTO definition). The district court applied the 2008 Guidelines, resulting in a higher range than would have applied under the 2007 Guidelines in effect at the time of Castillo's offense.

On appeal, Castillo naturally argued that this was an ex post facto violation, because he received a more severe sentence due to a guideline change that took effect after he had completed his offense. Unfortunately, Castillo raised this argument for the first time on appeal, resulting in plain error review.

Here's what the court said:

Castillo’s argument overlooks the Supreme Court’s subsequent decision in United States v. Booker, 543 U.S. 220, 125 S. Ct. 738 (2005), which rendered the sentencing guidelines merely advisory. In United States v. Rodarte-Vasquez, 488 F.3d 316 (5th Cir. 2007) (Jones, C.J., concurring), it was observed that the now-advisory guidelines should not raise ex post facto concerns because “the sentence imposed by the court need not be harsher under later guidelines than it would have been under the guidelines in effect when the offense was committed.” [Two circuits have agreed with that observation, two others have disagreed.]

We need not determine here whether ex post facto claims arising from the application of evolving sentencing guidelines are viable after Booker. Even if the district court’s application of the 2008 guidelines violated the ex post facto clause, the error would certainly not be “plain” in light of such post-Booker cases as Rodarte-Vasquez, Demaree, and Barton. To be “plain,” legal error must be “clear or obvious, rather than subject to reasonable dispute.” Puckett v. United States, 129 S. Ct. 1423, 1429 (2009). See also United States v. Peltier, 505 F.3d 389, 391 n.3 (5th Cir. 2007) (“Plain error must be ‘error so obvious that our failure to notice it would seriously affect the fairness, integrity, or public reputation of [the] judicial proceedings and result in a miscarriage of justice.’”). Because the caselaw reveals a “reasonable dispute” regarding the ex post facto implications of retroactive application of the advisory guidelines, the district court’s error, if any, was not plain.

But the Fifth Circuit has already determined that ex post facto claims arising from the application of evolving sentencing guidelines are viable after Booker. In United States v. Austin, 479 F.3d 363—a pre-Rodarte-Vasquez decision reviewing a sentence imposed post-Booker—the court began its analysis by noting that,

Post-Booker, we continue to review the district court's interpretation of the Sentencing Guidelines de novo and its fact findings for clear error, although the ultimate sentence is reviewed for unreasonableness. When calculating a Guidelines sentencing range, a district court applies the Guidelines in effect at sentencing, unless the Guidelines in effect when the offense occurred would yield a lesser penalty. In such a case, to avoid ex post facto concerns, the court uses the Guidelines yielding the lesser penalty.

Later, applying that rule, Austin stated that "the offense level under the 2004 Guidelines is higher than under the 1998 Guidelines, and the 1998 Guidelines apply." The court went on to hold that the district court had incorrectly calculated the range under the older Guidelines, and even vacated the sentence on plain error review.

Thus, in light of Austin, the district court's application of more severe post-offense 2008 Guidelines in Castillo's case was not only error, but clear and obvious error. (It might very well meet the other requirements for reversal under plain error review, too.) And that should hold true even though there's a circuit split on the ex post facto question.

The opinion in Castillo-Estevez does not discuss or attempt to distinguish Austin, so there is no way to know for sure why the court did not find it controlling. (It also is not clear whether Castillo brought up Austin. The opinion says only that he relied on United States v. Kimler, 167 F.3d 889 (5th Cir. 1999), a pre-Booker decision.) In any event, because one panel may not overrule another (absent something like a statutory change or intervening Supreme Court case law), Austin should control over Castillo-Estevez.

3/10/2010 UPDATE: In an apparent response to a petition for rehearing en banc, the panel has issued a revised opinion that includes a footnote addressing Austin:

Our post-Booker decisions in United States v. Reasor, 418 F.3d 466 (5th Cir. 2005), and United States v. Austin, 479 F.3d 363 (5th Cir. 2007), do not alter our conclusion that the district court committed no plain error. Reasor is distinguishable. See Rodarte-Vasquez, 488 F.3d at 325 n.2 (Jones, C.J., concurring). Moreover, the court’s acknowledgment in Austin that retroactive guidelines application might pose ex post facto problems was dicta; ex post facto claims were not at issue on appeal because the lower court had applied the guidelines version yielding a lesser penalty. 479 F.3d at 366–67.

To be fair, Austin is a difficult read (unless you delight in reading discussions of loss-amount calculations in Medicare fraud cases). But it's hard to see how Austin's ex post facto analysis could be dicta. As near as I can tell, Austin argued that the district court should have applied the 2004 Guidelines in effect at the time of sentencing, rather than the 1998 Guidelines in effect at the time of his offense, because the loss calculation would have been lower under the later version. He also argued that the district court erred in not crediting certain funds against the loss amount (presumably under the 1998 Guidelines, since it appears that he arrived at a loss of $0 under his interpretation of the 2004 Guidelines even without this credit). The court of appeals:

1. acknowledged that when the Guidelines range is higher under the version in effect at the time of sentencing than under the version in effect at the time of the offense, the court must apply the earlier version "to avoid ex post facto concerns,"
2. rejected Austin's proposed interpretation of the applicable 2004 provisions (which would have resulted in a loss amount of $0),
3. applied the ex-post-facto clause to conclude that because "the offense level under the 2004 Guidelines is higher than under the 1998 Guidelines, . . . the 1998 Guidelines apply[,]" and
4. ultimately held that the district court incorrectly calculated the loss under the 1998 Guidelines.

Because the court necessarily had to confront the ex post facto issue to resolve the case, Austin's discussion of and application of the rule was not dicta and Austin should be controlling on the question unless and until the en banc Fifth Circuit or Supreme Court conclude otherwise. (Which may very well happen at some point, as there is a circuit split on the question.)

Court Must Accept or Reject Entire (c)(1)(C) Agreement, May Not Accept Parts and Reject Others; Probably Goes for (c)(1)(A) Agreements, Too

United States v. Self, No. 08-40624 (5th Cir. Feb. 3, 2010) (Garza, DeMoss, Clement)

Self finds the court confronting an issue of first impression in the Fifth Circuit: whether a court can accept or reject a Rule 11(c)(1)(C) plea agreement on a "piecemeal basis." Answer: no. And the reasoning appears to be equally applicable to (c)(1)(A) agreements.

Self was charged with two counts of bank robbery, and two 924(c) carrying counts. He hammered out a (c)(1)(C) agreement with the Government, which entailed a guilty plea to one bank robbery and one 924(c), dismissal of the remaining counts, and a sentence of 171 months (87 months for the bank robbery + 84 months consecutive on the gun count). Additionally, "Self waived the right to appeal 'on all grounds,' but reserved the right to appeal the failure of the district court to impose a sentence in accordance with the terms of the agreement." A magistrate judge took Self's guilty plea, and recommended that the district court accept the agreement. So far, so good.

But then—stop me if you've heard this one before—the probation officer threw a spanner in the works: "The PSR determined that Self was a career offender and recommended a career offender enhancement on [the bank robbery count] that raised the advisory guidelines range to 188 to 235 months. When the mandatory minimum of 84 months for [the gun count] was added, the resulting guidelines range was 272 to 319 months’ imprisonment." At sentencing:

The district court informed Self of his right to withdraw his guilty plea and that if he did not withdraw his plea, he might receive a sentence less favorable than that agreed to in the plea agreement. Self declined to withdraw his guilty plea. The district court reiterated that it would accept all of the plea agreement’s terms except the recommended sentence as to [the bank robbery count]. Self again stated that he did not wish to withdraw his plea. Self allocuted and his defense attorney requested the minimum sentence. The district court then sentenced Self to the minimum sentence of 188 months for [the bank robbery count], taking into account the career offender enhancement, and 84 months for [the gun count], resulting in a sentence of 272 months.

Self appealed his sentence. The Government raised the appeal waiver. The court said:

Because we conclude, infra, that the district court rejected the plea agreement in toto, Self’s waiver of rights in that agreement does not bar his appeal. [cites]. Even assuming Self’s appeal waiver was enforceable, its terms do not apply here. The plea agreement reserved Self’s right to appeal “the failure of the Court, after accepting the agreement, to impose a sentence in accordance with the terms of this agreement.” The 272 month sentence imposed by the district court exceeded the 171 month sentence stipulated to in Self’s plea agreement; the sentence was not “in accordance” with the plea agreement’s terms. There is no obstacle to Self’s appeal.

On to the substance: "Self argue[d] that he is entitled to a sentence reduction because the district court accepted the plea agreement but did not comply with its terms." But because "Self did not raise any objection to the proceedings at the sentencing hearing and he did not assert this argument as a basis for a sentence reduction in either of his post-hearing motions[,]" the court reviewed for plain error.

Before reaching the error prong, the court first had to decide whether the district court had accepted the plea agreement but failed to comply with its sentencing terms (as Self argued), or whether it had rejected the plea agreement (the Government's position). Examining the district court's statements at the sentencing hearing against the background of Rule 11, the court of appeals concluded that the district court rejected the entire plea agreement. The district court, after finding that the probation officer's calculation was correct, said "[I] advise the defendant that I will not follow, cannot follow . . . that portion of the plea agreement, which would have set the sentence . . . [at] 70 to 87 months. . . . I must inform the parties of my rejection of that part of the plea agreement." The court then advised Self, consistent with Rule 11(c)(5), of his right to withdraw his guilty plea, saying "if you don’t withdraw . . . it would be that part of the plea agreement that would be changed, however, the rest of the plea agreement will remain the same." Self opted not to withdraw his plea, and was sentenced to 272 months' imprisonment, rather than the 171-month sentence in the plea agreement. Said the court:

Although we have found no case in our Circuit that explicitly addresses whether a plea agreement may be accepted or rejected on a piecemeal basis, based on the language of Rule 11, we conclude that it cannot. See FED. R. CRIM. P. 11(c)(3)(A) (finding that “the court may accept the agreement, reject it, or defer a decision until the court has reviewed the presentence report” (emphasis added)); [a couple of cases]. Rule 11 speaks in terms of a “plea agreement” and “does not distinguish between ‘sentence bargains’” and “so-called ‘charge bargains,’ in which a criminal defendant typically pleads guilty to a specific charge in exchange for the prosecution agreeing to drop other charges.” At least one other court has concluded that the “rejection of a stipulated sentence constitutes rejection of the entire plea agreement, thereby triggering the mechanisms in . . . Rule 11(c)(5).” We agree. By rejecting the agreed-to sentence of 87 months on Count III, the district court constructively rejected the plea agreement in toto. Further indicia that the district court rejected the plea agreement is that it admonished Self as required by Rule 11(c)(5), which sets forth the warnings that must be delivered to a defendant when the court rejects a plea agreement.

(most cites omitted). Lets pause for a moment to talk about a couple of things before we move on. First, the court explained how to avoid this mess in the first place: "We note that the district court unnecessarily muddied the waters by making statements to the effect that it was rejecting 'that portion of the plea agreement' specifying an 87-month sentence. The better practice is to make clear that the entire plea agreement is being rejected and then give the Rule 11 warnings." Second, remember that Rule 11(c)(3)(A) covers not only (c)(1)(C) agreements, but (c)(1)(A)'s, as well. Plus, when you think about it, this case actually involves a combination (c)(1)(A)/(c)(1)(C) agreement. Thus, the court's reasoning should apply equally to (c)(1)(A)'s. How might this come up in the context of a (c)(1)(A)-only plea? I'm not entirely sure, but keep the case in mind if you have one those that starts going south.

Okay, back to plain error. Although the district court had discretion to reject the agreement, it "was not permitted to reject the plea agreement and then re-impose it on the parties with terms that it found acceptable."

In this respect, the colloquy between Self and the district court is problematic. Although the district court followed the dictates of Rule 11(c)(5) in advising Self that he could withdraw his plea and might be sentenced more harshly, the district court at the same time stated that the plea agreement would stand in all respects except for the sentence. Thus, it is unclear whether Self understood that if he withdrew his plea of guilty, he could either proceed to trial or try to negotiate a new plea agreement, or if he persisted in his guilty plea, he would be sentenced without any of the constraints or benefits of the plea agreement, including the government’s agreement to drop two of the counts. Instead, it appeared that the district court was offering the same terms as the plea agreement, except that the sentence would be changed.

The error affected Self's substantial rights. Instead of getting the 171-month sentence he bargained for, "Self was subjected to an altogether different bargain—one of the district court’s making. Had the district court rejected Self’s plea agreement in toto and sent the parties back to the drawing board, we cannot say what agreement they might have struck."

Which brings us to the last prong: "Failure to properly inform Self that the entire plea agreement was being rejected, compounded by the district court’s error in reimposing all of the terms of the plea agreement on Self, except for the agreed-to sentence under Rule 11(c)(1)(C), requires reversal because such error 'seriously affects the fairness, integrity or public reputation of judicial proceedings.'" Hence, vacation of Self's conviction and remand for proceedings before a different judge.

By the way, notice anything missing from the court's analysis? Perhaps a discussion of whether the error was "plain?" Don't worry, it's there. It just isn't identified as such. The court's discussion of the fairness, etc. prong notes that "[s]trict compliance with Rule 11 is generally required," and that, "[f]or the same reasons that the district court may not involve itself in plea negotiations, it may not reject a plea agreement and then, nonetheless, impose it on the parties with modifications that it chooses." Plainness is also implicit in the court's explanation of why there was error, so that part of plain error review is covered.

Certificates of Non-Existence of Record Are Testimonial for Confrontation Clause Purposes, But That's Not the Silver Bullet You Might Have Hoped For

United States v. Martinez-Rios, No. 08-40809 (5th Cir. Jan. 28, 2010) (per curiam) (Jones, Smith, Elrod)

Think way, way back, to the mists of last July, when the world—well, this blog, at least—was abuzz over whether the Supreme Court's Confrontation Clause decision in Melendez-Diaz v. Massachusetts portended a new dawn of heroic struggle against the ease with which the Government can prove the elements of illegal reentry. It seemed as clear as clear could be that Melendez-Diaz would bar the Government from relying on certificates of non-existence of record to establish the lack-of-permission-to-reapply-for-entry element of illegal reentry. Notwithstanding that clarity, we were technically flying blind until the Fifth Circuit could weigh in one way or the other. But no longer. The era of uncertainty has ended, and the result is mixed.

First, the good news: Martinez-Rios holds that CNRs are testimonial in light of Melendez-Diaz, and that the Supreme Court's decision "implicitly overruled" Rueda-Rivera, the Fifth Circuit's 2005 decision to the contrary. (The Government, to its credit, conceded the point.) The court appears to have read Melendez-Diaz's discussion of a clerk's CNR as dicta, but quite strong dicta that "spoke directly" to the issue with CNR's in illegal reentry cases. The court also pointed to Melendez-Diaz's "reli[ance] on a key distinction between records that are kept in the ordinary course of business and those that are specifically produced for use at trial: The latter are “testimonial” and are at the heart of statements triggering the Confrontation Clause. CNR’s are not routinely produced in the course of government business but instead are exclusively generated for use at trial. They are, therefore, testimonial." Finally, the court recognized that the lab reports at issue in Melendez-Diaz served the same purpose that CNR's serve in illegal reentry cases: to prove an element of the offense. Thus, "Rueda-Rivera is overruled."

But then there's the bad news: the Government may be able to prove the lack-of-consent element BRD even without a CNR. How? Let's start with a look at this case.

Martinez was arrested by a Border Patrol agent who caught him and others hitching a ride on a train in Gardendale, Texas (which is about halfway between Laredo and San Antonio, or—for you hunters out there—between Cotulla and Dilley). At the ensuing illegal reentry trial, the Government introduced Martinez's A-file—which contained a CNR—through a Border Patrol agent. The ICE official who prepared the CNR did not testify; "instead, the prosecutor introduced that evidence through [the Border Patrol agent], who explained how a CNR is processed." Martinez's "counsel objected to the introduction of the CNR on grounds that [the agent] had no personal knowledge of the case and was not the custodian of the records[,"] but did not mention the Confrontation Clause.

And you know what that means for the appeal: plain error review. The error prong is easy:

Under Melendez-Diaz, 129 S. Ct. at 2539, Martinez-Rios’s Sixth Amendment rights were violated. [The ICE official] did not testify at Martinez-Rios’s trial, despite having prepared the CNR. The government introduced the CNR through the testimony of [a Border Patrol agent], who explained how a CNR is ordinarily prepared. [The agent] personally reviewed Martinez-Rios’s A-file but did not conduct a search of any of the computerized databases associated with the CNR. Martinez-Rios was unable to cross-examine the person who had prepared a testimonial statement to be used against him at trial. Therefore, the district court erred in admitting the CNR without providing the testimony of the records analyst.

The error looks pretty plain, too, although the court skipped past that to the affected-substantial-rights prong, which requires a defendant to show "'a reasonable probability that, but for [the Confrontation Clause violation], the result of the proceeding would have been different.'" The Government said "harmless," and the court agreed:

[T]he government introduced ample evidence, other than the CNR, to establish that Martinez-Rios lacked permission to reapply. Blanco testified that he observed three individuals hanging onto the outside of a train northbound from Mexico as it pulled into the station, that he discovered Martinez-Rios hiding in the brush, and that Martinez-Rios admitted that he was from Mexico and had “no documents.” Mireles testified that Martinez-Rios said he was from San Luis Potosi, Mexico, and had crossed into the United States by wading across the Rio Grande, walking through the brush to the Laredo train depot, and boarding a northbound train. Melendez testified that he personally searched Martinez-Rios’s A-file for an I-212 (a form granting permission to enter the United States) but did not find any such document. In the face of all that testimony, even if the CNR was not entered into evidence, there was no reasonable probability that Martinez-Rios would have been acquitted.

"Okay," you say. "Big deal. I'll make sure to object on Confrontation Clause grounds. If the CNR stays out, I've got my Rule 29 motion. If that's overruled, I've got sufficiency to argue on appeal. If the CNR comes in, I'll have preserved the confrontation error, and on appeal the Government will have to show it was harmless beyond a reasonable doubt." All true. But will that make any difference?

Note that, in addition to the evidence summarized above, the court cited as support for its substantial-rights finding three court decisions applying the harmless error standard to the same question presented here. In one of those, an unpublished decision from the Second Circuit, the court found the error harmless because the defendant/alien admitted on cross-examination at trial that she had not applied for permission to reapply for entry. That's not your run-of-the-mill illegal reenty trial testimony, to be sure. But look at the other two decisions the court cites: two orders from the same judge denying motions for new trial based on the confrontation violation. In both cases, the evidence was very similar to that in Martinez-Rios: testimony from a Border Patrol agent that a search of the alien's A-file and relevant databases did not turn up the required permission, and the defendant's admission at the time of arrest that he was either present in the U.S. illegally or had not applied for permission to reapply for entry. (Martinez-Rios also cites an unpublished Tenth Circuit decision finding the evidence insufficient to support a conviction absent the CNR, but that appears to have been all the Government relied on. Plus, the court remanded for retrial in order to allow the Government to provide a foundation for the CNR.)

Of course, absent a specific factual scenario, it's impossible to say for sure how things would play out on a Rule 29 motion, motion for new trial, or an appeal raising sufficiency or a preserved Confrontation Clause violation. The point to take away, though, is that while Martinez-Rios is great on the confrontation issue, you can't simply assume that keeping out a CNR will automatically be fatal to the Government's case in an illegal reentry trial. You'll have to consider all the other direct and circumstantial evidence of lack-of-permission that the Government may introduce, along with legal and factual challenges to that evidence, and arguments for insufficiency notwithstanding such evidence. In other words, just like most any other case.

Suspended Sentence Not a "Sentence Imposed" for §2L1.2 Purposes, But Troubling Plain Error Analysis

United States v. Rodriguez-Parra, No. 08-40708 (5th Cir. Aug. 26, 2009) (Higginbotham, Smith, Southwick)

There's good and bad in this opinion. The good is that a suspended sentence isn't a "sentence imposed" for purposes of the 12- and 16-level drug-trafficking enhancements in guideline §2L1.2. The bad is that the panel continues a recent trend of ever-stricter plain error review, and in this case employs an approach to "plainness" that appears contrary to Supreme Court and Fifth Circuit case law.

So what happened? In 1994, Rodriguez was convicted of delivery of marijuana in Texas state court. He was sentenced to five years' imprisonment, which was suspended in favor of five years' probation, and served no part of his sentence in jail. When Rodriguez later illegally reentered the United States, the Texas conviction formed the basis for a 12-level enhancement under guideline §2L1.2(b)(1)(B), as "a conviction for a felony drug-trafficking offense for which the sentence imposed was 13 months or less[.]" The district court sentenced Rodriguez to 46 months' imprisonment, which was the bottom of the advisory Guidelines range.

On appeal, Rodriguez argued that the prior Texas drug conviction "cannot be used to enhance the [illegal reentry] sentence, because the [drug] sentence was entirely suspended." Because Rodriguez did not raise this argument in the district court, the court of appeals reviewed for plain error.

First, the good news. The court found error:

We must first decide whether there was error. The court enhanced under § 2L1.2(b)(1)(B), which increases a sentence by twelve levels “[i]f the defendant previously was deported, or unlawfully remained in the United States, after . . . a conviction for a felony drug trafficking offense for which the sentence imposed was 13 months or less.” Because the sentence was entirely suspended in favor of probation, we must determine whether it is a “sentence imposed” for purposes of § 2L1.2.

Section 2L1.2’s application notes answer that question. “‘Sentence imposed’ has the meaning given the term ‘sentence of imprisonment’ in Application Note 2 and subsection (b) of § 4A1.2.” U.S.S.G. § 2L1.2 Application Note 1(B)(vii). Section § 4A1.2 states,

"(1) The term ‘sentence of imprisonment’ means a sentence of incarceration and refers to the maximum sentence imposed.

"(2) If a part of a sentence of imprisonment was suspended, ‘sentence of imprisonment’ refers only to the portion that was not suspended."

Under subsection (2), therefore, the suspended portion of a sentence is not part of a “sentence of imprisonment.” Thus, if all of Rodriguez-Parra’s sentence was suspended, there is no remaining portion of the sentence that could refer to “sentence of imprisonment.” Application note two confirms this: “To qualify as a sentence of imprisonment, the defendant must have actually served a period of imprisonment on such sentence.” U.S.S.G. § 4A1.2 Application Note 2 (emphasis added).

There is caselaw that supports this conclusion. In United States v. Brown, 54 F.3d 234, 240 (5th Cir. 1995), we examined what constituted a “sentence of imprisonment” under § 4A1.2. That term “require[s] that . . . some time actually be served on the sentence.” Id. The only other court of appeals to examine this issue after the 2003 guideline amendments reached the same conclusion. Because all of Rodriguez-Parra’s sentence was suspended, his prior sentence did not meet the requirements for “sentence imposed” under the guidelines, so there was error under the first prong of the plain-error rubric.

As you can see, the same reasoning would apply to a sentence of straight probation, although probably not if any jail time was imposed as a condition of probation.

Now for the bad news: the court found that the error was not plain, and appears to have departed from controlling Supreme Court and Fifth Circuit precedent in reaching that conclusion. The court said, "There is no plain error if the legal landscape at the time showed the issue was disputed, even if, as here, the district court turns out to have been wrong. United States v. Ellis, 564 F.3d 370, 377 (5th Cir. 2009)." (emphasis added).

There are two problems with that statement. First, it's contrary to Supreme Court precedent. In Johnson v. United States, the Court held that, "where the law at the time of trial was settled and clearly contrary to the law at the time of appeal[,] it is enough that an error be 'plain' at the time of appellate consideration." 520 U.S. 461, 468 (1997) (emphasis added). And, consistent with Johnson, the Fifth Circuit has held that "[t]he plain or obvious nature of the error is determined by the law as it exists at the time of appellate review, rather than the time of the district court's ruling." United States v. Davis, 487 F.3d 282, 284 (5th Cir. 2007). Rodriguez-Parra does not discuss or attempt to distinguish these cases, or others that have followed Johnson. Instead, it just cites Ellis, which brings us to the second problem.

Ellis does not say that plainness is assessed by reference to the state of the law at the time of sentencing. In fact, the court in Ellis never even decided whether there was error in the first place, because it concluded that the question at issue was so close that no error could be plain: "even now after full briefing and oral argument the error is not plain or obvious, indeed it is most uncertain whether there was any error at all." Here, on the other hand, the court found easily found error.

Rodriguez-Parra goes on to treat as controlling a Fifth Circuit case that found no plain error on the same issue presented here: United States v. Garcia-Rodriguez, 415 F.3d 452 (5th Cir. 2005).

In Garcia-Rodriguez, the defendant had been convicted in California on a drug offense and sentenced to three years’ probation. Id. The district court used the earlier conviction to enhance, despite the fact that the defendant had not been imprisoned as a result of the earlier conviction. Id. Focusing on the same “sentence imposed” language from § 2L1.2 that we are currently reviewing, we held that the enhancement was not plain error. Id. at 456.

Garcia-Rodriguez relied on the fact that, "at the time of . . . sentencing[,] . . . [t]wo courts of appeals had interpreted the relevant provision in the same manner as the district court, and this court's law was unsettled." As Rodriguez-Parra describes it, "the lack of Fifth Circuit controlling authority on the issue meant that there was not plain error. [Garcia-Rodriguez, 415 F.3d] at 456. The presence of Garcia-Rodriguez as binding precedent means, therefore, that the resolution of Rodriguez-Parra’s sentencing issue is less than obvious." Rodriguez-Parra goes on to say that its finding of error

is reached . . . only by a careful parsing of all the relevant authorities, including the sentencing guidelines and applicable decisions. “[T]o satisfy the second prong of plain error inquiry, ‘the legal error must be clear or obvious, rather than subject to reasonable [dispute].’” Ellis, 564 F.3d at 377-78 (quoting Puckett v. United States, 129 S. Ct. 1423, 1429 (2009)). Here the result is reached only by traversing a somewhat tortuous path.

But that's an odd description of the court's analysis of the issue, which is reproduced in full up above. The court only needed three relatively brief paragraphs of analysis to conclude that there was error. There was no indication here, as in Ellis, that the legal error was subject to reasonable dispute; indeed, the court apparently saw it as an easy question with a straightforward answer. And it's hard to say that that a simple cross-reference to another guideline amounts to a "somewhat tortuous path" (it's nothing compared to §2L1.2(b)(1)(C)'s incorporation of Title 8's "aggravated felony" definition, for example). Moreover, to the extent that the court is suggesting that error cannot, by definition, be "plain" whenever there is a lack of controlling case law on the question, that seems wrong, too. Especially if the relevant guideline provisions are clear, as they seem to be here.

Of course, this decision isn't all that surprising in light of what seems to be a trend towards a more stringent plain error review. There's that cite to Puckett, for example. There's also this:

That [tortuous] path . . . would have been more straight and level had Rodriguez-Parra made his objection in the district court, thus allowing each side to present its arguments on the question now at hand. This is the very point of the plain-error standard—it “serves powerful institutional interests, including securing the role of the United States District Court as the court of first instance.” [Ellis] at 378.

These concerns would carry more force if the question at issue were a closer one, but—again—that evidently wasn't the case here. They also sound more relevant to the third and fourth prongs of plain error review, not the second.

Given Rodriguez-Parra's apparent departure from established plain-error principles, this case would seem to be a good candidate for panel or en banc rehearing. We'll see what happens.

Fives Address Issues Concerning Reasonableness Review and Plain Error In Course of Rejecting Various Challenges to Revocation Sentence

United States v. Whitelaw, No. 08-50346 (5th Cir. Aug. 19, 2009) (Davis, Owen, Haynes)

If the court imposes a term of imprisonment on revocation of supervised release, may it order that term to run consecutively to any other sentence, even though the sentence for the original offense—as required by the terms of the plea agreement—ran concurrently with an already-imposed state sentence? That question, and others concerning plain error and reasonableness review (including a circuit split!), are answered in Whitelaw.

The lengthy backstory:

Alan Whitelaw was convicted of theft of money in an amount exceeding $200,000 in state court and was sentenced to 60 years of imprisonment. Whitelaw subsequently pleaded guilty to federal bank fraud charges in the Southern District of Texas. While the same type of fraudulent conduct was the basis for both of Whitelaw’s convictions, the federal and state convictions involved different specific conduct, dates, and victims.

Although Whitelaw pleaded guilty to the federal charge without a written plea agreement, the Government made an oral agreement at rearraignment that it would recommend that Whitelaw’s sentence run concurrently with his state court sentence. The district court accepted the agreement as a plea agreement under FED. R. CRIM. P. 11(c)(1)(C), ruling that Whitelaw would be allowed to withdraw his guilty plea if it did not order that his sentence run concurrently with his state court sentence. The district court sentenced Whitelaw to 46 months of imprisonment, the sentence to run concurrently with his state court sentence, and five years of supervised release.

For reasons not explained in the opinion, Whitelaw was released from state prison after serving just five years of his sentence. He then began serving his federal supervised release term.

A little over three years later, Whitelaw was arrested on a revocation violation warrant. The Government then filed two motions to revoke Whitelaw's supervised release, the latter of which alleged six violations, including "committing the felony offense of theft of copper wire valued in excess of $1,000 or possession of stolen copper wire valued in excess of $1,000," lying to his probation officer about it (both orally and with forged documents), and technical violations.

After a hearing, the court found all but one of the alleged violations (a failure to report) to be true. Whitelaw requested a sentence at the low end of the advisory 4- to 10-month Guidelines range. The court instead sentenced him to 36 months—the statutory maximum—"and ordered that sentence run consecutively to any other state or federal sentence."

Following the revocation of his supervised release, Whitelaw filed a pro se motion that included a claim that he should be allowed to withdraw his guilty plea because his sentence upon the revocation of supervised release was not ordered to run concurrently with any state court sentence he received. The district court struck the motion because Whitelaw was represented by counsel and, therefore, not entitled to make pro se filings. In the alternative, the district court denied the motion on its merits.

Which finally gets us to the issues on appeal . . .

Standard of Review
For at least the third time, the court declined to decide whether revocation sentences are reviewed for reasonableness or plain unreasonableness, because all the issues Whitelaw raised were subject to plain error review:

• "Whitelaw did not raise any of the specific claims of procedural error that he argues in this appeal when he was before the district court for sentencing[,]" so plain error review for those (failure to consider the policy statements in Chapter 7 of the Guidelines Manual, and an inadequate statement of reasons for the sentence).
• Whitelaw's motion challenging the consecutiveness order "was stricken by the district court because Whitelaw was represented by counsel. Accordingly, Whitelaw did not properly raise this claim of specific legal error below, and this court reviews it for plain error only."
• "Whitelaw also did not object to the substantive reasonableness of the sentence below. Whitelaw’s contest of the revocation charges and request for a sentence at the low end of the guidelines range are insufficient to preserve the substantive reasonableness of the sentence for review." See Peltier. (But don't forget that circuit split.)

No Error In Running Sentence Consecutive to Any Other Sentence, Notwithstanding Plea Agreement on Original Offense
Whitelaw challenged the consecutive order on his revocation sentence on two grounds, both of which the court rejected. First, he argued that the district court didn't indicate it realized that it could order the sentence to run concurrently with any other sentence. While the district court didn't specifically acknowledge it's discretion in the matter, the Government asked for the revocation sentence to run consecutively, and the district court express doubt that it could do that. "In the absence of evidence to the contrary, this court assumes that the district court knows the law and applies it correctly[,]" so no plain error.

Second, Whitelaw argued that the terms of his original plea agreement required any revocation sentence to run concurrently to any other sentence, as was the case with his original sentence. Not so, said the court. The rearraignment transcript wasn't in the record, but the PSR

describes the plea agreement as requiring that Whitelaw’s sentence run concurrently with the state court sentence Whitelaw was then serving. We see no indication that the plea agreement applied to any sentence other than for the crime upon which he was charged. The judgment of conviction provides only that Whitelaw’s sentence of imprisonment would run concurrently with the state court sentence that Whitelaw was then serving; it did not impose any restrictions upon possible sentences that could be imposed if Whitelaw’s later term of supervised release was revoked.

(emphasis added). Is that right? It may be the case that the plea agreement didn't purport to cover what would happen in the event of a revocation of supervised release (and given the lack of a written plea agreement or a rearraignment transcript, it's impossible to say the court's wrong about that). But to the extent that the court is suggesting that a revocation sentence is for a crime other than the one for which a defendant was originally convicted, the court applies a mistaken understanding of revocation. As the Supreme Court explained in Johnson v. United States, a post-revocation sanction is an additional punishment for the original offense, not a punishment for the violation of the terms of release. 529 U.S. 694, 699–701 (2000). Treating post-revocation sanctions as punishment for violations of conditions of supervised release would raise "serious constitutional questions," given that "the violative conduct need not be criminal and need only be found by a judge under a preponderance of the evidence standard, not by a jury beyond a reasonable doubt." Id. at 700. The court doesn't address Johnson here, but it's something to keep in mind if you find yourself with a case similar to Whitelaw.

Whitelaw Failed to Show that the District Court Failed to Consider the Guidelines' Revocation Policy Statements
Whitelaw argued on appeal that the district court failed to consider the Chapter 7 policy statements. But the court of appeals held that he failed to show that, because 1) the district court correctly calculated the advisory revocation range and referred to it before imposing sentence, 2) "[d]efense counsel referred to that range in its argument[,]" and 3) the Government also referred to that range in its motions to revoke, and "made arguments for an above range sentence related to the § 3553(a) factors[.]"

District Court Plainly Erred in Failing to State Reasons for the Sentence, but That Error Did Not Affect Whitelaw's Substantial Rights Because the Record Reveals the Reasons
Given that the district court imposed a sentence that was more than three times the high end of the advisory Guidelines range, Rita required some explanation. The district court gave no reasons. That was clear and obvious error. But did the error affect Whitelaw's substantial rights? And how does the court of appeals assess that?

As it turns out, up until now the Fifth Circuit "ha[d] not yet applied plain error review to a district court’s failure to state reasons for an above guideline sentence." (emphasis added). As you'll recall, Mondragon-Santiago reviewed such an error in the context of a within-Guidelines sentence. It held that that the defendant must "show that the error actually did make a difference [in the sentence]: if it is equally plausible that the error worked in favor of the defense, the defendant loses; if the effect of the error is uncertain so that we do not know which, if either, side it helped the defendant loses." Mondragon-Santiago rejected the approach of "other circuits [that] have relaxed this requirement in the sentencing context[.]"

But given that Mondragon-Santiago involved a within-Guidelines sentence, Whitelaw had to decide whether that approach to the third plain-error prong also applies to review of above-Guidelines sentences. To that end, Whitelaw canvassed what appears to be a circuit split on the question:

In cases involving above guidelines sentences, the D.C. Circuit and the Second Circuit have relaxed the showing required by a defendant to demonstrate that the error affected the defendant’s substantial rights. These courts reason that the absence of a statement of reasons affects the appellate court’s ability to perform a meaningful review of the sentence. Both circuits also conclude that the failure to state reasons undermines “the public reputation of judicial proceedings” because the statement of reasons indicates to the public that the sentencing judge has thoughtfully discharged his responsibilities and demonstrates that the judgment is not arbitrary. [So has the Sixth Circuit.]

Other circuits have declined to follow that lead. The Tenth Circuit affirmed a below guidelines sentence in which the government argued procedural error for failure of the sentencing court to cite reasons for a downward departure. The Tenth Circuit refused to find that the error affected the government’s substantial rights because, applying traditional plain error analysis, the government could not explain how a more detailed reasoning process might have led the court to select a higher sentence. In other words, the government could not show that the error made a difference in the sentencing outcome. The court refused to presume prejudice.

(cites omitted). Ultimately, Whitelaw found it unnecessary to take sides on this question, because "our review of the record reveals the reasons for Whitelaw’s sentence, even if not explicitly stated by the district court." The revocation hearing was extensive. The district court heard evidence, and both parties presented sentencing arguments. "The government then urged the court to sentence Whitelaw to 36 months in prison (the statutory maximum) for the five reasons set forth in its motion to revoke and that the sentence be consecutive to any other sentence. The district court then granted the government’s motion and sentenced Whitelaw to the government’s recommended 36 month term of imprisonment." (By the way, a portion of the Government's motion is reproduced in the opinion, and is predictably purple.) The court then reaches this troubling conclusion:

We have no trouble concluding that by granting the government’s Motion to Revoke and sentencing Whitelaw to the term of imprisonment recommended and supported in that motion, the district court implicitly adopted the government’s rationale for that sentence as set forth in the motion. These reasons thoroughly explain Whitelaw’s sentence. The factual basis supporting them is implicit in the district court’s findings that most of the revocation charges were true. By reviewing the record of the sentencing proceedings in this case, we are able to conduct a meaningful appellate review. Accordingly, the district court’s failure to state these reasons on the record has not prejudiced Whitelaw.

Will that give prosecutors an incentive to paper the case with overheated arguments for a stiff revocation sentence, in hopes that the district court won't state reasons of its own? Perhaps. If possible, you might consider trying to blunt the effect of such pleadings with your own written submission (which may or may not be possible, depending on whether the releasee challenges the alleged violations, and on whether the court will allow time for written arguments before deciding on a disposition). Of course, it's also a good reason to object to procedural errors at the hearing so you're not stuck with plain error review in the first place.

On Review for Substantive Reasonableness, Whitelaw's Sentence was not Plain Error
After all that, the court's resolution of Whitelaw's substantive reasonableness challenge was pretty brief. The court simply said that because the sentence did not exceed the statutory maximum, it wasn't plain error, citing other Fifth Circuit opinions that also so held. Is the court relying on a broader principle that a sentence within the statutory range can never be plain error? If so, that sounds questionable. Hopefully, the court is just saying that, as a descriptive matter, such sentences will have a very hard time satisfying all four plain error prongs. And given the poor track record of preserved substantive reasonableness challenges, that's probably true.

Courts Lack Authority to Suspend a Failure-to-Depart Sentence After the Sentence Has Begun

United States v. Garcia-Quintanilla, No. 08-50400 (5th Cir. July 7, 2009) (Higginbotham, Garza, Prado)

Prefatory disclosure: I did some work on this case.

Ever heard of 8 U.S.C § 1253? It criminalizes conduct relating to an alien's failure to depart the United States after having been ordered to do so. It also contains a unique provision, in subsection (a)(3), permitting a district court to suspend the sentence of an alien convicted of failure-to-depart. The question presented in this case: does § 1253(a)(3) permit a court to suspend a failure-to-depart sentence after the sentence has begun? Answer: no.

Some context: Garcia, a citizen of El Salvador, was ordered removed from the United States. He refused, however, to cooperate with Salvadoran consular officials in obtaining the travel documents necessary for him to return to El Salvador. Consequently, the Government indicted him for failure-to-depart. "Up until the eve of trial, the Government offered to drop the charges if Garcia-Quintanilla would cooperate in his removal. [He] refused these offers, and a jury later found him guilty."

Although Garcia's advisory Guidelines range topped out at only 6 months, the Government sought an upward variance to the statutory maximum of 4 years, "suggesting that such a sentence would deter Garcia-Quintanilla from persisting in his refusal to speak with the Salvadoran consulate." The Government also justified its request by pointing to § 1253(a)(3), which it claimed would allow the district court to suspend Garcia's sentence should he change his mind in the future and cooperate in his removal. The district court—concerned that a six-month sentence would be insufficient to prompt Garcia's cooperation, thus leading to a series of piecemeal prosecutions for what amounts to an ongoing offense—agreed with the Government's arguments and sentenced Garcia to 4 years' imprisonment. In effect, the court of appeals noted, the sentence operated like civil contempt.

Garcia appealed, arguing that § 1253(a)(3) does not permit a district court to suspend a sentence after it has begun, and that his sentence was unreasonable because it was predicated on such authority. As Garcia did not raise the lack-of-authority argument in the district court, the court of appeals reviewed for plain error.

On the first question—whether there was error—the court began by examining the text and history of § 1253(a)(3), and found both lacking. The statute itself says nothing about when a district court may suspend the sentence, only that it may do so in certain circumstances. Legislative history was also unhelpful. The statute was enacted as part of IIRIRA in 1996, although it amounted to a reenactment of a statute that was part of the Immigration and Nationality Act of 1952. Nothing in the legislative history of either Act addressed the timing-of-suspension point. Complicating matters further, in the mid-1980's Congress passed the Sentencing Reform Act, which eliminated district courts' authority to suspend sentences.

The court, however, found guidance in "the historical practice of sentence suspension[,]" which revealed two important points. First, courts had no inherent authority to order suspension; the practice could only be authorized by Congress. Second, the Supreme Court interpreted the pre-SRA statutory scheme to "permit suspension of a sentence only before that sentence had begun." Consequently, "if Congress intended for § 1253(a)(3) to permit the suspension of a sentence after it has begun, we might expect Congress to say explicitly that a court may suspend a sentence before or after its execution, or we might expect the outlines of a procedure for bringing to the court’s attention the circumstances that warrant a suspension after the sentence has begun." Because the text and structure of § 1253(a)(3) appear to focus on the time of sentencing, "[a]nd because there is nothing in § 1253(a)(3) to indicate that Congress intended to deviate from the baseline of sentence suspension, we must presume that Congress intended § 1253(a)(3) to provide for suspension when suspension has normally occurred."

So there was error. But was it plain? Yes. Although there's no case law on the question, "it has long been the law that district courts cannot suspend a criminal sentence after it has begun. This error did not, therefore, merely rest on a misinterpretation of a statute. It also represented a substantial step beyond the district courts’ traditional authority as defined by the Supreme Court."

But did the error affect Garcia's substantial rights? Yes again. There was a reasonable probability that the sentence would have been lower absent the error, because 1) the sentence was eight times longer than the high end of the Guidelines range, and 2) "the possibility of suspension was an essential aspect of this sentence."

Okay, last hurdle: did the error seriously affect the integrity and fundamental fairness of the proceedings. Indeed it did, for a couple of reasons. First, were the court of appeals "not to correct the error, the end result would be a sentence that no one ever intended and which the court lacked the power to craft as it did[,]" as "the possibility of suspension was an essential aspect of Garcia-Quintanilla’s sentence." Second—and here's where you really want to pay attention—"the length of Garcia-Quintanilla’s erroneously-imposed sentence, the magnitude of which might itself be sufficient under many of our prior decisions. See, e.g., United States v. Sanchez, 527 F.3d 463, 466 (5th Cir. 2008) (holding that an error affected the fairness of judicial proceedings when it resulted in a sentence over two times longer than the proper Guidelines range)." (emphasis added). That may be the most significant part of this opinion. While there are very few prosecutions for failure-to-depart, plain-error application comes up all the time, and the fourth prong can be a big hurdle.

A final point . . . Garcia suggested that § 1253(a)(3) is a dead letter, in light of the Sentencing Reform Act's elimination of suspension authority. The court declined to decide the issue: "The only question in the present case is whether § 1253(a)(3) permits suspension after a sentence has begun; we can simply and safely assume that § 1253(a)(3) authorizes suspension up to the moment when a sentence begins." The court, though, went on to express "serious doubt as to Garcia-Quintanilla’s suggestion[,]" in light of the fact that § 1253(a)(3) was enacted (or re-enacted, depending on how you look at it) after the SRA.

Guidelines Calculation Error Doesn't Require Reversal If Sentence Didn't Result From That Error, Plus More on Plain Error

United States v. Ruiz-Arriaga, No. 08-40242 (5th Cir. Apr. 9, 2009) (Jones, Wiener, Benavides)

We know that, when imposing sentence, the district court must correctly calculate the advisory Guidelines range. Let's say the court gets the Guidelines calculation wrong, but says something along the lines of, "Even if the Guidelines range was different, I would still impose a sentence of X months, for the following reasons . . . ." Does the procedural error require vacation of the sentence?

Not necessarily. The Fifth Circuit recently clarified that a procedural error does not require reversal if the error was harmless, meaning that the error did not affect the sentence imposed. That's essentially what Ruiz-Arriaga holds, although it couches the matter in slightly different terms.

In this case, Ruiz pleaded guilty to illegal reentry. He had a prior Texas conviction for sexual assault of a child. The PSR recommended a 16-level COV enhancement for that prior conviction, producing an advisory range of 46 to 57 months. Ruiz objected to the enhancement (the details of which the opinion doesn't explain), arguing that the correct range was 10 to 16 months (presumably based on treating the sexual assault prior as a 4-level "any other felony" under §2L1.2(b)(1)(D), although the opinion doesn't explain that, either). The district court overruled the objection and sentenced Ruiz to 46 months, explaining that,

in the event that the court is incorrect about the guideline range, that a sentence at certainly something more than 10 to 16 months, which if the court were wrong is what—I haven’t recalculated, but that’s what [defense counsel] says would be the range. The court believes that a sentence of 46 months in custody would be reasonable, even if the court is not correct about the guideline range.

Ruiz appealed. Without addressing whether the district court erred in applying the 16-level enhancement, the court of appeals concluded that any error was harmless:

As we noted in United States v. Bonilla, 524 F.3d 647 (5th Cir. 2008), “[n]ot all errors in determining a defendant’s guideline sentence require reversal.” Id. at 656. Where the district court “considered the possible guideline ranges that might apply to the defendant with and without a disputed enhancement for a crime of violence under U.S.S.G. § 2L1.2(b)(1)(A),” and otherwise explains its sentence in terms of 18 U.S.C. § 3553, the resulting sentence does not result from an incorrect application of the Guidelines. Id. Here, the district court considered the 46 to 57-month range with the crime of violence enhancement and the 10 to 16-month range advocated by the defense.

Of course, this brings up the question of just how much 3553 explanation is necessary to avoid a causal link between a Guidelines-calculation error and the sentence imposed. It must be more than the above quote from the district court, because the opinion also mentions that "[d]uring the sentencing hearing, the district court stated that it had considered the non-enhancement range offered by the defense and would have imposed the same sentence independently of the Guidelines in light of the seriousness and recent nature of the statutory rape conviction." So an explanation stating simply that "this court would impose the same sentence even if it is wrong about the applicable Guidelines calculation" shouldn't cut it. The district court must still address the defendant's argument, and provide 3553-based reasons for the alternative non-Guidelines sentence.

And in keeping with a recent theme, the opinion offers another sign that the Fifth Circuit is getting stricter about plain error review. Although Ruiz argued in the district court that the correct advisory range was 10 to 16 months, he argued on appeal that it was actually 8 to 14 months due to an error in the criminal history calculation. As Ruiz didn't make the argument below, plain error review applied. Quoth the court:

First, we cannot attribute “plain” error to a district court decision when defense counsel affirmatively represented to the district court a sentencing range that appellate counsel now disavows. Trial counsel’s misstatement essentially waives the argument for any other sentencing range in this appeal. It is well established that appellate courts may correct errors of law under a plain standard where trial counsel simply stood mute at sentencing and failed to object to the PSR. . . . This is because a “plain” error is one that the district court, in its oversight of sentencing, should have been alert to correct. Sentencing under the Guidelines is so complex, however, that the court should not be faulted for “plain” error when counsel’s affirmative statements allay any possible concern.

(emphasis added). Plus, no prejudice: "It is hardly likely that the court would have responded differently to a two month difference in range when it so clearly rejected the slightly higher range."