Wednesday, August 19, 2020

8/19/20: Big day in the Ninth Circuit

 The Ninth Circuit published eight opinions today, including two criminal and two habeas.  I'm going to cover the two criminal decisions and one of the habeas decisions.

First, United States v. Valencia-Lopez, --- F.3d ---, No. 18-10482 (9th Cir. 2020), is an important decision for anyone practicing in a border district. 

The Court vacated the defendant's convictions for drug importation, holding that the district court erred in allowing expert testimony that the likelihood drug trafficking organizations would entrust a large quantity of illegal drugs to the driver of a commercial vehicle who was forced or threatened to comply was “[a]lmost nil, almost none.”  

The Court concluded the district court erred in admitting this testimony without a proper reliability determination under Daubert.  There is lots of helpful language.  The takeaway is that expert testimony without a specific reliability determination for that testimony is error.  Here are the highlights:
  • Under Rule 702, before admitting expert testimony, “the district court must perform a ‘gatekeeping role’ [to] ensur[e] that the testimony is both ‘relevant’ and ‘reliable.’” This gatekeeping obligation “applies to all (not just scientific) expert testimony.” 
  • But “the trial court’s broad latitude to make the reliability determination does not include the discretion to abdicate completely its responsibility to do so.” We have explained that a district court abdicates its gatekeeping role, and necessarily abuses its discretion, when it makes no reliability findings.
  • Daubert and Kumho Tire may be harder to apply when the expert testimony is “experience-based” rather than “science-based.” But any such difficulty cannot simply lead to a “that goes to weight, not admissibility” default, as here. Indeed, we see a strong argument that reliability becomes more, not less, important when the “experience-based” expert opinion is perhaps not subject to routine testing, error rate, or peer review type analysis, like science-based expert testimony.
  • Here, the district court made no reliability findings about Agent Hall’s testimony. Agent Hall had sufficient experience and knowledge to qualify as an expert. But the record contains no evidence as to why that experience, by itself, equals reliability for his testimony that there was “[a]lmost nil, almost no[]” likelihood drug cartels would do what Valencia-Lopez testified happened here—coerce him, at gunpoint, to carry illegal drugs across the border.
  • While the district court here did not have to hold a Daubert hearing,  it still had an unwavering gatekeeping obligation to determine the reliability of Agent Hall’s testimony. 
  • With no Daubert hearing or voir dire, “‘some [other] reliability determination must be apparent from the record’ before we can uphold a district court’s decision to admit expert testimony.”  We find none in the record; rather, the district court qualified Agent Hall as an expert without explicitly finding his proposed testimony reliable. That was an abuse of discretion.
  • The district court also did not fulfill its gatekeeping role for a second, more important reason. The court admitted Agent Hall’s testimony despite the government establishing no reliable basis for Agent Hall’s “[a]lmost nil, almost none” expert testimony. 
  • It is one thing for a witness with Agent Hall’s expertise to testify as to the risks to a cartel of using a coerced courier. But that is a far cry from him essentially testifying that the cartel never does it. Such testimony would not be allowed, for example, in a civil case involving a death allegedly caused by an inoculation without a rigorous reliability evaluation of the causation opinion.7 We know the nature of such evaluation is not carved in stone, either for that hypothetical case or this real one. But we also know that this evaluation is required.
  • Agent Hall never explained the methodology, if any, that he relied on to arrive at the near zero probability of drug trafficking organizations using coerced couriers. Rather, the district court appeared to rely on Agent Hall’s general qualifications to qualify him as an expert, contrary to our admonition in Hermanek. See id. Moreover, the record does not establish the method Agent Hall employed to arrive at his “almost nil” opinion. 
  • His qualifications and experience are relevant, and indeed necessary. But they cannot establish the reliability and thus the admissibility of the expert testimony at issue. Rather, like the expert in Hermanek, Agent Hall “failed to explain in any detail the knowledge, investigatory facts and evidence he was drawing from,” to eventually conclude the probability of coercion by drug trafficking organizations was “almost nil.” Crucially, he failed to link his general expertise with his “almost nil” conclusion, and by never explaining how his expertise lent itself to that conclusion, we cannot sort out what “reliable principles and methods underlie the particular conclusions offered.” 
  • The issue is not whether Agent Hall had knowledge and experience sufficient to allow him to testify as an expert on the modus operandi of drug cartels. He did. Nor is the issue whether he had sufficient “background for his opinions.” Rather, the issue is whether he provided a reliable basis for his opinion that the likelihood of drug cartels using coerced couriers is “[a]lmost nil, almost none.” As explained above, he did not.
  • We do not question that expert modus operandi testimony is admissible in drug smuggling cases involving unknowing or coerced couriers. But the government must still establish that its expert opinions are reliable under the standards mandated by Daubert and Kumho Tire. The government failed to do so here.
Moving to whether the error was harmless, the Court explained, "[w]e begin with the presumption of prejudice, which the government can rebut by showing “it is more probable than not that the jury would have reached the same verdict even if the evidence had not been admitted.” 

The Court held: "the government bears the burden of proving the error was harmless, and it cannot merely speculate that the jury both disregarded Agent Hall’s testimony and disbelieved Valencia-Lopez’s testimony. The government has not carried its burden -- the error here was not harmless. We thus vacate Valencia-Lopez’s convictions and remand for a new trial.


Next, United States v. Swenson, --- F.3d ---, No. 18-30215 (9th Cir. 2020), deals with restitution under the Mandatory Victims Restitution Act (MVRA).

In very short, the defendant was convicted of fraud crimes and ordered to pay restitution.  To enforce the restitution order, the government obtained a writ of garnishment against the defendant's wife's bank account.  The account held her Social Security benefits.  The government's theory was that those funds were subject to garnishment as community property.

The Ninth Circuit said no. It held, the wife's "Social Security benefits are not subject to garnishment pursuant to the MVRA in connection with her husband’s criminal restitution order."

Also, as to jurisdiction, the Court explained, "[a] disposition order (or termination of the garnishment otherwise) concludes litigation of the writ on the merits and is thus a final judgment for purposes of appeal."


Finally, in Kipp v. Davis, --- F.3d ---, No. 16-99004 (9th Cir. 2020), the Court reversed the denial of the defendant's 2254 petition challenging his California conviction and death sentence for first degree murder and attempted rape.

Two things to highlight about this case.  

First, it was decided under 2254(d)(2).  The Court found the California Supreme Court based its decision  on an unreasonable determination of the facts.  Thus, the Court considered the underlying claim without the deference AEDPA otherwise requires.  There is a very detailed and helpful discussion of 2254(d)(2) review.

Second, as to the underlying claim, the Court held that the admission of other act evidence (an unadjudicated murder) violated the defendant's due process rights.  

The general test is “whether the admission of evidence rendered the trial so fundamentally unfair as to violate due process.”

But there is also a more detailed test: [T]he admission of other crimes evidence violate[s] due process where: (1) the balance of the prosecution’s case against the defendant was “solely circumstantial;” (2) the other crimes evidence . . . was similar to the [crime] for which [the defendant] was on trial; (3) the prosecutor relied on the other crimes evidence at several points during the trial; and (4) the other crimes evidence was “emotionally charged.”

Applying this test, the Court "conclude[d] that there is no highly distinctive pattern that would justify an inference of the same intent or perpetrator behind both crimes. Rather, there is a high risk that a juror would have assumed that if Kipp committed the Frizzell homicide, he had the propensity to commit the Howard homicide as well."

Thus, "the trial court’s admission of the Frizzell evidence deprived Kipp of a fundamentally fair trial in violation of his due process rights."  The Court found actual prejudice and reversed. 

[In a different opinion, the Court affirmed the denial of the defendant's other 2254 petition from a different murder conviction and death sentence].