Sunday, April 22, 2012

Gain? Wait!

United States v. Zangari, No. 10-4546-cr (2d Cir. April 18, 2012) (Cabranes, Pooler, Wesley, CJJ)

In this decision, the court found that the district court’s restitution order, which was based on the defendant's gain instead of the victims' loss, was error, but not plain error. It accordingly affirmed.

Defendant Zangari was a securities broker in the securities-lending departments of two major banks.  He engaged in unauthorized stock-loan transactions with financial institutions that had a relationship with one of his co-workers, and received a portion of the kickbacks, approximately $65,000.  His employers  suffered “losses in the form of unrealized profit.”

Zangari pled guilty to a Travel Act conspiracy, and was sentenced under USSG § 2B4.1, the commercial bribery guideline. The PSR used the $65,000 figure as the loss calculation, recommending an enhancement for a loss between $30,000 and $70,000. Although neither bank had submitted a loss affidavit, the PSR used Zangari’s gain as a proxy for their losses. The report equated the kickback amount the banks’ lost profit on the transactions, but did not detail how it reached this conclusion.

Zangari never objected to the loss calculation, and the district court used it both for calculating his guidelines and for fixing the amount of restitution.  On appeal, however, Zangari argued that the restitution order was illegal because the victims suffered no loss.

The circuit agreed that the restitution order was erroneous. The restitution statutes require that the order reflect the “full amount of each victim’s loss.” But here, the district court based the order instead on Zangari’s gain. While this is acceptable for calculating the loss for guidelines purposes - an application note permits using gain as a substitute for loss where the loss “reasonably cannot be determined” - the substitution is not permissible for calculating restitution.  Indeed, every circuit to consider the question has reached the same result, even in “hard cases.”

It is true that there are cases where there is a direct correlation between the defendant’s gain and the victim’s loss; in such situations the gain is a “measure of” but not a “substitute for” the loss.  But here, there was no such correlation. In the stock-loan transactions at issue the securities and collateral involved were returned to their previous owners at the end of the loan. “Therefore, any loss to the identified victims in this case could only have come in the form of opportunity cost.” And those losses were not equivalent to the sham finders fees that produced Zangari’s gain.

Accordingly, the order was error, and the error was “plain.” Nevertheless, the circuit declined to exercise its discretion to correct it because Zangari failed to show prejudice by demonstrating that the amount would have been less had it been properly calculated.  Rather, he insisted only that the banks’ failure to file affidavits of loss showed that they suffered no loss at all. The circuit disagreed: “the fact that the victims did not claim a loss does not mean that they did not sustain” one.

The circuit also thought this might even be a case of “salutary error,” in that it was possible that the restitution award “in fact understated the victims’ actual losses.”  Zangari pled guilty to being a member of an “industry-wide” conspiracy, and could have been held liable for all of the losses of all of the victims affected by it along with, under § 3663A(b)(4), the expenses they incurred in their internal investigations, including attorneys fees and accounting costs.

Labels: ,

Remorse Code

United States v. Aleynikov, No. 11-1126 (2d Cir. April 11, 2012) (Jacobs, Calabresi, Pooler, CJJ)

Sergey Aleynikov, a former Goldman Sachs computer programmer, stole a portion of Goldman’s proprietary high frequency trading (“HFT”) computer code, apparently in preparation for taking a related, but higher paying, job at a startup company.  A jury convicted him of violating 18 U.S.C. § 2314, which makes it a crime to transport stolen “goods” in interstate commerce, and § 1832, which makes it a crime to steal a trade secret that is related to or included in a “product” that is “produced for or placed in” commerce.  Two months ago, the circuit reversed these convictions in a one-line order with an opinion to follow.

And here it is. While we were all expecting a sufficiency-of-the-evidence opinion, the court instead concluded that the indictment charging Aleynikov with those crimes was itself insufficient because it failed to allege a crime within the terms of the applicable statutes.

First, as to the stolen property count, the court concluded that the stolen HFT code did not constitute “goods,” “wares,” or “merchandise” within the meaning of the statute.  

In 1966, the court had held that stolen photocopies of documents detailing a pharmaceutical manufacturing process were “goods” because the documents were tangible items that were transported across state lines. But that case also suggested that if intellectual property were stolen in an intangible form, the statute would not apply. Moreover, in 1985, the Supreme Court held that the interstate transmission of intangible property - there it was the musical compositions contained in bootleg recordings - did not violate § 2314. Several circuits have subsequently held that intangible property is not a “good” “ware or “merchandise.” 

Aleynikov’s case is no different. He uploaded Goldman’s HFT code to a server in Germany, thereby stealing “purely intangible property embodied in a purely intangible format.”  The indictment did not allege that he “physically seized anything tangible from Goldman,” such as a CD or a flash drive.  And, while he later copied some of the code onto his own laptop and flash drive and transported it interstate, this did not violate the statute either, since the “thing stolen” must be a good or ware “at the time of the theft.”  “The later storage of intangible property on a tangible medium” - just like the bootleg phonograph records in the Supreme Court case - “does not transform the intangible property into a stolen good.”

For the trade secret count, the court concluded that code was not “related to or included in a product that is produced for or placed in” commerce, as required by § 1832.  This language covers two distinct categories of products: those “placed in” commerce are already in the marketplace, while those “produced for” commerce are still in development.  Clearly the source code here was not a product “placed in” commerce. But was also not “produced for” commerce simply because its function was to execute trades in financial markets.  And Godlman had no intention of selling its HFT system or of licensing it to anyone else. Thus, since that code “was not designed to enter or pass in commerce, or to make something that does,” its theft did not violate § 1832. At a minimum, the statute was ambiguous as to whether it would cover stolen HFT code, and under the rule of lenity the outcome would be the same.

In a concurring opinion, Judge Calabresi noted that, while it certainly appeared that Aleynikov’s theft of Goldman’s code was the type of “mischief” that Congress intended to be covered by § 1832, he agreed with the majority’s textual analysis. He suggested that Congress “return to the issue.”


Thursday, April 12, 2012

Scott Free

United States v. Scott, No. 10-3978-cr (2d Cir. April 6, 2012) (Pooler, Parker, CJJ)

In 2009, two NYPD detectives arrested defendant Scott after witnessing him engage in what they said was a hand-to-hand drug sale. At trial, the district court permitted the detectives to testify, over objection, that they had seen Scott several times before, and had spoken to him several times, for as long as twenty minutes. The circuit, finding that this evidence violated both Rule 404(b) and Rule 403, vacated the judgment and remanded the case for a new trial.

The circuit first concluded that the evidence was indeed Rule 404(b) evidence, and not something else. Rule 404(b) covers other “acts,” not other “bad acts,” and here, the detectives’ description of their prior contacts with Scott clearly would bear adversely on the jury’s assessment of his character. The court distinguished this case from those where the evidence was limited to officers’ testifying that they had seen the defendant in the past. “The difference between a police officer’s mere observations of a defendant in an area and testimony that two different detectives had had occasion to speak to him up to five times and for up to twenty minutes ... is substantial.” This latter scenario “indicates to a jury” that the defendant is “at a minimum, the sort of person who warrants a level of police observation to which law-abiding citizens are unaccustomed” and thus that he is a person “with a propensity to engage in wrongful, criminal or otherwise unusual behavior that would attract the attention of the police.”

Next, the court found that the evidence was inadmissible under Rule 404(b), following the Huddleston factors. The court found that the evidence was not admissible as to Scott’s identity, because Scott’s identity was not at issue. His defense was that he did not engage in a drug sale, and not that the detectives arrested the wrong person. But even if identity were disputed, that would only justify permitting the officers to testify that they had seen Scott before, not that they had had numerous lengthy conversations wit him.

The court also rejected the government’s novel argument that the officers’ familiarity with Scott would somehow free up their powers of observation to focus better on what he was doing. In fact, this argument only supported the view that the evidence went to Scott’s criminal propensity. “There is only one reason a person is better able to tell what someone he knows is doing, which is that he knows what someone he is familiar with is likely to be doing.”

The court also held that the evidence was not admissible to bolster the credibility of the detectives by explaining their actions, which the court sometimes permits to ward off a defense argument that police testimony was suspect because the officers appeared to have singled out the defendant, or that the evidence corroborated the detectives’ other testimony. Here there was no fact that could be corroborated by the officers’ having seen and spoken with Scott before.

Finally, the court found both that the district court did not engage in a full Rule 403 analysis and that, in any event, the admission of the evidence violated Rule 403.

The district court’s only comment about any possible prejudice was that the evidence would not lead the jury to conclude that Scott had been previously arrested. But that was insufficient, because Rule 404(b) covers any acts that might relate to a persons’ character, and the evidence here surely implied that Scott had had “substantial contact with the police that was not benign.” The lower court's “too narrow” view of prejudice was error. Under its own review, the circuit found “no probative value” to the evidence, and that any minimal value was “substantially outweighed by the risk of prejudice” since the testimony clearly telegraphed what Scott was “likely to be doing” - dealing drugs.

Finally, the court found that the error was not harmless. The government’s case was “not particularly strong”: the observations were made through a patrol car’s tinted windows, from a great distance, and using only the side- or rear-view mirrors. Some of the testimony was also of suspect credibility: the supposed drug sale took place in broad daylight, with no lookout, and no one noticed the obvious police vehicle nearby. Moreover, the officers did not find drugs on Scott or the alleged buyer. The evidence was also clearly important to the prosecution, which opened on it and dedicated one sixth of its summation to it. The summation was particularly prejudicial because it perpetuated the propensity argument that the officers were better able to tell what Scott was doing because, since they knew him, they knew what he was “likely to be doing.”

Labels: ,

Friday, April 06, 2012


United States v. Pitera, No. 10-1564-cr (2d Cir. April 3, 2012) (Jacobs, Miner, Katzmann, CJJ)

In this opinion by the late Judge Miner, the circuit rejected the claim of Thomas Pitera, formerly of the Bonanno crime family, that the district court erroneously rejected his application for DNA testing of “newly discovered” evidence.

Pitera was convicted in 1992 of various racketeering and CCE offenses that were predicated on seven murders, and received a life-plus-thirty-year sentence. Since then, he has regularly sought post-conviction relief. Most pertinent to the issues on this appeal is a 1999 habeas petition in which Pitera unsuccessfully alleged that evidence seized from one Frank Gangi, the main cooperator against him, would prove that Gangi was the “true killer.” The district court denied relief, noting that Gangi had admitted at Pitera’s trial that he was a participant in many of the murders, and explained that he had done them with Pitera. The habeas court also noted that Pitera’s involvement in the murders was corroborated by other evidence, including the recovery from Pitera’s home of items belonging to the victims and a recorded phone call where he discussed disposal of the bodies.

In the current litigation, Pitera returned to the claim that Gangi committed the murders of which Pitera was convicted. He filed a claim under the Innocence Protection Act, 18 U.S.C. 3600 (“the Act”), asking the government to compel DNA testing of six items of evidence allegedly seized from Gangi. Pitera alleged that a finding of either a victim’s or Gangi's DNA would raise a reasonable probability that Gangi, and not Pitera, was the murderer. The district court denied the application, and the circuit affirmed.

The circuit started by answering a question it had not yet reached - the standard of review. It concluded that whether a prisoner is entitled to DNA testing under the Act is reviewed de novo, with the related findings of fact reviewed for clear error. Here, the court had little trouble affirming. The issues Pitera raised were the same ones he raised in 1999, albeit with a “speculative forensic twist.” Even if the materials allegedly recovered from Gangi showed only DNA from Gangi and one or more victims, this would not raise a “reasonable probability that” Pitera did not commit the offense because, as noted above, the government’s theory was that Pitera and Gangi committed the murders together. The circuit again relied on Gangi’s trial testimony and the same corroborating evidence that the district court relied on in the earlier habeas litigation.


The Lyin’ King

United States v. Oyewumi, No. 10-3427(L) (2d Cir. March 29, 2012) (Wesley, Carney, CJJ, Cedarbaum, DJ)

Defendant-appellant Saeed went through the entire district court process - arrest, trial, safety-valve proffer and sentence - under the name Reginald Davis, a stolen identity. He also, according to a footnote in this opinion,tried to continue using that identity in the circuit, but the court would not permit it. It is his use of that identity that generated the most action on his appeal.

Saeed was arrested in 2009 after law enforcement agents seized a package at Newark Airport that contained 787 grams of heroin. A controlled delivery, followed by some monitored telephone calls, ultimately implicated Saeed. Saeed’s attorney told the government that Saeed might be eligible for safety-valve relief. Warning that Saeed would have to reveal his true identity, the government invited him to a proffer. Saeed attended, and continued to insist that he was Reginald Davis, although he also appeared to give a truthful account of his involvement in the drug conspiracy.
Saeed was subsequently charged with four additional crimes, all resulting from his use of a false identity during the safety-valve proffer, although two of the four were dismissed before trial. He was convicted by a jury of the narcotics conspiracy, one count of violating 18 U.S.C. § 1001, and one aggravated identity theft count. At sentencing, the government opposed safety-valve treatment because Saeed had lied about his identity in the proffer. The court agreed that this disentitled him to the safety-valve, and sentenced him to 110 months’ imprisonment.

On appeal, Saeed argued that his § 1001 conviction - which was the predicate for the aggravated identity theft count - was based on legally insufficient evidence because his false statements about his identity were not material. The circuit disagreed. “As a matter of common sense, providing a false identity to officials conducting a safety-valve proffer has both a ‘natural tendency to influence’ and is ‘capable of distracting’ those officials.” This type of deception affects the government’s ability to develop information about the crime, the defendant’s participation in it, and any relevant criminal history. The government was not required to prove that the lies actually affected those inquiries.

Saeed also argued that the government should not have been permitted to introduce his safety-valve proffer statements at trial, on the theory that the government acted in “bad faith” in continuing the proffer after it realized that he would not meet the pre-condition of disclosing his true identity. The circuit again disagreed. Saeed was informed of the government’s pre-condition for recommending safety-valve relief but nevertheless lied about his identity. The government was “under no obligation to save Saeed from himself once he failed to reveal his true identity.”

Labels: , ,