United States v. Sorto

Opinion

UNITED STATES DISTRICT COURT FOR THE DISTRICT OF COLUMBIA

UNITED STATES OF AMERICA, v.

YONAS ESHETU, et al., Case No. 13-cr-262 (CRC)

Defendants.

MEMORANDUM OPINION

Defendants Pablo Lovo, Joel Sorto, and Yonas Eshetu were tried by a jury and convicted of conspiracy to commit Hobbs Act robbery of a liquor store doubling as a drug stash house. In actuality, this stash house was fictitious: Defendants were caught red-handed in a reverse-sting operation orchestrated by the Washington, D.C. Metropolitan Police Department (“MPD”). After the jury returned its guilty verdict and following a direct appeal, Defendants now move to vacate their convictions based on claims of ineffective assistance of counsel. They contend that their trial counsel performed below constitutional standards by failing to (1) investigate and present an entrapment defense, (2) object to the admission of Spanish-language recordings capturing Defendants’ meetings with the undercover officers, and (3) seek dismissal based on selective enforcement.

The Court disagrees on all counts. For each charge of ineffective assistance, Defendants cannot satisfy Strickland v. Washington’s dual requirements of deficient performance and prejudice. Because this outcome is clear based on the record, an evidentiary hearing is not required. The Court will, accordingly, deny Defendants’ motion to vacate their convictions.

I. Background The Court draws the following factual background from the D.C. Circuit’s opinions on direct appeal, transcripts of the trial and pretrial proceedings, and other pleadings and docket entries in the case.

Back in the summer of 2012 and again in February 2013, Defendant Lovo helped his longtime friend, Jonathan Avila, obtain drugs to sell to an individual named “Santos.” See United States v. Eshetu, 863 F.3d 946, 949 (D.C. Cir. 2017) (Eshetu I). Unbeknownst to Lovo, Avila was no longer a true friend. He was now cooperating with law enforcement, and “Santos” was in reality an undercover MPD Officer named Miguel Rodriguezgil. Id. These drug transactions were the opening salvo of a reverse-sting operation designed to catch Lovo committing a crime.

The operation continued into summer 2013, but its focus shifted to a more serious transgression: conspiracy to commit an armed robbery of a liquor store and purported stash house. Id. This conspiracy came into being over the course of five meetings that summer. Id. The first took place on August 13 when Lovo and Avila broke bread with Rodriguezgil at Camino Real, a local restaurant in the Northwest quadrant of Washington, D.C. See ECF No. 191 (“Trial Day 2 Tr.”) at 166–71. At this initial encounter—which is the only one not recorded—Rodriguezgil testified that he probed Lovo’s interest in committing robberies and explained that he knew a narcotics courier who could tip them off when large quantities of drugs and cash were being stashed inside a liquor store. Id. While Rodriguezgil explained that he could use a New York-based crew to handle the job, Lovo insisted there was no need to look elsewhere. Id. at 171–74. Lovo assured him that he had a “crew” that was well-suited for the job given their extensive experience robbing brothels and illegal gambling houses and their

3

expansive stockpile of weapons. Id. at 174–76. The plotters parted ways after agreeing to reconnect in the coming days to hammer out the details. Id. at 176. Rodriguezgil then phoned Lovo two days later, and they arranged to meet the next day with the purported “drug courier”— Janice Castillo, a special agent with the United States Bureau of Tobacco, Firearms, and Explosives (“ATF”). Id. at 188–93.

This second meeting occurred on August 16 at Camino Real. Id. at 48, 55. As with all meetings that followed, this conversation was secretly recorded on both audio and video. The recording quality was poor, however, and the dialogue drifted back and forth between Spanish and English. Agent Castillo posed as a disgruntled courier for a local liquor store that trafficked large quantities of cocaine. Eshetu I, 863 F.3d at 949. Lovo actively engaged the undercover officers throughout the conversation as they plotted their scheme and, once more, reiterated that he had a crew that was well experienced in robbing brothels and gambling joints and equipped with all the weapons they might need to carry out the heist. Id.; Trial Day 2 Tr. at 73–79. Although Lovo bragged about his crew’s criminal resume, he never mentioned any member by name. Trial Day 2 Tr. at 78. The parties left the restaurant with an agreement to divide the robbery proceeds equally. Id. at 84.

After more back and forth phone communication, Rodriguezgil and Lovo agreed to meet again on August 24 at another nearby restaurant. ECF No. 192 (“Trial Day 3 Tr.”) at 54. This time, Lovo brought one of his “crew” members, Defendant Sorto. Id. It was the first time Rodriguezgil had heard of Sorto. Id. at 58–61. At the meeting, Rodriguezgil informed Lovo and Sorto that the courier had concerns about their experience and was considering using a veteran squad from New York to handle the heist, id. at 78, but Lovo again touted their bona fides and insisted they were ready for the job, id. at 65–66. Lovo explained he had a variety of firearms

4

available for the robbery, including a “TEC-9” semiautomatic pistol. Id. at 68–70. Sorto then interjected that he would be armed with a machete which, from experience, he viewed as more intimidating. Id. at 115–20. The three delved into details about the robbery’s logistics and reaffirmed their pact to share evenly in the booty. Id. at 72, 94. Rodriguezgil concluded the meeting by telling the two unwitting suspects he would be in touch once he received more information from the courier. Id. at 121–22.

After speaking over the phone several days later, Lovo and Rodriguezgil met again on September 2 so Rodriguezgil could showcase the vehicle that he planned to use for the robbery. Id. at 148–59. Meeting outside a restaurant on 14th Street, Rodriguezgil popped the trunk of his minivan and allowed Lovo to examine a secret compartment which could be used to store their guns and loot. Id. Lovo seemingly approved of the proposed get-away car and repeated that his crew had its own firearms, so Rodriguezgil need not make any arrangements to procure weapons. Id. at 158.

The fifth and final meeting took place on September 5, the day of the would-be robbery.

The parties had agreed to stage the operation from a storage unit outfitted to resemble a cocaine processing facility. Id. at 176–77. Rodriguezgil pulled up to the spot in the minivan he had previewed three days prior, and Lovo arrived in a rented Kia accompanied by Sorto, Defendant Eshetu, and two other men. ECF No. 193 (“Trial Day 4 Tr.”) at 9–10. This was the first Rodriguezgil had heard of Eshetu and the first time he learned the identities of Lovo’s other crew members. Id. at 18. Before entering the storage unit, Rodriguezgil placed a gun inside the van’s secret compartment and directed the others to do the same. Id. at 11–13. Lovo and Sorto initially responded by opening their Kia and manipulating a bag inside, but they left the bag in the trunk and did not retrieve any guns. Id. at 13–15. Lovo then told Rodriguezgil that he had

5

left the weapons inside the Kia because they might use two vehicles for the robbery, and the men entered the storage unit. Id. at 14–15.

Once inside, Raul Cruz, Jr., another member of the crew, directed everyone to lift their shirts and drop their pants to ensure they were not hiding weapons. Id. at 45. All complied, including Cruz who pulled out an eight-inch knife. Id. at 45–50. The parties then remained in the storage unit for around 40 minutes as they rehashed their agenda. Id. at 24. Rodriguezgil showed the crew members photographs of the target liquor store as well as the owner, and the co-conspirators asked about the target, the courier, and the game plan. Id. at 39–42. As on prior occasions, Rodriguezgil gave the conspirators an out, informing them they were free to leave if they did not trust him. But Defendants refused the offer in unison and opted to go forward. Id. at 50–53. Once again, Lovo reassured Rodriguezgil about their experience and restated the terms of their agreement. Id. at 56–58. Eshetu, meanwhile, began assigning responsibilities to the other men. Id. at 65. The meeting came to a surprise end, however, when the storage unit gates lifted to reveal an MPD squad standing at the ready to arrest the unsuspecting conspirators. The officers later searched the Kia and found, among other items, a TEC-9 and other firearms, wire, ammunition, magazine clips, facemasks, and two long machetes. See Eshetu I, 863 F.3d at 950; ECF No. 211 (“Trial Day 6 Tr.”) at 48–115.

The following week, a grand jury indicted all five crew members with one count of conspiring to interfere with interstate commerce by robbery (“Hobbs Act robbery”), in violation of 18 U.S.C. § 1951, and a second count of using, carrying, or possessing a firearm during a crime of violence and abetting that offense, in violation of 18 U.S.C. §§ 924(c) and 2. Eshetu I, 863 F.3d at 950. Lovo, Sorto, and Eshetu proceeded to trial while the other two members entered guilty pleas. Id. at 950 & n.3. In the run up to trial before now-retired Judge Rosemary

6

M. Collyer, the parties litigated two issues relevant here: (1) the possibility that Defendants would raise an entrapment defense and (2) the admissibility of transcripts of the government’s audio and video recordings.

Regarding the first issue, the government sought a pretrial order prohibiting Defendants from presenting an entrapment defense on the grounds that, as a matter of law, there was no inducement and each Defendant had shown a clear predisposition to engage in the robbery. See Mot. in Limine & Mem. of Law Regarding Entrapment Defense, ECF No. 59. Defense counsel successfully parried this effort, however, asserting it would be prudent for the Court to reserve judgment until Defendants resolved whether to mount such a defense. See Pretrial Hearing Tr., ECF No. 188 at 6–7. Defendants thereby preserved the right to present an entrapment defense at trial without committing to doing so. As it turned out, they eschewed that option and did not request an entrapment instruction. They instead advanced a theory that Lovo had agreed to sell firearms to Rodriguezgil but never entered a conspiracy to commit robbery.

As to the recordings, given that significant portions of the conversations were in Spanish, the government turned to a court-certified translator for assistance in producing transcripts of the surveillance audio from the recorded meetings with Defendants. In a pretrial memorandum, the government acknowledged that the translation would be a heavy lift considering that some of the recordings were “not so clear” and included background noise. Mem. on Admissibility of Trs., ECF No. 63 at 3–4. The government therefore pledged to produce the translations well in advance of trial so the parties could “work out most, if not all, discrepancies and concerns about the transcripts.” Id. It did not carry through on this promise. On the eve of trial, the government acknowledged it had not completed the translations on schedule due to funding issues and just turned over the last of the transcripts the night before. Mot. Hearing Tr., ECF No. 149 at 29, 52–

7

53, 79. Defense counsel objected to this late production and raised several concerns about the proposed transcripts. Id. at 30–39. After translators for the parties agreed it was infeasible to resolve the discrepancies before the looming start of trial three days later, Judge Collyer agreed with the defense’s proposal to “exclude the transcripts” and force the government to “go to trial . . . without the transcripts.” Id. at 55, 60, 65–67. Having scored a major victory in excluding the transcripts, trial counsel did not push their advantage to the hilt by moving to exclude the underlying recordings as well on the basis that significant portions were in Spanish and even the English sections were, at times, difficult to decipher due to their poor quality.

When trial began a few days later, the government opened its case-in-chief by calling Officer Rodriguezgil and Agent Castillo to the stand. During their testimony, the government played the video and audio recordings of their conversations with Defendants as the witnesses described the substance of the conversations. The prosecutor would play a clip, press pause, and have the witness explain his or her recollection of what transpired during the scene. The prosecutor would then unfreeze the video, and the cycle repeated. All throughout, trial counsel for Defendants did not object to the government playing the audio recordings. They instead selectively challenged the quality of the recordings and the witnesses’ recollections of what had transpired.

After the government rested its case-in-chief, Lovo testified on his own behalf and offered a “starkly different version of events.” Eshetu I, 863 F.3d at 951. Lovo contended that he first met with “Santos” (i.e., Officer Rodriguezgil) on August 13, 2013, to discuss granite work for ongoing renovations to the pawn shop where Avila and Lovo had several months earlier sold Santos cocaine. Id. During this initial meeting and a subsequent one with Agent Castillo, whom Lovo purportedly believed to be Santos’s girlfriend, Lovo acknowledged that a drug heist

8

was floated but maintained he declined the offer to participate and agreed only to sell weapons to Santos. ECF No. 195 (“Trial Day 7 Tr.”) at 74–76, 89. Consistent with this story, Lovo testified that he arrived at the storage facility on September 5 to complete the gun sale and averred that the wire and rope in his rental car’s trunk were for his granite work, not a robbery, and that he used the machetes for his landscaping projects. Id. at 95–105. During his testimony, Lovo repeatedly contested the officers’ interpretations of the recordings and offered his own account of these meetings. See, e.g., id. at 59–60.

The ten-day trial concluded with closing arguments on May 17, 2014. Lovo’s counsel used this opportunity to critique the quality of the audio recordings and the officers’ recollections of their meetings, urging the jury to credit Lovo’s contrary account. See ECF No. 197 (“Trial Day 10 Tr.”) at 27–48. Trial counsel for Sorto did much the same, calling out the audibility issues and asserting that the Spanish-language portions redounded to a game of he said/she said between Lovo and the officers. Id. at 61. The government responded to these arguments by reminding the jury that its case did not hinge on these recordings. “This case is not the United States versus audio tape number one . . . [or] audio tape number two,” the prosecutor insisted. Id. at 78. “This is United States versus Defendant Sorto, Defendant Lovo and Defendant Eshetu. It’s important to make that distinction because based on the defense counsel’s arguments just this morning, you wouldn’t think that we had witnesses who were there.” Id. at 79. The prosecutor repeated that Rodriguezgil and Castillo were present at each of the meetings and that their testimony alone was enough to carry the day. Id. (“What was it about Officer Rodriguezgil’s testimony that would cause you to doubt what he said and whether or not there was anything to corroborate his testimony. Special Agent Castillo was there [anyway] to corroborate [it].”). In short, the government explained that it was “not asking [the jury] to convict these defendants

9

based on the tape” but rather “based on the testimony.” Id. at 82. It also pointed to video evidence of Lovo scoping out a get-away vehicle and entering a storage unit three days later, where he and his crew were shown photos of the purported target and discussed plans for 40 minutes as their rental car sat outside stocked full with all the tools one might need for a robbery. Id. at 79–85.

When the Court delivered jury instructions after closing arguments, Defendants’ trial counsel did not request an instruction concerning the entrapment defense. See id. at 89–125. The Court instead instructed the jury that it was simply to decide whether the government had proven each element of the charged offenses beyond a reasonable doubt. Id. at 119–20. In doing so, Judge Collyer reminded the jurors they could not try to translate the Spanish portions of the recordings as they must consider only the testimony describing those conversations. Id. at 101. The jury returned a verdict the next day, finding all three Defendants guilty of Hobbs Act robbery and Defendants Lovo and Sorto guilty of violating 18 U.S.C. § 924(c).

Nearly ten months after the verdicts, now represented by new counsel, Lovo and Sorto filed motions for judgment of acquittal or a new trial. See Am. Mot. for Acquittal, ECF No. 207; Mot. to Adopt Lovo’s Mot. for J. of Acquittal & New Trial, ECF No. 212. In these motions, Defendants for the first time pressed that they had been unlawfully entrapped. This entrapment defense was partially predicated on what Defendants framed as newly discovered evidence— namely, telephone records showing that Lovo’s friend-turned-cooperator Jonathan Avila had called him 131 times between June 15 and August 24, 2013. See Am. Mot. for Acquittal at 8. Defendants also contended they were the victims of selective enforcement based on law enforcement’s racially skewed reverse-sting tactics involving fictitious stash houses. Id. at 22– 24.

10

At a hearing three days later, Judge Collyer denied the motions as untimely and otherwise meritless. See Post-Trial Hearing Tr., ECF No. 278 at 17–18. On the entrapment issue, Judge Collyer noted that this argument was “really a motion for ineffective assistance of counsel,” but trial counsel preserved the right to raise such a defense during the pretrial hearing when staving off the government’s motion in limine to preclude it. Id. at 19. From the fact that the issue was preserved but never asserted, Judge Collyer divined that “as a matter of strategy and legal acumen defense counsel [must have] decided it wouldn’t run, it had no legs, it was not legitimate.” Id. She went on to endorse that assessment. While she “assumed going into this that entrapment would be a defense,” Judge Collyer explained she changed her view after learning what transpired in the meetings with the undercover officers. Id. at 20. She emphasized “[t]here were multiple opportunities[,] particularly on the last day in the storage facility,” for the Defendants to back out of the robbery. Id. Despite Rodriguezgil offering an off-ramp at every juncture, each Defendant said “no, no, no, I have my knife.” Id. at 20. Any new evidence that Lovo’s “very good friend[] . . . called him twice a day for any reason under the sun” could not change this decisive fact. Id. at 19. Judge Collyer therefore proceeded to sentencing.

Defendants appealed. One of their challenges was successful, as the D.C. Circuit held that their § 924(c) convictions were invalid in light of intervening Supreme Court caselaw. See United States v. Eshetu, 898 F.3d 36, 38 (D.C. Cir. 2018) (Eshetu II). But their remaining challenges to the Hobbs Act robbery convictions were less fruitful. With respect to the admission of the Spanish-language recordings without translations, the Circuit applied the plain error standard because counsel had not raised this objection at trial and found that “the district court erred but not plainly so.” Eshetu I, 863 F.3d at 956. In doing so, it noted that portions of the recordings were in English, that trial counsel was able “to probe Rodriguezgil’s and

11

Castillo’s understanding of the recordings’ contents,” and that “the jury had before it video footage of certain critical events—among them, Lovo’s September 2 inspection of Rodriguezgil’s SUV and the September 5 storage-facility meeting.” Id. at 957. The Circuit similarly did “not think any alleged sound-quality problem with the recordings was serious enough to make their admission plain error.” Id. at 957 n.8. Turning to the allegations of ineffective assistance of counsel, the Circuit recognized that the trial court is the best forum to resolve ineffective-assistance claims in the first instance and, accordingly, remanded the matter to this Court. Id. at 957–58.1 On remand, Defendants have moved to vacate their convictions based on three ineffective-assistance claims: (1) failure to investigate or present an entrapment defense, (2) failure to object to admission of the recordings, and (3) failure to pursue a selective-enforcement claim. In support of the motion, Defendants attached a single affidavit from Defendant Sorto asserting that he asked his trial counsel to make an entrapment argument but was rebuffed by his lawyer, who allegedly insisted that he was not permitted to raise a derivative-entrapment theory or advance conflicting defenses. See Mot. to Vacate, Ex. 1 (“Sorto Aff.”) ¶¶ 6–8, 12. Defendants also request an evidentiary hearing on their claims. The government opposes the motion and has submitted affidavits from each Defendant’s trial counsel justifying each of the challenged decisions. See Opp’n, Ex. 1 (“Carroll Aff.”) (trial counsel for Sorto); id., Ex. 2 (“Kiersh Aff.”) (trial counsel for Eshetu); id., Ex. 3 (“Balarezo Aff.”) (trial counsel for Lovo). II. Legal Standard “The Sixth Amendment right to counsel in ‘all criminal prosecutions’ is the right to the effective assistance of counsel.” United States v. Burroughs, 613 F.3d 233, 238 (D.C. Cir. 2010)

1 This Court was assigned the case due to Judge Collyer’s intervening retirement.

12

(quoting Strickland v. Washington, 466 U.S. 668, 684–86 (1984)). A defendant is denied this right when (1) his attorney represents him deficiently and (2) the deficiency causes him prejudice. Id.

Under Strickland’s first prong, a defendant must show that “counsel’s representation fell below an objective standard of reasonableness.” Strickland, 466 U.S. at 688. The Supreme Court has cautioned that “[j]udicial scrutiny of counsel’s performance must be highly deferential,” and that defense attorneys are to be afforded “wide latitude . . . in making tactical decisions.” Id. at 689; see also Padilla v. Kentucky, 559 U.S. 356, 371 (2010) (“Surmounting Strickland’s high bar is never an easy task.”). Applying this highly deferential standard of review, the Court’s role is limited to determining whether a defendant’s attorney “made errors so serious that counsel was not functioning as the ‘counsel’ guaranteed to the defendant by the Sixth Amendment.” Strickland, 466 U.S. at 687.

To establish prejudice under Strickland’s second prong, the “defendant must show that there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Id. at 694. A “reasonable probability” means one “sufficient to undermine confidence in the outcome.” Id. A different outcome need not have been “more likely than not.” Id. at 693. Still, the likelihood of a contrary outcome must be “substantial, not just conceivable.” Harrington v. Richter, 562 U.S. 86, 112 (2011). And naturally, an adverse judgment with substantial evidentiary support is less likely to have been tainted by deficient performance than one with weaker grounding in the record. Strickland, 466 U.S. at 696.

A defendant’s failure to make the required showing on either Strickland prong defeats an ineffective-assistance claim, and the order of analysis is at the court’s discretion. Id. at 697. “If

13

it is easier to dispose of an ineffectiveness claim on the ground of lack of sufficient prejudice, . . . that course should be followed.” Id.

III. Analysis Based on the record and materials in the briefing, the Court finds that Defendants cannot establish ineffective assistance based on any of the three theories advanced in their motion to vacate. First, on the entrapment defense, the evidence at trial did not support an entrapment instruction and there is no reasonable probability that further inquiry would have converted this doomed defense into a viable option. This outcome is even more evident for Eshetu and Sorto because any derivative-entrapment defense was dead on arrival given that the police never targeted these two crew members individually. Second, although the trial court was deemed to have erred in admitting the recordings without translated transcripts, it is doubtful that the outcome would have been different absent these recordings considering the significant evidence showing that Defendants entered into a conspiracy to commit robbery and followed through with their plans until the very last second. Third, trial counsel did not perform deficiently by failing to raise a moonshot selective-enforcement defense, and it is vanishingly unlikely that such a defense would have succeeded in any case.

Because these outcomes are readily apparent based on the present record, the Court will deny Defendants’ motion to vacate their convictions as well as their request for an evidentiary hearing on the matter.

A. Entrapment Starting with entrapment, Defendants fault their trial counsel for failing to move for an entrapment instruction or to investigate this line of defense. The problem with this argument,

14

however, is that there is no reasonable probability that this “relatively limited defense” would have prevailed here. United States v. Russell, 411 U.S. 423, 435 (1973).

Entrapment “has two related elements: government inducement of the crime, and a lack of predisposition on the part of the defendant to engage in the criminal conduct.” United States v. Glover, 153 F.3d 749, 754 (D.C. Cir. 1998) (quoting Mathews v. United States, 485 U.S. 58, 63 (1988)). Inducement must be “such that a law-abiding citizen’s will to obey the law could have been overborne.” Id. (quoting United States v. Kelly, 748 F.2d 691, 698 (D.C. Cir. 1984)). Solicitations, “even repeated government solicitations,” do not constitute inducement “unless the requests are coupled with persuasive overtures, or unless there is evidence of reluctance on the defendant’s part demonstrating that the repetition of the requests may have moved an otherwise unwilling person to commit a criminal act.” Id. (quoting United States v. McKinley, 70 F.3d 1307, 1312 (D.C. Cir. 1995)). If the defendant carries his “initial burden of showing government inducement . . . , the burden then shifts to the government to prove the defendant was predisposed to commit the crime.” Id. “[T]he predisposition which must be shown by the prosecution is a ‘state of mind which readily responds to the opportunity furnished by the officer or his agent to commit the forbidden act for which the accused is charged.’” United States v. Burkley, 591 F.2d 903, 916 (D.C. Cir. 1978) (quoting Hansford v. United States, 303 F.2d 219, 222 (D.C. Cir. 1962) (en banc)). In other words, the government must show the defendant was “presently [r]eady and willing [t]o commit the crime.” Id. While inducement and predisposition are normally jury questions, an entrapment instruction is appropriate only where there is sufficient evidence from which a reasonable jury could find entrapment. Glover, 153 F.3d at 754.

15

Defendants may also advance a “derivative entrapment” defense where an unknowing government agent (Lovo in this case) recruits others to commit a crime. But this derivative theory is even more constrained, “extend[ing only] to cases in which unwitting intermediaries— at the government’s direction—deliver the government’s inducement to a specified third party.” United States v. Washington, 106 F.3d 983, 993 (D.C. Cir. 1997) (emphasis added). “The defense should not apply if, in response to pressure put on him by the government, the unknowing intermediary on his own induces the [third person] to engage in criminal activity.” Id. (quoting United States v. Layeni, 90 F.3d 514, 520 (D.C. Cir. 1996)). Derivative entrapment is thereby unavailable as a defense when government officials do “not designate particular” third parties “to whom the intermediaries should” direct their inducement. Id. at 995. Put simply, “a derivative entrapment theory should only apply to sting operations conducted through unwitting intermediaries if the intermediaries are directed to target specific individuals and follow these instructions.” Id. at 996.

Here, Lovo maintains that his trial counsel should have investigated and presented a direct-entrapment defense. Eshetu and Sorto meanwhile parrot that their trial attorneys should have mounted a derivative-entrapment defense because they were pulled into the scheme by Lovo, an unwitting intermediary. Neither argument passes Strickland’s demanding test.

1. Direct Entrapment: Lovo Lovo first faults his trial counsel for not moving for an entrapment instruction, which he submits was warranted based on the facts presented at trial. He contends that the trial testimony clearly demonstrated that the “government proactively induced his participation in the fake drug robbery” by concocting an artificial scheme “that minimized the danger and nature of the crime,” including by “claim[ing] to have an inside person to grease the skids.” Mot. to Vacate at 11–12.

16

Based on this record, Lovo sees no good reason why counsel “abandon[ed] the defense that they told the trial judge [they planned to pursue] on the eve of trial.” Id. at 14.

As the government points out, however, Lovo contorts the record when claiming that trial counsel abandoned the entrapment defense after informing Judge Collyer they planned to pursue it. See Opp’n at 32. Rather, the record demonstrates that defense counsel simply preserved the option of raising this defense depending on the evidence presented at trial. See Pretrial Hearing Tr. at 6–7. Lovo’s trial counsel explains in his affidavit that he ultimately concluded this defense was too risky because it conflicted with Lovo’s own account of how things transpired and, regardless, would fall flat given Lovo’s actions and statements to the officers, “which plainly indicated that Mr. Lovo was a willing participant in criminal activity.” Balarezo Aff. ¶¶ 10–12. Lovo does not refute that reasonable judgment call in any real way, and this explanation tracks Judge Collyer’s own assessment post-trial that counsel must have resolved that an entrapment defense “wouldn’t run” after watching the testimony unfold. See Post-Trial Hearing Tr. at 19.

From its own review of the record, the Court agrees that the evidence presented at trial did not support an entrapment defense. The defense would have stumbled out the gate because nothing in the trial record indicates that Lovo was impermissibly induced into the conspiracy. It is no doubt true that undercover agents recruited Lovo and that Officer Rodriguezgil initiated all conversations. See, e.g., ECF No. 194 (“Trial Day 5 Tr.”) at 90–93, 108–110. This evidence establishes only that Rodriguezgil solicited Lovo, however, and it is black-letter law that “simply offer[ing the defendant] the opportunity to” commit a crime is not inducement. Jacobson v. United States, 503 U.S. 540, 550 (1992). Lovo never pinpoints where exactly Rodriguezgil’s advances crossed the line, and the Court cannot locate any such overreach. Lovo testified at trial that Rodriguezgil never threatened him, see ECF No. 199 (“Trial Day 8 Tr.”) at 67–68, 76–77,

17

and his account of what transpired does not resemble the sort of hounding that could constitute entrapment given that, in his own telling, the drug heist was broached only during the second meeting on August 16, see id. at 37–42.2 Lovo avowed that he declined the offer and that Rodriguezgil accepted this refusal, stating “it was okay [and] that we could leave that out of the way” while proceeding with a weapons sale instead. ECF No. 196 (“Trial Day 9 Tr.”) at 57–58. It is also irrelevant that Agent Castillo posed as a disgruntled insider with intel on the target. This sort of subterfuge is par for the course in sting operations, and it is no defense to plead that the criminal opportunity was simply too good to pass up.

Even assuming Lovo somehow could show actual inducement, the record is clear as day that he was all too “ready and willing” to commit the crime. Burkley, 591 F.2d at 916. At no point during the events did Lovo ever express uneasiness with going through with the robbery. On multiple occasions the undercover officers showed him an off ramp, but he declined to take it every time. See, e.g., Trial Day 2 Tr. at 98–99; Trial Day 3 Tr. at 78–79; Trial Day 4 Tr. at 50– 53. Not only did he refuse the exit offer, Lovo reiterated time and time again that he and his crew were fit for the role because they had conducted numerous robberies of brothels and gambling establishments. See, e.g., Trial Day 2 Tr. at 57, 74–76, 81, 93, 173–74; Trial Day 3 Tr. at 103–09; Trial Day 4 Tr. at 91; Trial Day 5 Tr. at 21–22, 39. Lovo actively advertised his criminal pedigree to assuage any concerns and convince his new partners in crime that they had found the right man for the job. See, e.g., Trial Day 3 Tr. at 79 (“[H]e’s reassuring me after I give him an opportunity that they’re ready, that they have basically everything that they need in order to commit this and they’re ready.”); id. at 124 (“[Lovo’s] just trying to make me feel better

2 Lovo now claims Rodriguezgil first mentioned the robbery at the initial meeting. See Trial Day 6 Tr. at 21–22. This is consistent with Rodriguezgil’s own account, but it moves the relevant timeline up only a few days and therefore does not change the analysis.

18

by saying that he could describe how many times him and his crew have been at risk of losing their lives.”); id. at 125 (“[He is] talking about how there’s . . . [no] crew that’s more experienced in D.C.”). His aura during their meetings also was calm and collected, and he displayed none of the jitters one would expect from a person reluctantly roped into a robbery. See, e.g., Trial Day 2 Tr. at 57 (“He was very comfortable.”); id. at 71 (“He was just engaging.”); id. at 79 (showing no concern about the prospect of armed guards); id. at 89 (stating he “wanted to do a robbery like this before”); Trial Day 3 Tr. at 20 (“He was really relaxed, very confident.”); id. at 27 (noting Lovo never indicated any hesitation); id. at 113 (“He was very relaxed[,] and he was a very active participant of the conversation.”); Trial Day 4 Tr. at 55–56 (Lovo calming down Rodriguezgil in the storage unit); id. at 79 (“Lovo’s demeanor . . . is very relaxed.”); id. at 88 (taking charge of the crew on the day of the robbery). Lovo and Sorto even touted the weapons they planned to use, including the TEC-9 and machetes police later recovered from the rental car at the scene. See, e.g., Trial Day 3 Tr. at 68–70, 115; id. at 91–92 (discussing the rope he will use “to tie people up with”). With scant evidence of inducement and significant indications of predisposition, it is hard to fault trial counsel for not requesting an entrapment instruction and even harder to fathom how this decision prejudiced Lovo.

Perhaps sensing the weakness in this line of argument, Lovo contends that his counsel’s error was not just the failure to request an entrapment instruction at the close of trial but also his inadequate investigation into the defense beforehand. In support of this claim, Lovo identifies two concrete pieces of evidence that he contends a more thorough probe would have unearthed: (1) a phone log revealing that Jonathan Avila called Lovo 131 times between June 15 and August 24, 2013, Mot. to Vacate at 4–5; and (2) his “subjective vulnerabilities” which made him particularly susceptible to entrapment, including troubles with alcohol, id. at 12. The Court finds

19

that neither of these proffers sufficiently raise a likelihood that a more vetted entrapment defense would have succeeded.

Regarding the phone calls, Lovo first presented this “new” evidence to Judge Collyer in his motion for a judgment of acquittal. Judge Collyer, who presided over the trial and was well- versed with the record, was unconvinced that these dial records showed inducement. She accurately noted that the call logs “don’t tell us what was discussed, don’t tell us whether it’s just friends talking to each other,” and therefore do not support the conclusion “that this was all hounding” because the calls could have been made for “any reason under the sun.” See Post- Trial Hearing Tr. at 19, 21–22. Despite knowing this flaw in his supposed smoking gun, Lovo sheds no additional light on the contents of these conversations in his current motion. What limited information Lovo provides actually undercuts his suggestion that the calls were all efforts to coax him into agreeing to the robbery. While Lovo testified at trial that the robbery never came up until the second meeting on August 16, Trial Day 7 Tr. at 40, he now contends that the meeting on August 13 “was the first that [he] had ever heard of, or considered, a stash house robbery” but that he “did not agree to do the robbery” that day, Mot. to Vacate at 4. If this is true, the sheer volume of calls from Avila is irrelevant considering that the vast majority took place before this August 13 meeting.3 See Am. Mot. for Acquittal, Ex. D. And after August 13, Officer Rodriguezgil testified that Avila was “left out” of the picture at Lovo’s request and never appeared again at any meeting during the crucial next three weeks. Trial Day 3 Tr. at 22. Moreover, most of the conversations lasted between 30 and 120 seconds, making it even more

3 Officer Rodriguezgil testified that the focus of the investigation already had shifted from the drug sales to the stash-house heist before their first meeting. See Trial Day 6 Tr. at 33– 34. While this timeline better supports Lovo’s suggestions that these calls were somehow related to the robbery, he continues to disclaim awareness of the robbery until August 13. And regardless, Rodriguezgil also stated that he first proposed the robbery in the August 13 encounter, see id. at 21–22, 77, rendering the earlier phone calls immaterial here.

20

implausible that each call was an effort to cajole Lovo into the conspiracy. See Am. Mot. for Acquittal, Ex. D. Viewed in proper context, then, these phone calls do not seriously call into question whether Lovo’s will was overborne—especially considering that the D.C. Circuit has “never found such a plea [based on friendship] sufficiently strong” to “satisfy the inducement prong of an entrapment defense.” See United States v. Evans, 216 F.3d 80, 90 (D.C. Cir. 2000).

And even if these phone records raise the quantum of evidence supporting a finding of inducement a tick above where it stood at trial (which was nearly at nil), that still would not get Lovo to first base because nothing on the calls could negate the predisposition that he displayed throughout his meetings with the undercover officers. Judge Collyer recognized this reality when denying Lovo’s motion for a judgment of acquittal. While Judge Collyer initially assumed “that entrapment would be a defense,” she soon recognized that “entrapment will [not] do when there was such [a] voluntary, [‘]yes, I want to do this,[’] response to . . . offers to say you can walk out and not do this.” Post-Trial Hearing Tr. at 20–21. That Lovo’s friend called him twice a day was thus immaterial in Judge Collyer’s eyes because Lovo jumped on the opportunity, flaunted his purported criminal experience,4 and refused numerous opportunities to back out. Id. at 19. Concurring with that assessment, the Court fails to see how these call logs could have made a difference in this case.

Lovo’s other “new” evidence—his “subjective vulnerabilities”—fares no better. Lovo asserts that Avila knew he had a difficult upbringing, had dropped out of high school, and had a problem with alcohol. Mot. to Vacate at 12. Without citing the record, Lovo then posits that

4 Trial counsel suggested that Lovo’s boasts about his criminal history could have been mere bluster given that the police did not verify his accounts of robbing brothels and other illegal establishments. See Trial Day 2 Tr. at 122–23, 155–56. But even if this was just puffery, the fact that Lovo advertised his talents, refused invitations to leave, and instead assembled a crew which showed up to the storage facility with guns at the ready was more than enough to tip the scales against any entrapment defense.

21

Officer Rodriguezgil took advantage of these vulnerabilities when he “plied Mr. Lovo with liquor” at their meetings. Id. Such unfounded assertions do not carry the day. Regarding Avila’s purported awareness of Lovo’s difficult personal life, any claim that this knowledge aided the alleged inducement is again refuted by trial testimony from both sides indicating that Avila was uninvolved in planning the robbery, aside from introducing Lovo to “Santos.” Nor does Lovo’s troubled upbringing negate the overwhelming predisposition evidence recounted above. The only newly proffered evidence that even touches on predisposition is the charge that Rodriguezgil used alcohol to encourage Lovo to agree to the robbery. But this bald allegation is not only unfounded; it is contradicted by the record. Lovo contends the officers held the meetings in restaurants to induce him into drinking as part of a ruse to cloud his judgment. Id. at 13–14. Rather than drawn-out drinking sessions, however, trial testimony (and video evidence for the latter two) show that each of the first three meetings in restaurants lasted a breezy 30 minutes. See Trial Day 2 Tr. at 57, 174; Trial Day 3 Tr. at 62. While Rodriguezgil ordered a beer at one meeting, see Trial Day 5 Tr. at 104, this is a far cry from plying Lovo with alcohol. There was also no alcohol at the fourth meeting when Rodriguezgil showed Lovo his vehicle’s hidden compartment in a parking lot. See Trial Day 3 Tr. at 148–59. The only time hard alcohol appeared at trial was inside the storage unit on the fifth and final meeting when Rodriguezgil told the crew he had Coronas and Hennessy (as well as pizza) available. See Trial Day 4 Tr. at 33, 64, 75–76 (Rodriguezgil stating that he and Cruz shared a shot of Hennessy at Cruz’s request). Lovo also testified that he was offered and had a beer in the storage unit. See Trial Day 7 Tr. at 144. But, once again, nothing suggests that Lovo was plied with alcohol during the 40 minutes he spent in the storage unit before his arrest. See Trial Day 4 Tr. at 24. And regardless, Lovo

22

already had agreed to the robbery, brought the weapons, and recruited his crew. The die was cast by this point.

In sum, even with the “new” evidence, the record conclusively shows that Lovo was not induced into the robbery and that he was predisposed to committing the crime. Any decision not to investigate or raise an entrapment defense therefore did not prejudice Lovo because, as Judge Collyer noted, such an argument “had no legs” and “wouldn’t run” from the jump. Post-Trial Hearing Tr. at 19.

2. Derivative Entrapment: Eshetu & Sorto Turning to Defendants Sorto and Eshetu, their arguments that trial counsel should have sought a derivative-entrapment instruction fails at the threshold because there is no evidence that Officer Rodriguezgil or Agent Castillo targeted them specifically.

As noted above, the “limited” defense of derivative entrapment applies only when an intermediary “deliver[s] the government’s inducement to a specified third party.” Washington, 106 F.3d at 993 (emphasis added). A “general instruction by a government agent to an intermediary” to recruit others “is not sufficient to support a derivative entrapment defense.” Id. at 996. The record in this case shows that neither Rodriguezgil nor Castillo targeted Sorto or Eshetu in any manner. At trial, Rodriguezgil repeatedly testified that he never heard of Sorto or Eshetu before Lovo independently brought them to the third and fifth meetings, respectively. Trial Day 3 Tr. at 58–59, 65, 176; Trial Day 4 Tr. at 18; Trial Day 5 Tr. at 115–16, 194–95, 213. Similarly, Castillo testified that while she asked to meet “the rest of [Lovo’s] crew members” during the second meeting, Trial Day 2 Tr. at 127, she did not know any crew member’s identity and Lovo never mentioned Sorto or Eshetu by name, see id. at 112–13, 127–28. This unrebutted testimony, reaffirmed under cross-examination, demonstrates that Rodriguezgil and Castillo did

23

no more than instruct Lovo to assemble a “crew” for the job. This general request is indistinguishable from the unwitting intermediary in Washington who was asked to recruit other “dirty” cops. There, the Circuit held that this instruction did not support a derivative-entrapment defense. Washington, 106 F.3d at 995–96. The same is true here. By gathering his “crew” himself, Lovo “acted independently—that is, on his own—in deciding who would be [the] targets of [Rodriguezgil’s] offer.” Id. at 996. Such independent action cannot serve as the basis for a derivative-entrapment defense.

Sorto’s and Eshetu’s trial counsel thus did not perform deficiently by failing to raise a doomed defense, and Defendants were not prejudiced by this decision.5 B. Admission of Spanish-Language Recordings Defendants next maintain that they received ineffective assistance because trial counsel failed to object to admitting the Spanish-language recordings of the meetings with Rodriguezgil and Castillo. See Mot. to Vacate at 15–17. Defendants reason that these recordings prejudiced them because, without a translated transcript, “defense counsel was unable to cross-examine the government witnesses as to what was on the tapes” and, as a result, “the jury necessarily was left with the impression that . . . the government witnesses’ testimony as to the contents of the tapes had to be the truth.” Id. at 16. “The only option the jury had,” they assert, “was to accept what the government witnesses testified was on the tapes.” Id. The Court disagrees. Even assuming trial counsel fell down on the job by not objecting to these recordings, this error did not prejudice Defendants given the multiple avenues to probe the officers’ accounts and the copious evidence against Defendants.

5 Because they were not specifically targeted, the proffered evidence concerning Sorto’s and Eshetu’s personal histories of depression, drinking, or drug use is irrelevant. Mot. to Vacate at 12–13.

24

Before getting to prejudice, a word on deficient performance. The government contests Defendants’ initial premise that trial counsel’s failure to object to the recordings constituted ineffective assistance. Attached to its opposition are declarations from each Defendant’s trial counsel explaining that they made a tactical decision to use the recordings (absent the transcripts) to their advantage by pointing to the recordings’ poor quality and interpretative ambiguities to argue that the testifying officers either misinterpreted, misremembered, or misrepresented the statements in the audio. Opp’n at 34–35; Carroll Aff. ¶ 15; Kiersh Aff. ¶ 15; Balarezo Aff. ¶ 9. The record supports these explanations. All throughout, trial counsel called out the recordings’ poor quality and the officers’ purported lack of proficiency in Spanish to cast doubt on their accounts, arguing that the audibility issues made their testimony unreliable. See, e.g., Trial Day 2 Tr. at 110–11, 115–17; Trial Day 3 Tr. at 99; Trial Day 5 Tr. at 53–55, 107–08. They returned to this theme during closing arguments, asking the jury “who’s more believable . . . [a] guy who’s admitting to three crimes under oath . . . [o]r police officers who are giving you summaries of what happened from a video and a tape that you can’t even hear.” Trial Day 10 Tr. at 42; see also id. at 41–42, 61, 68. This sustained attack was sufficiently effective that the prosecution felt compelled to counteract it by reminding the jury that “[t]his case [was] not the United States versus [the] audio tape[s]” and redirecting attention to the officers’ first-hand recollections of these meetings. Id. at 78–79.

Defendants counter that it would have been preferrable to exclude the recordings in the first place, leaving the prosecution with only their fallback, i.e., contested recollections of events that transpired many months ago without any external verification. See Reply at 10. This line of Monday morning quarterbacking may well be justified. But whether any potential tactical error here raises a Sixth Amendment issue is another question entirely. The Supreme Court repeatedly

25

has cautioned that “it is all too tempting for a defendant to second-guess counsel’s assistance after conviction or adverse sentence” and has directed courts to “indulge a strong presumption that counsel’s conduct falls within the wide range of reasonable professional assistance.” Strickland, 466 U.S. at 689. The Court doubts that trial counsel’s decision not to object to the recordings rises to the sort of grave error necessary to overcome this strong presumption. Regardless, the Court need not resolve this issue because—even assuming deficiency— Defendants cannot satisfy the flip side of the Strickland inquiry: prejudice.

Based on the record in this case, there is no “reasonable probability” that the exclusion of the recordings would have altered the outcome here. Id. at 694. The Circuit strongly suggested as much on direct appeal when holding that admitting the Spanish-language audio recordings, while a legal error, was not a “plain error.” See Eshetu I, 863 F.3d at 956–57. “Pursuant to plain error review,” it explained, “an appellant must demonstrate (1) that there was an error, (2) that the error was clear or obvious, (3) that it affected the appellant’s substantial rights, and (4) that it seriously affected the fairness, integrity, or public reputation of the judicial proceedings.” Id. at 956 (quoting United States v. Gooch, 665 F.3d 1318, 1332 (D.C. Cir. 2012)). While Defendants had met the first requirement of proving an error, the Circuit held that this error did not meet “the stringent plain-error standard.” Id. at 957. Contrary to Defendants’ current assertions that the recordings prevented them from adequately cross-examining the officers, the Circuit found that trial counsel was able “to probe Rodriguezgil’s and Castillo’s understanding of the recordings’ contents,” aided by the fact that “numerous Spanish speakers testified.” Id. It also observed that “the meetings were not conducted exclusively in Spanish,” as “portions were in English,” and that the jury “had before it video footage of certain critical events—among them, Lovo’s September 2 inspection of Rodriguezgil’s SUV and the September 5 storage-facility meeting.”

26

Id. The Circuit was not entirely clear as to which plain-error factor these considerations mapped onto, yet these comments appear to speak directly to whether admitting these recordings violated Defendants’ “substantial rights.” An error affects substantial rights when there is “a reasonable probability that, but for the error, the outcome of the proceeding would have been different,” United States v. Reynoso, 38 F.4th 1083, 1092 (D.C. Cir. 2022) (citation omitted), a standard which mirrors Strickland’s prejudice prong. Thus, to the extent the Circuit’s finding of no plain error was rooted in a finding that Defendants suffered no violation of their substantial rights, any argument that admitting the recordings prejudiced Defendants is squarely foreclosed.6 Even if not outright precluded, the Court cannot find prejudice here for many of the same reasons the Circuit highlighted. First, Defendants’ assertion that admission of the recordings left them “no way . . . to counter the evidence” is at odds with the trial record. Mot. to Vacate at 16. As described above, the record shows that trial counsel routinely challenged Rodriguezgil’s and Castillo’s descriptions of the conversations by drawing attention to the poor audio quality and questioning the officers’ fluency in Spanish. Lovo also gave his own account of what transpired during both direct and cross-examination, based on his recollection as well as his interpretation of the recordings. See Trial Day 7 Tr. at 47–51, 60–63, 72, 136; Trial Day 8 Tr. at 71, 82–85, 89, 96–99; Trial Day 9 Tr. at 50. There were thus multiple ways that the defense could and, in fact, did counter the government’s narrative. Second, the recordings were not entirely in Spanish. Some portions were in English, including segments where Lovo discussed plans to tie up the courier during the robbery, see Trial Day 8 Tr. at 99, and bragged that his crew had done “this shit too many times,” see id. at 82. These statements required no translation, and a more

6 It is worth noting that, consistent with its customary practice, the Circuit nonetheless remanded the ineffective-assistance claim, commenting that “the extent of resultant prejudice” from the tapes’ admission, “if any,” was not “clear from the record.” Eshetu I, 863 F.3d at 958.

27

selective challenge based on the poor audio quality was unlikely to rule out all incriminating statements. Third, and most importantly, there was overwhelming evidence beyond the audio recordings. Defendants were caught in the act in the police storage unit with weapons, wire, and masks stashed in their rental car outside. Their only defense was that Lovo had agreed to sell the officers guns but, instead of charging him with this serious offense, the police opted to fabricate a story that he had agreed to rob a fictitious liquor store. As implausible as this defense sounds on its own terms, it was refuted by the video evidence of Lovo scoping out the get-away vehicle and of all Defendants examining photographs of the fictional robbery target while waiting for 40 minutes inside the storage unit. To top it off, Rodriguezgil and Castillo still could have testified about the conversations from memory even if the recordings were excluded—and, to the extent they did not recall conversations, presumably could have refreshed their recollections with the recordings. On this record, it is implausible that any marginal boost to the officers’ credibility that the recordings supplied made a difference in the verdict.

C. Selective Enforcement Defendants’ final claim of ineffective assistance is that trial counsel failed to investigate or present a defense that this reverse-sting case was tainted by selective enforcement because “ATF and local police” reverse-sting operations involving Hobbs Act robbery of fictitious stash houses are generally directed at “minorities.” See Mot. to Vacate at 18–20. The Constitution does not compel counsel to pursue long-shot legal defenses, however, and this theory would not have appeared viable at the time of the trial nearly a decade ago.

To sustain a selective-enforcement or prosecution claim under the Equal Protection Clause, defendants must prove “(1) [that they were] singled out for prosecution [or enforcement] from among others similarly situated and (2) that [the challenged action] was improperly

28

motivated, i.e., based on race, religion or another arbitrary classification.” United States v. Washington, 705 F.2d 489, 494 (D.C. Cir. 1983). Parties are “similarly situated” for purposes of a selective-enforcement claim when there are no legitimate factors that might justify making different law-enforcement decisions with respect to them. See Mahoney v. United States Capitol Police Bd., 566 F. Supp. 3d 1, 14 (D.D.C. 2022). Discovery is almost always essential to make out a selective-enforcement claim, and it poses a high hurdle. For many years, courts in this District and all others carried over the standard for discovery in selective-prosecution cases that the Supreme Court established in United States v. Armstrong, 517 U.S. 456 (1996), to selective- enforcement claims challenging police action. See, e.g., United States. v. Dixon, 486 F. Supp. 2d 40, 44 (D.D.C. 2007) (applying Armstrong to a selective-enforcement claim ). Under the Armstrong test, a defendant must proffer “some evidence” of both discriminatory effect and discriminatory intent before a court will order discovery—a task that, in part, requires defendants to come forward with evidence that similarly situated persons were intentionally treated differently. Armstrong, 517 U.S. at 468–69. Recognizing this was a near impossible task in most cases and citing to the absence of “special solicitude shown to prosecutorial discretion,” some jurisdictions recently have lowered the standard for discovery in selective-enforcement cases by jettisoning the “similarly situated” requirement for the discriminatory-effect prong and abandoning the discriminatory-intent prong entirely. See, e.g., United States v. Washington, 869 F.3d 193, 220–21 (3d Cir. 2017). Under this new test, a party seeking discovery must show only “some evidence” of “discriminatory effect” in the form of a factual proffer “contain[ing] reliable statistical evidence, or its equivalent,” demonstrating enforcement disparities. Id.; accord United States v. Lopez, 415 F. Supp. 3d 422, 427 (S.D.N.Y. 2019) (holding that “a defendant who is a member of a protected group” can access discovery if he can “show that that group has been

29

singled out for reverse sting operations to a statistically significant extent in comparison with other groups”). Regardless of the standard for discovery, however, ultimate success on a selective-enforcement claim still requires proving that the challenged law enforcement action had both discriminatory effect and discriminatory purpose. See Dixon, 486 F. Supp. 2d at 45.

Defendants fault their trial counsel for failing to investigate, seek discovery on, or seek dismissal of the case based upon the Government’s selective enforcement of reverse-sting cases against non-white persons. See Mot. to Vacate at 18–20. In support of this argument, they point to three pieces of evidence which, in their view, tend to show discriminatory effect: (1) two cases from this District—United States v. Hopkins, No. 13-cr-109-ESH, and United States v. Green, 13-cr-208-RBW—in which most or all of the defendants in reverse stings were Black; (2) two academic articles discussing racial disparities in stash house stings; and (3) recent cases from other jurisdictions detailing racial disparities in these operations and granting discovery on the matter. Although the evidence from this jurisdiction is slim, it is plausible that the statistical evidence showcasing disparities across the country might be enough to warrant discovery today in a jurisdiction that has abandoned Armstrong for selective-enforcement claims in favor of the less demanding test. Whether this would have been enough for Defendants to access discovery back in 2014 (or even today in this District)—and, even more importantly, whether this defense ultimately would have succeeded—is far less likely.

Perhaps the strongest evidence that Defendants have presented tending to show an issue with selective enforcement is a 2020 law review article citing studies across several jurisdictions finding significant racial disparities in reverse-sting operations involving stash-house robberies. See Alison Siegler & William Admussen, Discovering Racial Discrimination by the Police, 115 Nw. U. L. Rev. 987, 990–91 & nn.10–12 (2020). While this article might support a request for

30

discovery now, it was published in 2020—six years after trial in this case—and the sources cited by the article for the statistical disparities likewise post-date Defendants’ trial. That is not to say there was no evidence of racial disparities in these operations before 2014. Indeed, Defendants cite to another article published in October 2014, just a few months after their trial, detailing the issues with selective enforcement in stash-house stings and describing developments in this area in the Seventh Circuit, see Katharine Tinto & Kathryn O. Greenberg, Fighting the Stash House Sting, Champion 16, 17–18 (Oct. 2014), as well as a Ninth Circuit dissent discussing selective enforcement in ATF reverse-sting operations a mere two weeks before their trial, see United States v. Black, 750 F.3d 1053 (9th Cir. 2014) (Reinhardt & Kozinski, JJ., dissenting from denial of rehearing en banc). However, these sources also highlight the novelty of these types of claims and do not provide good reason to believe that these disparities were so widely known that trial counsel should have been on notice of them. Matter of fact, the 2014 article describes selective- enforcement challenges to reverse-sting operations involving stash houses as “an exciting area of legal and factual development.” Tinto & Greenberg, supra at 18.

Putting this timing issue to the side for the moment and assuming trial counsel could have presented some evidence of discriminatory effects in stash-house reverse stings, this alone would not have been enough for Defendants to access discovery on this matter. As noted above, prior to 2015, courts uniformly required defendants to satisfy the demanding Armstrong standard for accessing discovery in selective-enforcement claims. See Sigler & Admussen, supra at 1005–06. To obtain discovery under this standard, defendants must present some evidence tending to show both discriminatory effect and discriminatory intent—a showing that requires identification of similarly situated individuals outside a defendant’s racial group who could have been targeted for enforcement but were not. Id. at 1000–01, 1005–06. The Seventh Circuit first broke ranks in

31

2015 when it lowered the discovery standard for selective-enforcement claims in United States v. Davis, 793 F.3d 712 (7th Cir. 2015) (en banc). The Third and Ninth Circuits followed suit in 2017 and 2018, respectively. United States v. Washington, 869 F.3d 193 (3d Cir. 2017); United States v. Sellers, 906 F.3d 848 (9th Cir. 2018). Notably, all these deviations from the norm occurred after Defendants’ trial, and their counsel would have had no way of forecasting these future changes in the law.7 Indeed, the law in this District has not changed, as none of these cases have been adopted (or even cited) by courts in the District of Columbia. Defense counsel therefore reasonably would have assumed that the Armstrong standard would have governed discovery in any selective-enforcement claim. This standard is exceedingly onerous, as evidenced by the materials Defendants cite in their briefs. The 2020 law review article describes Armstrong as an “[i]nsurmountable [b]arrier” to discovery, Sigler & Admussen, supra at 1016, adding that in “only a handful of cases have defendants even been able to meet Armstrong and gain discovery,” id. at 1003. The fact that courts in other jurisdictions have abandoned Armstrong and now have permitted defendants to conduct discovery in some such cases is thus immaterial to the inquiry here. What matters is that, under the prevailing standard in 2014 which has never been ditched in this District even today, Defendants faced an almost impossible task just to access discovery. See United States v. Mason, 774 F.3d 824, 829 (4th Cir. 2014)

7 To be sure, the discovery order underlying the Seventh Circuit’s decision in United States v. Davis was issued several months prior to trial in this case. See United States v. Davis, 13-cr-63 (N.D. Ill. Oct. 30, 2013) (ECF No. 124). However, that case purported to apply the Armstrong standard (even if, in actuality, it only paid lip service to the rule). See id. at 2. Indeed, the district court there was obligated to do so because, as other courts in the district recognized around the same time, the Seventh Circuit had held that the “same analysis governs both [selective prosecution and selective enforcement] claims: a defendant seeking discovery on a selective enforcement claim must meet the same ‘ordinary equal protection standards’ that Armstrong outlines for selective prosecution claims.’” United States v. Alexander, No. 11-cr- 148, 2013 WL 6491476, at *3 (N.D. Ill. Dec. 10, 2013) (quoting United States v. Barlow, 310 F.3d 1007, 1010 (7th Cir. 2002)) (alternations in original).

32

(rejecting a similar claim and noting that the Sixth Amendment does not require counsel to pursue “novel” arguments).

And discovery is only the first step on this uphill march. No matter what standard applies to discovery, the party pursuing a selective-enforcement claim under the Fifth Amendment or Equal Protection Clause always has the burden of proving discriminatory effect and discriminatory intent. See Frederick Douglass Foundation, Inc. v. District of Columbia, 82 F.4th 1122, 1147 (D.C. Cir. 2023) (holding that while selective-enforcement claims under the First Amendment do not require showing bad motive, a defendant making a selective-enforcement claim under the Fifth Amendment or Equal Protection Clause must show discriminatory intent). This is no easy feat. The 2020 law review article asserts that “[s]ince the Court established Armstrong’s demanding discovery standard, there has not been a single successful selective prosecution or selective law enforcement claim on the merits.” Sigler & Admussen, supra at 1002. Even if there are some cases that deviate from the norm, the point remains intact: Defendants would have had to break new ground or hoe an exceedingly tough row to succeed on this defense.

“[A]n attorney may not be faulted for a reasonable miscalculation or lack of foresight or for failing to prepare for what appear to be remote possibilities.” Harrington, 562 U.S. at 110. With the deck so stacked against a selective-enforcement defense, other courts have held that defendants cannot ground an ineffective assistance based on counsel’s failure to mount this “long-shot” defense. See, e.g., Mason, 774 F.3d at 829–30. So too here. Because trial counsel’s failure to mount this defense neither constituted deficient performance nor prejudiced Defendants, the Court finds no Sixth Amendment violation.

33

D. Evidentiary Hearing In addition to asking the Court to vacate their sentences, Defendants request an evidentiary hearing on the matter. Yet such hearings are not required to decide a motion to vacate, and one is not needed here because the outcome is clear based on the existing record.

A district court need not conduct an evidentiary hearing before denying a § 2255 motion when “the motion and the files and records of the case conclusively show that the prisoner is entitled to no relief.” 28 U.S.C. § 2255(b). “[I]f it plainly appears from the face of the motion and any annexed exhibits and the prior proceedings in the case that the movant is not entitled to relief in the district court, the judge shall make an order for its summary dismissal.” United States v. Morrison, 98 F.3d 619, 625 (D.C. Cir. 1996) (citation omitted). A defendant whose allegations are “vague, conclusory, or palpably incredible,” rather than “detailed and specific,” is not entitled to an evidentiary hearing. Machibroda v. United States, 368 U.S. 487, 495 (1962); see also United States v. Pollard, 959 F.2d 1011, 1031 (D.C. Cir. 1992). “The party seeking a hearing in a [§] 2255 proceeding carries a fairly high burden of demonstrating need for a such a hearing, and the decision whether to grant one is ‘committed to the district court's discretion.’” United States v. Wilson, No. 96-cr-319-01 (CKK), 2019 WL 4990752, at *14 (D.D.C. Oct. 7, 2019) (quoting Pollard, 959 F.2d at 1031). These rules apply with full force to claims of ineffective assistance of counsel. See Morrison, 98 F.3d at 626 (“[S]ummary denial of a § 2255 motion is appropriate when the ineffective assistance claim is speculative.”).

In this case, the Court finds that the resolution of the three ineffective-assistance claims is clear based on the record and that Defendants’ vague suggestions that there are further details they wish to uncover during a hearing do not entitle them to one. See, e.g., Mot to Vacate at 15. Defendants’ assertions that a hearing is needed to probe trial counsel’s affidavits claiming that

34

their decisions not to proceed with these defenses was based on strategy—rather than mere oversight or neglect—also miss the mark. These explanations are well-supported by the record and, no matter the actual rationale, none of these purported errors prejudiced Defendants. No probing of counsel’s accounts will change that determinative fact, and all claims therefore can be dismissed without a hearing. See Morrison, 98 F.3d at 626 (holding that ineffective-assistance claim “did not give rise to the need for an evidentiary hearing” where the claim did “not necessitate the consideration of any information not within the record” and trial counsel’s affidavit describing his trial strategy was well supported by the record). IV. Conclusion The Court will, accordingly, deny Defendants’ motion to vacate their sentences. A separate Order follows.

CHRISTOPHER R. COOPER

United States District Judge

Date: November 8, 2023