United States v. Morosco

822 F.3d 1, 2016 U.S. App. LEXIS 8753, 2016 WL 2756000

• Court: Court of Appeals for the First Circuit
• Decided: May 12, 2016
• Docket: 15-1802P
• Status: Published

Opinion

THOMPSON, Circuit Judge.

Stage-Setting

Years back, Michael McLaughlin, James Fitzpatrick, and Bernard Morosco worked for the Chelsea Housing Authority (“CHA”), a public agency principally responsible for providing low-income housing in Chelsea, Massachusetts. McLaughlin served as CHA’s executive director, Fitzpatrick as CHA’s director of modernization, and Morosco as CHA’s paid consultant.

The federal Department of Housing and Urban Development (“HUD”) funds three of CHA’s properties — properties that have a combined total of about 350 housing units. As required by regulation, HUD periodically inspects a randomly-selected, “statistically valid sample of [ ] units” to help ensure that CHA’s federally-funded housing is “decent, safe, sanitary ... and in good repair.” See 24 C.F.R. §§ 902.22(e), 902.20(a). The Real Estate Assessment Center (“REAC”) — an agency within HUD — performs these evaluations, though it usually has REAC-trained independent contractors do the inspecting. Getting a high inspection score (90 or above) meant CHA would be considered a “high performer,” which meant fewer inspections (every two years rather than every year), less oversight, and more capital funding (a 3% annual increase). And CHA got designated a “high performer” in three consecutive inspections — in 2007, 2009, and 2011.

But not all was right at CHA, it turns out. McLaughlin abruptly resigned his post in 2011 after a newspaper reported that he made about $360,000 a year, even though he told state officials that he made $160,000. As he left, McLaughlin wrote himself checks from CHA’s account for $200,000, supposedly for unused leave— talk about throwing gasoline on a fire!

McLaughlin’s salary scandal sparked a criminal investigation that led agents to Vitus Shum, CHA’s finance director. Shum copped to helping McLaughlin with the salary scheme. Receiving immunity, Shum also later told agents about how he and others at CHA had rigged the HUD inspections. And his revelations helped a grand jury indict McLaughlin, Fitzpatrick, and Morosco for “knowingly and unlawfully” conspiring to defraud the United States and its agency, HUD — a violation of 18 U.S.C. § 371, which makes it a crime for “two or more persons [to] conspire either to commit any offense against the United States, or to defraud the United States, or any agency thereof in any manner or for any purpose.” As for the indictment’s allegations, all you need to know is this: Morosco was a REAC-inspection consultant — though he principally advised housing authorities on how to handle the REAC-inspection process. And using his REAC-inspector status, he (the indictment added) accessed the REAC database and software, figured out the sample of CHA units to be inspected, and passed the information on to Fitzpatrick, McLaughlin, or both — allowing CHA employees to get those units up to snuff before the inspectors came a-ealling.

McLaughlin pleaded guilty and got a 12-month prison sentence and a $3,000 fine, on top of the 36 months he previously got for pleading guilty to charges stemming from his salary chicanery. He did not testify at Fitzpatrick and Morbsco’s seven-day trial — Fitzpatrick did, but Morosco did not. A jury found them guilty as charged. And a judge later sentenced Fitzpatrick to 3 months in prison, plus 1 year of supervised release, and Morosco to 6 months in prison, followed by 1 year of supervised release.

Fitzpatrick and Morosco now appeal. Between them, they raise a battery of arguments — though not every one requires a lot of analysis. To make the opinion easier to follow, we organize our discussion thematically, issue-by-issue, providing more background as needed. And — spoiler alert — after working through their claims, we affirm.

Void-for-Vagueness Claim

Fitzpatrick and Morosco complain that section 371’s defraud clause— criminalizing any conspiracy “to defraud the United States, or any agency thereof in any manner or for any purpose” — is unconstitutionally vague as applied to them. For those not in the know, a law is unconstitutionally vague if it fails to give ordinary people fair notice of what is forbidden, or if it fails to give the designated enforcers (police, prosecutors, judges, and juries) explicit standards (thus creating a risk of arbitrary enforcement). See Welch v. United States, — U.S.-, 136 S.Ct. 1257,-, 194 L.Ed.2d 387, at *3 (2016). Of course the requisite fair warning can come from judicial decisions construing the law. See, e.g., United States v. Lanier, 520 U.S. 259, 266, 117 S.Ct. 1219, 137 L.Ed.2d 432 (1997). And judges have no business junking a statute simply because we could have written it “with greater precision.” Rose v. Locke, 423 U.S. 48, 49, 96 S.Ct. 243, 46 L.Ed.2d 185 (1975).

Helpfully, both sides agree — rightly— that Fitzpatrick and Morosco preserved their vagueness claim below (via a motion to dismiss the indictment) and that our review is de novo. See, e.g., United States v. Hussein, 351 F.3d 9, 14 (1st Cir.2003). Also helpfully, both sides concede that binding precedent squarely forecloses this claim. 1 And we second that assessment.

Start with Fitzpatrick’s and Mor-osco’s most loudly trumpeted point. As they tell it, section 371’s “defraud” clause only bans conspiracies to deprive the government of property and money by dishonest schemes, a reading (they add) that jibes with the common-law understanding of “defraud.” And such a reading would help them (they continue) because they never scammed the government out of property or money. Unhappily for them, years’ worth of Supreme Court precedent holds that section 371 “is not confined to fraud as that term has been defined in the common law,” see Dennis v. United States, 384 U.S. 855, 861, 86 S.Ct. 1840, 16 L.Ed.2d 973 (1966); that defrauding the government under section 371 means obstructing the operation of any government agency by any “deceit, craft or trickery, or at least by means that are dishonest,” see Hammerschmidt v. United States, 265 U.S. 182, 188, 44 S.Ct. 511, 68 L.Ed. 968 (1924); and that the conspiracies need not aim to deprive the government of property or money, see id., because the act is written “broad enough ... to include any conspiracy for the purpose of impairing, obstructing, or defeating the lawful function of any” government “department,” see Haas v. Henkel, 216 U.S. 462, 479, 30 S.Ct. 249, 54 L.Ed. 569 (1910). Ever faithful to high-Court holding, our caselaw rejects the idea that section 371 only bars conspiracies to defraud the government out of property or money. See United States v. Barker Steel Co., 985 F.2d 1123, 1136 (1st Cir.1993) (relying on Supreme-Court cases interpreting section 371 and its basically “similar predecessors”); Curley v. United States, 130 F. 1, 6-10 (1st Cir.1904) (explaining that “defraud” in section 371’s forerunner has a broader meaning than the common-law definition — and justifiably so because the statute’s aim is to protect the government, and deceit can impair the workings of government even if the conspiracy does not take the government’s property or money). Obviously then, this facet of Fitzpatrick’s and Morosco’s vagueness thesis goes nowhere.

Undaunted, Fitzpatrick and Moros-co also suggest that because no statute or regulation criminalizes receiving a list of sample units before any HUD inspection, the government could not prosecute them under section 371. But our cases take all the wind out of their sails, holding as they do “that lawful activity may furnish the basis for a” section-371 conspiracy conviction. See United States v. Hurley, 957 F.2d 1, 4 (1st Cir.1992) (finding unconvincing “defendants’ asserted lack of ‘fair warning’ that their ‘legal’ conduct could be the basis for a criminal prosecution,” noting that “[t]he statutory prohibition against defrauding the government adequately put defendants on notice that a scheme designed to frustrate tax collection was prohibited”); accord Barker Steel Co., 985 F.2d at 1131 (emphasizing that section 371 bans both “(1) conspiracies to commit a specific offense against the United States, included elsewhere in the criminal code, and (2) conspiracies to defraud the United States,” and rejecting defendants’ argument “that if no other federal law or regulation proscribes alleged conduct, then [they] cannot be held criminally responsible pursuant to § 371” — “[i]f the second clause were interpreted to require commission of a specific offense, it would have the same meaning as the first clause thus rendering the second clause redundant”); United States v. Tarvers, 833 F.2d 1068, 1075 (1st Cir.1987) (stressing that section 371 “does not require that the means used to achieve the unlawful goal of the conspiracy be unlawful”). So this aspect of Fitzpatrick’s and Morosco’s vagueness theory also goes nowhere.

In what is basically a Hail Mary pass, Morosco argues that two fairly recent cases signal a new willingness on the high Court’s part to entertain vagueness challenges — a willingness (the argument goes) that we must emulate. The two cases are (1) Skilling v. United States, 561 U.S. 358, 130 S.Ct. 2896, 177 L.Ed.2d 619 (2010), limiting “honest services” fraud so that it only applies to defendants involved in either bribery or kickback schemes, and (2) Johnson v. United States, — U.S.-, 135 S.Ct. 2551, 192 L.Ed.2d 569 (2015), declaring the Armed Career Criminal Act’s residual clause — a provision dealing with crimes that “involve[] conduct that presents a serious potential risk of physical injury” — too vague to be enforced. His pass falls incomplete, however, and for a simple reason. Neither Skilling nor Johnson overruled the Haas/Hammerschmidt line of section-371 cases. And because overruling Supreme Court precedent is the Court’s job, not ours, we must follow Haas/Hammerschmidt, etc. until the Court specifically tells us not to — something that is true even if these long-on-the-books cases are in tension with Skilling and Johnson (and we do not suggest that they are). See Hohn v. United States, 524 U.S. 236, 252-53, 118 S.Ct. 1969, 141 L.Ed.2d 242 (1998) (declaring that Supreme Court “decisions remain binding precedent until [the Court] see[s] fit to reconsider them, regardless of whether subsequent cases have raised doubts about their continuing vitality”); Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S. 477, 484, 109 S.Ct. 1917, 104 L.Ed.2d 526 (1989) (instructing that “[i]f a precedent of this Court has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, the Court of Appeals should follow the case which directly controls, leaving to this Court the prerogative of overruling its own decisions”); see also United States v. Coplan, 703 F.3d 46, 61-62 (2d Cir.2012) (rejecting the idea that a circuit court should use Skilling to rework controlling section-371 precedent, noting that lower courts should leave any reworking to the Supreme Court); Specter Motor Serv. v. Walsh, 139 F.2d 809, 823 (2d Cir.1943) (L.Hand, J., dissenting) (cautioning lower courts against “embracing] the exhilarating opportunity of anticipating” the overruling of a Supreme Court decision), vacated sub nom. Spector Motor Serv. v. McLaughlin, 323 U.S. 101, 65 S.Ct. 152, 89 L.Ed. 101 (1944).

With the vagueness issue out of the way, we press on.

Insufficient-Evidence Claim

Basically 1 page of Morosco’s 68-page brief contains an attack on the judge’s decision not to acquit him because of insufficient evidence. The parties correctly agree that he preserved the issue for appeal — so our review is de novo, taking all facts and inferences in the light most friendly to the government, and drawing all credibility choices in the government’s favor as well. See, e.g., United States v. Munyenyezi, 781 F.3d 532, 536 (1st Cir.2015). Sufficiency arguments seldom succeed. See United States v. Correa-Osorio, 784 F.3d 11, 26 (1st Cir.2015). So it is here.

Morosco’s main argument is that the evidence did not demonstrate that he had sabotaged HUD’s quality-control efforts, meaning (his theory runs) that the government’s case against him floundered because prosecutors never “show[ed] that a function of the government was targeted.” But we beg to differ.

Viewed from a government-friendly perspective, the trial record reveals the following (we only hit the highlights):

• HUD fears that if housing-authority employees get advance notice of which units REAC planned on inspecting, the entire inspection regime — designed to ensure the units were “decent, safe, sanitary ... and in good repair,” remember, see 24 C.F.R. § 902.20(a) — would be compromised. And that is because tipped-off employees could then concentrate their energies and resources on just fixing those units up for review.

• Unsurprisingly then, REAC’s inspector-training guide makes clear that inspectors cannot “[p]rovid[e] the property owner with, the sample units ahead of time, so that the owner can clean up the units to be inspected,” and that such conduct constitutes “gaming” and “an attempt to cheat the system.” Also unsurprisingly, to log on to REAC’s server, inspectors have to accept what are called “Rules of Behavior” — rules that say that inspectors’ user IDs and passwords “are to be used solely in connection with the performance of [their] responsibilities as set forth in [their] job description, contract or agreements with [HUD].” Tellingly, Morosco admitted in an email that as “an actively certified REAC inspector,” he was “very familiar with the REAC inspection process.”

• In late 2006, Morosco told Fitzpatrick and Shum that he could access an REAC database and come up with the units REAC planned on inspecting in 2007 — even though he (Morosco) was not the inspector for that job. Fitzpatrick and Shum passed that juicy tidbit on to McLaughlin, who was gung-ho about the idea — McLaughlin badly wanted that “high performer” designation we talked about earlier. So Morosco downloaded key data for CHA’s upcoming inspection ■ from REAC’s server (the data included “demographic” info, e.g., CHA’s buildings, the units within the property), info he could get because of a glitch in the server’s security features.

• Armed with all this data, Morosco generated the sample of the units REAC would inspect in 2007. He gave the list to Fitzpatrick, who then gave it to Shum. And Shum went through the list and matched the numbers there with CHA’s rent roll (REAC only inspects “inhabited units”).

• After getting clued in on the list, McLaughlin organized “SWAT teams.” Comprised of CHA’s management and administrative personnel (Fitzpatrick was involved), SWAT-team members checked and re-checked the to-be-inspected units — and those units only. CHA’s maintenance department then fixed any problems flagged by the SWAT teams, thus ensuring that those units were “perfect.” Meanwhile, maintenance work on the other units “slowed down.”

• Inspection day 2007 eventually came and went, with the REAC inspector generating a list of to-be-inspected units that matched Morosco’s and with REAC’s scoring netting CHA the coveted “high performer” designation— which, again, meant fewer inspections and less oversight by HUD, and more money for. CHA. A guick word about REAC’s scoring system: Inspectors inspected not only the inside of the units but also the property’s common areas, exterior, and building systems. Plus they inspected other elements, like CHA’s management and finances. Without getting bogged down with the math, we simply note that unit inspections accounted for 10% of the overall score.

• The next two inspections — in 2009 and 2011 — involved the same basic script: Morosco would generate a list of to-be-inspected units, using data he got from REAC’s server; he would give the list to Fitzpatrick, who would give it to Shum; McLaughlin would then send the SWAT teams to the selected units; and after the inspections, REAC would designate CHA a “high performer.”

• During all this, Fitzpatrick and McLaughlin warned Shum not to tell a soul about how they had gotten the list. Fearing that CHA personnel might wonder why the SWAT and maintenance’ teams spent so much time and effort on only a few units, McLaughlin came up with a false cover story — that Shum had devised a “formula” for predicting which units REAC would check out. Fitzpatrick clued Shum in on that plan.

• At some point, Fitzpatrick got an email from an official with another housing authority asking if CHA used a REAC consultant. “I guess we need to talk to [McLaughlin] about whether we mention Bernie [Morosco] (and I’m sure we don’t mention Bernie’s ‘extra services’!!!),” Fitzpatrick said in an email to a CHA colleague, adding “[t]his is a little bit of a dilemma!!” Also, in an email he sent to Morosco entitled “Information embargo,” Fitzpatrick “specifically ... reminded] [Morosco] again” that a certain CHA manager “is not in the REAC inner circle.” McLaughlin described that manager as having a “big mouth.”

• Fitzpatrick also gave Morosco both his personal email address and Shum’s so that communications about the inspection-rigging scheme would not be on CHA’s email system.

So despite what Morosco argues, the evidence sufficed for a reasonable jury to conclude that the conspiracy did target a legitimate HUD function — namely, assessing the physical condition {e.g., habitability) of CHA’s federally-funded properties.

Noting that “the physical inspections of the units only constituted 10% of the overall score,” Morosco theorizes that it is “possible that the CHA would have been deemed a high performer regardless of [his] assistance.” 2 But his theorizing is undone by our standard of review, “which is heavily stacked against him” — don’t forget, we must take the facts in the light most flattering to the government’s theory of the case, not his. See, e.g., United States v. Guerrier, 669 F.3d 1, 8 (1st Cir.2011); United States v. Lee, 790 F.3d 12, 13 (1st Cir.2015) (explaining that we must analyze the evidence “in the light most favorable to the jury’s guilty verdict”).

Ever persistent, Morosco contends that CHA fixed units “throughout the year and not just in preparation for the inspections,” which (he suggests) shows that any conspiracy did not “affect[ ]” or “under mine[ ]” CHA’s “quality control.” The claim is both wrong and irrelevant. It is wrong because the evidence — considered in the proper light (afresh, and in the light most agreeable to the government)— shows maintenance work on other units had “slowed down.” And it is irrelevant because — as the government correctly notes — “the crime did not consist of having shabby housing units but of conspiring to keep HUD from accurately assessing them.”

Finally, in something of an offhand suggestion, Morosco calls the evidence inadequate because it (supposedly) did not show “that [he] had an agreement with CHA, that he joined an illegal conspiracy with the required intent or that the purpose of the scheme was to defraud the government.” This single-sentence suggestion is both unaccompanied by a discussion of the relevant evidence and unsupported by citation to legal authority. What we have here “ ‘is hardly a serious treatment of ... complex issue[s]’ ” and is “not sufficient to preserve these points for review” — so these arguments are waived for lack of adequate development in briefing. See Rodríguez v. Mun. of San Juan, 659 F.3d 168, 176 (1st Cir.2011) (quoting Tayag v. Lahey Clinic Hosp., Inc., 632 F.3d 788, 792 (1st Cir.2011)); see also United States v. Zannino, 895 F.2d 1, 17 (1st Cir.1990) (warning that litigants must do more than mention arguments “in the most skeletal way, leaving the court to do [their] work”). But even if we were inclined to overlook this waiver — and we are not — there is no merit to Morosco’s arguments: given the bullet-point summary above, we think that the government presented enough evidence for a reasonable jury to conclude that Morosco knowingly conspired with other CHA-connected colleagues to impair REAC’s inspections and that he intended to achieve the conspiracy’s goal through deceitful means.

Enough said about Morosco’s sufficiency claim.

Jury-Pool-Contamination Claim

McLaughlin’s arrests, indictments, and guilty pleas related to his salary-hiding and inspection-rigging activities received pervasive publicity. So before trial in their case, Fitzpatrick and Morosco asked the judge to question persons in the venire pool about this. Fitzpatrick, for example, wanted the judge to ask them: “Have you read any newspaper accounts of events at the [CHA] in the recent past? If you remember, what do you think of what you read or heard?” And after telling them that “one of the co-conspirators is the [CHA’s] former Executive Director Michael McLaughlin,” Morosco wanted the judge to ask: “Do you know or have you read of any information concerning this housing authority or this person?” Trying to respond to these requests, the judge told the venire panel that “[o]ne of the persons who the [government alleges was involved in this case was” CHA’s executive director, “a fellow by the name of Michael McLaughlin.” And then to “sharpen this a bit more,” the judge said (emphasis ours):

Mr. McLaughlin has pleaded guilty in this case for himself. It received some publicity, the case has received some publicity. Do any of you recall hearing anything about the case with that in mind?

A few jurors responded affirmatively. So the judge stressed two things. First, “the fact that you have been exposed to some information about the case is not disabling, necessarily,” he said, “but it is something I want to explore.” And second, “now that you are involved in the jury selection process,” he added, “[y]ou have got to keep yourself immune from any outside influence at all” — “[y]ou have just got to put it out of your mind.” The judge then questioned the potential jurors who had acknowledged some pretrial-publicity exposure. One — juror no. 14 — had seen “there was an investigation” but did' not know the details. Another — juror no. 5— had heard there was a “scandal, basically,” but nothing “in depth.” And a third— juror no. 18 — had heard “there was a problem” at the CHA, ie., possible “cheating and a man pleaded guilty.” Yet they all confirmed that they could decide the case based solely on evidence presented in court. The parties agree that the judge did not specifically ask any potential juror if knowing about McLaughlin’s guilty plea could affect his or her ability to decide the case impartially — however no party asked the judge to inquire further. Ultimately, the defense struck jurors no. 14 and 18; juror no. 5 sat on the jury, however.

After a break for lunch, Morosco’s lawyer said at sidebar that she was “very concerned” that the judge had “told the jury” about McLaughlin’s guilty plea, adding that she “had no idea” that the judge “was going to do that” and that “[p]art of the reason” she was “so concerned” was “the defense here is whether or not a crime is committed and the issue of criminal intent.” The judge, however, thought “this is belated,” noting that he had made the comment “about two hours ago ... in response to your request” — “I do not see it as a problem,” he said, and “I believe that it has been waived.”

Before the trial got underway, the judge told the chosen jurors that “this case is to be decided solely on the basis of the evidence that is presented here in the courtroom and in light of the law that I give you.” And after warning them not to do their own research during the trial, the judge told them that they did “not have to do anything at all except pay attention here in the courtroom while the evidence is submitted.”

Later at the final-instruction conference (held shortly before the trial ended), the prosecutor brought up the judge’s pre-trial reference to McLaughlin’s plea and asked if the judge intended to “make it clear” that the jury “should not consider anything about him,” including “whether he was prosecuted or not.” As a suggestion, the prosecutor proposed that the judge tell the jury to “keep in mind that whether anyone else should be or was prosecuted for this crime is not a proper matter for you to consider.” Responding, the judge said that he “may have said” that McLaughlin “had been convicted” but thought that he “did not say of what, in what circumstance.” The judge, however, added that “if you want something more on that, I will” do it.

In his final charge, the judge told the jurors that they had “to decide this case solely on the basis of the evidence that was actually admitted here, not something else,” and that they had to “analyze” the question of guilt “separately as to each” defendant. The judge expanded on this point:

You have heard about other people who the parties believe are culpable in some fashion or another, but we are concerned about [Fitzpatrick and Morosco], and you are concerned about [Fitzpatrick and Morosco] individually in your evaluation. And so, I emphasize, again, that you must give separate and individual consideration to the charge against each defendant, and the fact that you find one defendant not guilty or guilty does not mean that you find the other defendant guilty or not guilty.

And in describing the crime of conspiracy, the judge did refer to McLaughlin, but without mentioning McLaughlin’s plea: Prosecutors, the judge said, have proceeded against two persons

that they say are members of the conspiracy. There are others. The [prosecution] has suggested, more than suggested, that Mr. McLaughlin was part of it, Mr. Shum was part of it. So, you are considering any two people who have created this agreement, and then you are asking did these guys, these two defendants, willfully join that conspiracy with the knowledge that I have outlined for you.

The judge then stressed that in deciding whether Fitzpatrick or Morosco had joined the conspiracy, jurors had to consider each defendant’s “own words and actions,” plus “the acts and statements of other persons [jurors] may find to have been members of the conspiracy made during and in furtherance of the conspiracy.” And after telling jurors that “arguments and statements of counsel” were not “evidence in the case,” the judge explained that “it is the evi dence” — i. e., “what you heard on the stand, the documents that you have seen” — that matters. He also reminded them that Fitzpatrick and Morosco enjoyed a presumption of innocence until proven guilty beyond a reasonable doubt.

Fitzpatrick’s and Morosco’s lawyers objected to none of these instructions. And, by the way,' no one — not the prosecution, not the defense, and not the judge — ever mentioned McLaughlin’s guilty plea in the jury’s presence after the jury-empanelment process wrapped up.

Before us, Fitzpatrick and Morosco say that the judge’s telling potential jurors that McLaughlin had “pleaded guilty in this case” denied them their constitutional right to have the jury’s verdict based solely on the trial evidence. Noting that his trial defense pivoted off his belief that Morosco “had legitimate access to the list,” Fitzpatrick writes that the judge’s comment about McLaughlin’s pleading “guilty to the charged conspiracy” (actually, the judge said that McLaughlin had pleaded guilty to something “in this case,” without saying what he had pled to) was really “tantamount to informing” prospective jurors that McLaughlin did not believe the legitimate-access claim — and so neither should they. Similarly, Morosco thinks that the judge’s remark about McLaughlin’s guilty plea “planted the seed” in the would-be jurors’ minds that Morosco had to have been part of the conspiracy— something Morosco could not counter on cross-examination because McLaughlin did not testify.

The government candidly (and commendably) concedes that the judge “should not have” mentioned McLaughlin’s guilty plea to the jury pool. We agree, for case-law has long recognized that a jury’s “ ‘exposure to extrinsic information deprives a criminal defendant of the protections of the Sixth Amendment’ ” — e.g., “ ‘his right of confrontation, of cross-examination, and of counsel’ ” — and that “[t]he jury’s exposure to extrinsic facts is especially troubling when the trial judge is the source of the information.” See United States v. Ofray-Campos, 534 F.3d 1, 18 (1st Cir.2008) (quoting United States v. Santana, 175 F.3d 57, 65 (1st Cir.1999)) (alteration omitted).

Still, the parties bicker over plenty of stuff, like whether Fitzpatrick preserved this argument for appeal: Fitzpatrick says that he did preserve it — agreeing that he did not object below, Fitzpatrick notes that Morosco did and suggests that Morosco’s objection gave the judge a chance to fix things before irreparable harm occurred, and so we should consider the claim preserved for both of them. The government, contrastingly, contends that Fitzpatrick did not preserve it — agreeing with Moros-co that his objection preserved the issue for him (a point on which we offer no opinion, since no one has put that issue in play), the government argues that Fitzpatrick cannot piggyback on Morosco’s objection, and so plain-error review is called for. For simplicity we will assume that Fitzpatrick preserved the argument, because it does not change the outcome.

We review preserved jury-contamination claims for abuse of discretion. See id. at 20-22. But there is a wrinkle: If the jury’s contact with outside info “did not occur inadvertently” and was “accompanied by ‘egregious circumstances,’ ” and if the judge’s actions were not curative, then we will presume prejudice and review for harmless error. See id. That is, we will see if the government has proved “beyond a reasonable doubt” that the complained-of constitutional error “did not contribute to the verdict,” id. at 22 (quoting Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967)) — a fact-specific exercise that requires us to consider (among other things) the evidence’s “centrality” and “prejudicial impact,” as well as “the use to which the evidence was put, and the relative strength of the parties’ cases,” id. (quoting United States v. García-Morales, 382 F.3d 12, 17 (1st Cir.2004)); see also generally United States v. Schneiderhan, 404 F.3d 73, 80 (1st Cir.2005) (noting that harmless-error review turns on an evaluation of the totality of the evidence).

The sides battle over whether we should presume prejudice — Fitzpatrick and Mor-osco say we should; the government says we should not. But we can duck the question. And that is because even assuming — favorably to Fitzpatrick and Moros-co — that the presumption applies, we can classify the judge’s comment about McLaughlin’s guilty plea as harmless beyond a reasonable doubt.

The Evidence’s Centrality and Prejudicial Effect

On whole-record review, we, unlike Fitzpatrick and Morosco, think that the judge’s off-base pretrial remark — that “Mr. McLaughlin has pleaded guilty in this case for himself’ — was (at best) minimally “central” and “prejudicial.” Just hear us out, please.

The info about McLaughlin’s plea could suggest simply that he knew more about the inspection-rigging scheme than Fitzpatrick or Morosco. It could also provide a scapegoat so jurors would know that someone got pinched for the scheme. And it could suggest that Fitzpatrick and Mor-osco must really believe in their innocence. The defense’s argument — that info about McLaughlin’s plea suggested that he thought a'crime had occurred, thus eviscerating the defense’s no-crime-had-occurred defense — is both speculative and weak, because it is clear that some people were indeed trying to cheat the inspection regime.

Keep in mind as well what the judge told prospective and then selected jurors: He, for example, told the jury panel that prior knowledge about the case or the defendants may be disqualifying and must play no role in the verdict. And after speaking to those potential jurors claiming to know something about the case, the judge concluded that they could remain impartial. More, the judge told the panel members that they must decide “the case in light of the evidence that is presented.” Defining evidence as what “you heard on the stand, the documents you have seen,” the judge also told the seated jurors that they had to assess the guilt or innocence of each defendant separately — so although the parties may believe “other people ... are culpable in some fashion or another, ... we are concerned about” Fitzpatrick and Morosco, the judge stressed. True, the judge never told prospective or sitting jurors (either during his inquiry or during his instructions) to ignore what. he had said about McLaughlin’s plea. But neither Fitzpatrick nor Morosco ever asked the judge to do or say more on that score. Also and importantly, the judge’s single mention of McLaughlin’s plea happened before the seven-day trial began — way before the jury retired to deliberate. Critically too, no party made use of McLaughlin’s plea during the trial — further weakening any argument concerning the centrality and prejudicial impact of the judge’s out-of-bounds comment.

Not so fast, Fitzpatrick and Morosco argue: the level of prejudice here is on par with the level of prejudice in Ofray-Cam-pos, a conspiracy case in which the judge had told the jury that 37 nontestifying codefendants were sitting in prison — an error we found harmful to some- of the appellants. Actually, however, their cases are worlds apart from Ofray-Campos. There we found a “direct” link between the disclosed info and the verdict because (among other things) the jury had asked for that data while deliberating (not our situation) and had returned with a guilty verdict soon after hearing from the judge (not our situation either) — circumstances, we concluded, that suggested “that the jury attributed weight to the trial judge’s response, and indeed considered the ... response to be important, if not critical, in arriving at the verdict.” See 534 F.3d at 24-25. Given these night-and-day differences between these cases — there is nothing approaching that kind of “direct” link here, after all — Ofray-Campos adds no oomph to Fitzpatrick’s and Morosco’s prejudice claims.

The Evidence’s Strength

That leaves us with the task of weighing the strength of each side’s case. Focusing on Fitzpatrick’s first, we emphasize the following (some of which we’ve already noted in this opinion): One, Fitzpatrick pinned his not-guilty hopes below on a “legitimate access” defense — a defense premised largely on his testifying at trial that Morosco had told him that he (Moros-co) had “legitimate access” to an “algorithm” that CHA could use to come up with a list of to-be-inspected units. Two, according to testimony from others (as Fitzpatrick is at pains to point out), both he and Morosco told CHA colleagues that getting advance notice of the list was not a problem. Three, the judge (Fitzpatrick argues) blew that defense out of the water when he told the jury pool about McLaughlin’s guilty plea — info (the theory continues) that suggested to the pool members that McLaughlin thought the legitimate-access idea was hooey. And four, Fitzpatrick testified that he had “never organized or directed any SWAT team” and that he had told Shum that he did not think Morosco’s plan would work. Cf generally id. at 28 (commenting that in assessing the evidence’s relative strength, it is “significant” that the defendant “testified on his own behalf’).

Turning now to the government’s side of the ledger, we note the following (some of which we’ve also mentioned before):

• Shum testified that he and Fitzpatrick told McLaughlin about Morosco’s idea for how to rig the REAC inspections.

• Fitzpatrick got the list of to-be-inspected units from Morosco. And then Fitzpatrick sent the list Shum’s way, Shum also said on the stand.

• Shum further testified that Fitzpatrick told him “not to tell anybody” about how he (Fitzpatrick) had gotten the list of units from Morosco.

• And Shum noted that Fitzpatrick had given Morosco their personal email addresses to bypass CHA.

• Testifying under an immunity agreement, Richard Russell — formerly with CHA’s maintenance department — explained in court that before the 2011 inspection, Fitzpatrick had given him the list of units that REAC would inspect. Russell added that he saw the REAC inspector generate the random sample on inspection day — and that REAC’s list matched up with the units on the list Fitzpatrick had given him before the inspection.

Prosecutors did not just rely on cooperating-witness testimony. They also relied on documentary evidence, like this:

• After a person working for a different housing authority asked him if CHA used a REAC consultant, Fitzpatrick emailed a coworker, writing that “I guess we need to talk to [McLaughlin] about whether we mention [Morosco] (and I’m sure we don’t mention [Mor-osco’s] ‘extra services’!!!).” Fitzpatrick’s email — which the judge admitted as an exhibit — closed with, “[t]his is a dilemma!!”

• In another email exhibit — bearing the subject heading “Information embargo”' — Fitzpatrick “reminded” Morosco that a particular CHA manager “is not in the REAC inner circle.”

• Despite claiming that he had “never organized or directed any SWAT team,” an email exhibit showed that Fitzpatrick had asked others to “clear” their calendars for SWAT-team inspections.

• And despite also claiming that he did not think Morosco’s plan would work, another email exhibit showed that Fitzpatrick had encouraged Shum to “work your magic” in getting ready for the next inspection-^-remember, the first inspection under Morosco’s plan had netted CHA the high-performer designation.

Significantly too, Fitzpatrick’s own words — drawn out on cross- and recross-examination — weakened his case, indicating that he did not believe it was “legitimate” to mess around with HUD’s random-sampling procedure:

• Fitzpatrick conceded knowing that REAC used a random-sample methodology for its inspections.

• He admitted knowing (thanks to a letter he got from REAC) that REAC would “inform” CHA “of the units that have been selected for inspection on the day of the inspection” — repeat, “on the day of the inspection.” He conceded knowing (thanks to that same letter) that REAC would come up with those to-be-inspected units from an “on-site” sample drawn on inspection day.

• He initially denied telling the grand jury that the SWAT team inspected all units. But his grand-jury testimony— read into the record at trial — confirmed that he had: “Every time we got ready for a REAC inspection,” he told the grand jury, “[w]e would go out, be a ‘SWAT team,’ ... do inspections of all the units.... ” First he called this part of his grand-jury testimony a “misstatement,” and then he denied he was talking about SWAT teams at all.

• He also admitted never telling any inspector that he knew in advance which units REAC would inspect.

• Relatedly, he conceded that if he had told an inspector about getting the list in advance, then “perhaps” the inspections would be “void.”

• And he admitted never checking with HUD or with CHA’s legal counsel to see if Morosco’s plan was legit.

As for Morosco’s case, we need not say much. Being “an actively certified REAC inspector” Morosco was “very familiar with the REAC inspection process” (quotes lifted from an email he sent to an REAC inspector; Morosco did not testify, remember) — with HUD making clear that an inspector’s giving “property owners ... the sample of units ahead of time, so that [they] can clean up the units to be inspected,” constitutes “gaming” the system, which is an “illicit activity]” (quotes lifted from a HUD instructor training guide). In other words, as the government notes, the evidence showed that he knew his actions were forbidden by HUD because they could defeat the random inspection’s very purpose. His theory below (expressed in his counsel’s closing argument) was that while he had varied from HUD protocol, no regulation or law criminalized, his conduct — a theory that he resurfaces here. But because the judge — without reversibly erring — did not instruct the jury that Morosco had to know that he was committing a federal crime (more on that later, in the mens-rea section), this theory can in no way lessen the strength of the government’s evidence against him.

The Net Result

We know that if “record review leaves the conscientious judge in grave doubt about the likely effect of an error,” we should treat the error “as if it affected the verdict.” See O’Neal v. McAninch, 513 U.S. 432, 435, 115 S.Ct. 992, 130 L.Ed.2d 947 (1995) (adding that “ ‘grave doubt’ ... mean[s] that, in the judge’s mind, the matter is so evenly balanced that he feels himself in virtual equipoise as to the harmlessness of the error”). But — given the minimal role and prejudicial effect that the judge’s line-crossing remark had here (again, he made the comment pretrial, well before jury deliberations; checked with prospective jurors who said they knew something about the case; and instructed seated jurors to consider each defendant’s fate separately), and given the relative strength of the government’s case compared with the defendants’ (what with the government’s putting in evidence undermining Fitzpatrick’s position and cross- and recross-examining him so effectively, plus the government’s showing that Mor-osco knew his actions were forbidden and that it mattered not if he knew that he was committing a federal crime) — we believe that the guilty verdicts returned here were “surely unattributable” to the judge’s error. See Ofray-Campos, 534 F.3d at 26 (quoting Sullivan v. Louisiana, 508 U.S. 275, 279, 113 S.Ct. 2078, 124 L.Ed.2d 182 (1993)).

On to the next claim, then.

Credibility-Bolstering Claim

Like Shum, Richard Russell, the previously-mentioned CHA maintenance official, cooperated with investigators and testified before the grand jury too. Believing that Russell was a “probable government witness,” Fitzpatrick’s lawyer moved pretrial to have counsel appointed to represent him. She argued that as an “unindicted co-conspirator” Russell needed counsel “to advise him regarding his potential Fifth Amendment right not to testify.” And Fitzpatrick’s lawyer explained at a pretrial hearing that she intended to cross-examine Russell “about what his understanding is of what’s going to happen to him, if anything, for testifying” and about “whether he has some understanding that he is going to walk away from this without getting his hair mussed or whether he genuinely has no idea that he’s placing himself at risk.”

The judge (we are told) met .privately with him to see if Russell wanted counsel. Hard on the heels of this meeting, Russell got a lawyer. And sometime before trial started, Russell entered into an immunity agreement with the government covering his testimony.

At trial the prosecutor asked Russell about the immunity agreement near the end of his direct testimony. After Russell said — without objection — that he had testified about these CHA “matters” previously, the prosecutor asked: “Did you testify differently?” Russell answered “no.” Fitzpatrick’s counsel then objected. But the judge overruled his objection.

Near the end of her cross-examination of Russell, Morosco’s lawyer brought up the immunity agreement and introduced it into evidence. Russell then agreed with Morosco’s lawyer that the agreement “indicates that in return for testifying [he] will not be prosecuted.”

On redirect examination, Russell acknowledged that the immunity agreement got signed “last week,” that the government had not “promised anything” before then, that law enforcement had interviewed him “about REAC,” and that he had earlier testified before the grand jury pursuant to a subpoena. Over Fitzpatrick’s counsel’s objection, Russell answered “yes” to the question whether he had “testifped] essentially to the same matters that [he] had testified here today[.]” And again over Fitzpatrick’s counsel’s objection, Russell testified that he had told the truth to the grand jury.

Later, the judge told the parties why he had let Russell testify about “his prior statements to the grand jury.” Expressing regret for having spoken to Russell about the counsel-appointment matter (because doing so had “intruded the Court into [Russell’s] choices about how he wanted to present himself in this case”), the judge intimated that Fitzpatrick’s lawyer had engineered Russell’s immunity agreement by the motion to appoint counsel and then “exploited” the situation by hinting to the jury that Russell’s “statements here were the subject of some form of promises, reward and inducement, and, inferentially,” that his “prior statements would be somehow less than credible” and “perhaps inconsistent.” And the judge said that he had overruled counsel’s objections on the prior-consistent-statements questions “to rectify that litigation gavotte or strategy.”

Fitzpatrick contends that the judge basically let the prosecutor bolster Russell’s credibility at trial by letting him testify that he had told the grand jury the truth, a textbook example of an abuse of discretion — or so Fitzpatrick thinks. See, e.g., United States v. Simonelli, 237 F.3d 19, 28 (1st Cir.2001) (discussing the judge’s “discretion” in this area); United States v. DeSimone, 488 F.3d 561, 574 (1st Cir.2007) (noting that we review decisions to admit evidence for abuse of discretion). He notes — and the government agrees — that the challenged testimony was admissible if it had “some rebutting force beyond the mere fact that the witness has repeated on a prior occasion a statement consistent with his trial testimony.” See Simonelli 237 F.3d at 27-28 (quoting United States v. Pierre, 781 F.2d 329, 331 (2d Cir.1986)). But he is adamant that his lawyer never attacked Russell’s credibility so as to make “consistency” an issue — “[t]o the contrary,” writes Fitzpatrick, “the defense took the position that Russell gave truthful testimony,” noting how “counsel for both defendants relied upon Russell’s testimony in closing as corroboration for the theories of defense.” Also, Fitzpatrick says, the judge was wrong in thinking that his attorney had masterminded the immunity agreement via the counsel-appointment motion, let alone that she had done so to challenge Russell’s credibility — all she had done, according to Fitzpatrick, was to raise some concerns about Russell’s potential “exposure.” And, Fitzpatrick quickly adds, while his counsel did say that any immunity agreement would be fodder for cross-examination, that does not mean that she would have suggested that Russell is a liar — counsel could have simply appealed to the jury’s sympathy, for example by highlighting the “unfairness” of punishing 'one person (Fitzpatrick) when another “with a comparable level of culpability” (Russell) escaped prosecution.

The government counters that because defense lawyers usually use immunity agreements to attack a witness’s credibility, the judge abused no discretion in allowing the prosecutor to show on direct examination that Russell had not changed his testimony after getting immunity. And as for Russell’s testimony on redirect examination, by then Morosco’s lawyer had mentioned the immunity agreement and had introduced it into evidence, “implied[ly]” attacking Russell’s credibility — or so the government asserts.

We need not referee this tussle, however. And that is because even assuming (without deciding) that the judge did err, the mistake was harmless. “A non-constitutional evidentiary error is harmless” if “it is highly probable” that the mistake “did not influence the verdict.” United States v. Piper, 298 F.3d 47, 56 (1st Cir.2002). Fitzpatrick, remember, argues that his lawyer did not torpedo Russell’s credibility at trial but instead “took the position” that Russell had testified “truthfully]” — and he notes that his attorney and Moroseo’s went so far as to “rel[y] upon Russell’s testimony” during closing argument to “corroborate]” the defense’s “theories.” If so, then evidence that Russell had testified consistently in the past could only have helped Fitzpatrick — hence our harmless-error holding. 3

With that said, we move right along.

Prejudicial-Cross-Examination Claim

At trial, Fitzpatrick’s counsel did his best to destroy Shum’s credibility, telling the jury during his opening statement, for example, that Shum was a “conniv[er]” who had “very corruptly handled [CHA’s] books” to help McLaughlin out. And during cross-examination by the government, Fitzpatrick himself said that Shum had “kind of skewed his [Shum’s] testimony.” Fitzpatrick then conceded that earlier he had “talk[ed]” and “jok[ed]” with Shum outside the courtroom, that CHA had sued him and Shum civilly over “this inspection-rigging business,” and that the two men were “commiserating” about the suit. Fitzpatrick’s lawyer protested, saying “I think this is improper.” Overruling counsel’s objection, the judge gave a limiting instruction, telling jurors that “the nature of the case opens up questions of the relationships between various of the persons who have testified here and their relationships now,” and so they could consider the Fitzpatrick/Shum “relationship! ]” and the “regard” they had for each other.

Fitzpatrick now argues that the judge should have excluded this testimony. As he sees it, the evidence had little or no probative value but was highly prejudicial since the jury “would have inferred” that CHA had sued him because he had actually played a role in the inspection-rigging scheme. See Fed.R.Evid. 403 (providing that a judge “may exclude relevant evidence if its probative value is substantially outweighed” by other things, like “a danger of ... unfair prejudice”). His is a difficult argument to win, however, given how our review here is tempered by deference and looks only for abuses of discretion, see United States v. Rodríguez-Soler, 773 F.3d 289, 293-94 (1st Cir.2014) — indeed, the degree of the judge’s discretion in an evidentiary ruling like this is “particularly” wide, see Sprint/United Mgmt. Co. v. Mendelsohn, 552 U.S. 379, 384, 128 S.Ct. 1140, 170 L.Ed.2d 1 (2008).

That Fitzpatrick was still friendly with Shum and apparently thought him an ally in the civil suit helped counter the defense’s credibility attacks. So the complained-of testimony had probative value. And as for prejudice, yes, the testimony had some of that too — most evidence is prejudicial, to one side or another, courts commonly say. See, e.g., Rodríguez-Soler, 773 F.3d at 296. But because jurors obviously already knew that the government had criminally indicted Fitzpatrick on the inspection-rigging scheme, we see little indication of unfair prejudice from the testimony about the scheme also triggering a yet-unadjudicated civil case. Certainly we see no unfair prejudice “substantially” outweighing the testimony’s probativeness— particularly given the judge’s unobjected-to instruction clarifying that jurors could consider the testimony insofar as it related to Fitzpatrick and Shum’s relationship. See, e.g., United States v. Mehanna, 735 F.3d 32, 64 (1st Cir.2013) (upholding the judge’s ruling, highlighting his limiting instruction); United States v. Tejeda, 974 F.2d 210, 214 (1st Cir.1992) (same). The bottom line is that this is not one of those “rare[]” and “extraordinarily compelling” situations requiring our intervention. See Mehanna, 735 F.3d at 59 (quoting United States v. Pires, 642 F.3d 1, 12 (1st Cir.2011)) (internal quotation marks omitted). 4

Mens-Rea Claim 5

At oral argument before us, Mor-osco (through his lawyer) conceded that while what he did (giving CHA the list of to-be-inspected units before the inspection) “does not appear to be innocent,” he “did not see it as criminal.” That tees up his mens-rea argument, which is sort of a corollary to his void-for-vagueness claim: Morosco contends that section 371’s defraud clause — which, to repeat, outlaws conspiracies “to defraud the United States, or any agency thereof in any manner or for any purpose” — lacks a mens-rea requirement. And, he intimates, to cure this problem, the judge should have told jurors (but did not) that they could only convict if they found that he knew his actions constituted a crime — an instruction, he says, that would have resulted in his acquittal, because, as he knew, no HUD regulation criminalized giving housing-authority officials a list of to-be-inspected units before the inspections. Color us unconvinced.

Mens rea (for the uninitiated) is the mental state — “knowingly” or “willfully,” for example — required to convict. The idea behind the mens-rea requirement “is that a defendant must be ‘blameworthy in mind’ before he can be found guilty” — an idea that “is ‘as universal and persistent in mature systems of law as belief in freedom of the human will and a consequent ability and duty of the normal individual to choose between good and evil.’ ” See Elonis v. United States, — U.S. -, 135 S.Ct. 2001, 2009, 192 L.Ed.2d 1 (2015) (quoting Morissette v. United States, 342 U.S. 246, 250, 252, 72 S.Ct. 240, 96 L.Ed. 288 (1952)). So important is this concept that we will usually read criminal statutes as implicitly requiring proof of mens rea even when they do not have a mens-rea component explicitly written into them, id. — though in doing so we read into them “only that mens rea which is necessary to separate wrongful conduct from ‘otherwise innocent conduct,’ ” id. at 2010 (quoting Carter v. United States, 530 U.S. 255, 269, 120 S.Ct. 2159, 147 L.Ed.2d 203 (2000)).

But “[t]his is not to say that a defendant must know that his conduct is illegal before he may be found guilty.” Id. Far from it. Instead, he “generally must ‘know the facts that make his conduct fit the definition of the offense.’ ” Id. (quoting Staples v. United States, 511 U.S. 600, 608 n. 3, 114 S.Ct. 1793, 128 L.Ed.2d 608 (1994)). We say “generally,” however, because in certain .situations — like where a statute presents a danger of criminalizing apparently innocent acts — we sometimes require proof that the defendant knew his conduct infracted a specific law. See, e.g., Cheek v. United States, 498 U.S. 192, 200-01, 111 S.Ct. 604, 112 L.Ed.2d 617 (1991).

Back to our case. Essentially parroting a pattern-jury instruction, the parties— Morosco included — asked the judge to tell the jury that a section-371 conviction requires proof that the defendant acted “willfully,” ie., with “bad purpose, either to disobey or to disregard the law.” See Nancy Torresen, 2015 Revisions to Pat-tem Criminal Jury Instructions for the District Courts of the First Circuit 119 (2015), available at http://www.med. uscourts.gov/pdf/crpjilinks.pdf (instruction 4.18.371(3)); see generally United States v. Charlton, 502 F.3d 1, 3 n. 2 (1st Cir.2007) (noting that the pattern instructions, though often helpful, “have not been officially adopted by th[is] court”). And the judge agreed to do just that. First, though, he told the jurors that section 371 reaches conspiracies to defraud that “have been agreed upon willfully to impair, impede or defeat the proper operation of the federal government by ... deceit, craft, trickery, or dishonest means.” He also told them that the government had to prove “two types of intent”: an intent to “willfully and knowingly join[ ] the conspiracy” and “an intent to violate, whether reasonable or not, ... the underlying” section-371 offense. As for what “willfully” means, the judge said that

[t]o act willfully ... means to. act voluntarily and intelligently and with the specific intent that the underlying crime, that is, interfering with the proper operation of the [HUD] program, ... be committed. That is, when we talk about acting “willfully,” we talk about acting with bad purpose either to disobey or disregard the law, and not to act simply because of ignorance, or accident, or mistake.

Defending the judge’s charge, the government (to quote its brief) says that section 371 neither explicitly nor implicitly “require[s] ‘willful’ action” — a “knowing[ ]” mens rea suffices — so, the government asserts, the instruction here actually required prosecutors to prove “a level of mens rea” higher than what the statute demands. 6 Interesting as the govern ment’s thought may be, the only mens-rea issue relevant here is the one Morosco raises: ie., his claim that the judge should have said more than he did, instructing them that to convict they had to find that he knew his actions were not just improper (he basically concedes that they were) but were “criminal.” The problem for Moros-co is that he never asked for such an instruction, meaning we review only for plain error — a hard-to-meet standard that requires a person in his shoes to show “error, plainness, prejudice to [him] and the threat of a miscarriage of justice.” See United States v. Torres-Rosario, 658 F.3d 110, 116 (1st Cir.2011); see also United States v. Frady, 456 U.S. 152, 163, 102 S.Ct. 1584, 71 L.Ed.2d 816 (1982) (noting that plain error means an error so obvious that a judge is “derelict in countenancing it, even absent the defendant’s timely assistance in detecting it”). But Morosco cites no authority — and we know of none— saying that the mens rea for a case like his is that the defendant knew his conduct constituted a crime. So the situation here is not within a country mile of plain er ror — ie., an “ ‘indisputable’ ” error by the judge, “given controlling precedent.” See Correa-Osorio, 784 F.3d at 22 (quoting United States v. Jones, 748 F.3d 64, 70 (1st Cir.2014), which in turn cited United States v. Marcus, 560 U.S. 258, 262, 130 S.Ct. 2159, 176 L.Ed.2d 1012 (2010)).

We are not done.with “willfully,” however.

Supplemental-Instruction Claim

After deliberating for about two hours, the jury sent the judge a note saying, “Can we have an expanded definition of what constitutes ‘willfulness’ in regards to this charge?” Admitting that he was “not exactly sure what ‘expanded’” meant, the judge talked with the parties’ attorneys and proposed “simply repeating] what I said before here, to which no objection was made.... ” Because the jury had sought an “expanded definition,” lawyers for Fitzpatrick and Morosco asked the judge to say more. Fitzpatrick’s attorney offered some language. 7 But the judge declined to take counsel’s suggestion, concluding that the recommendation did not add usefully to what he proposed to say, Fitzpatrick’s lawyer then asked the judge to caution the jurors that they should not regard the supplemental instruction “as a substitute for the earlier instruction....” Calling counsel’s request “silly” — because, the judge said, the earlier instruction and the supplemental instruction were “the same thing” — the judge then gave the jury a written supplemental instruction, which read (cross-outs omitted):

To act “willfully” means to act voluntarily and intelligently and with specific intent that the underlying crime — conspiracy to impair, impede and defeat the proper operation of the physical condition assessment of federally-funded housing units of the Chelsea Housing Authority by the United States Department of Housing and Urban Development’s Real Estate Assessment Center (“REAC”) — be committed, that is to say with bad purpose, either to disobey or to disregard the law and not because of ignorance, accident or mistake.

The jury returned guilty verdicts roughly an hour later.

Fitzpatrick thinks that the judge’s actions here constituted an abuse of discretion, the standard (the parties agree) that governs our oversight of preserved claims, see United States v. Rivera-Hernández, 497 F.3d 71, 83 (1st Cir.2007), with unpre-served claims getting plain-error review. Ultimately, though, we see no reason to reverse.

Fitzpatrick’s lead argument is that the judge should have given an “expanded” definition of “willfully” since that is what the jury asked for. But he does not tell us what the judge should have said differently in defining that term — e.g., he does not argue that the judge should have given the supplemental instruction that counsel suggested. And given this situation, we can hardly say that his argument adds up to an abuse of discretion. Cf. generally Lussier v. Runyon, 50 F.3d 1103, 1111 (1st Cir.1995) (saying that, “[i]n general, the abuse of discretion framework is not appellant-friendly”); Dopp v. Pritzker, 38 F.3d 1239, 1253 (1st Cir.1994) (emphasizing that that most “appellants who consider themselves aggrieved by discretionary decisions of the district court ... are destined to leave this court empty-handed”).

Conceding that the judge “did not wrongly define ‘willfully’ ” in the original charge, Fitzpatrick next blasts the judge for (supposedly) “omittfing] the thrust of the defense,” first by not reminding the jury that conviction required “dual in tent” — ie., proof that he had “knowingly and willfully joined the conspiracy,” plus had “the specific intent to commit the underlying crime”; and then by “omitting the earlier emphasis” that “ ‘mere presence’ ” at the scene of a crime does not implicate the bystander in that offense (the judge had given a “mere presence” charge in his original instructions). Fitzpatrick preserved neither claim, however. And he makes no attempt to explain how he satisfies the requisites of plain error. We are under no obligation to do his work for him. See, e.g., United States v. Etienne, 772 F.3d 907, 918 n. 7 (1st Cir.2014); United States v. Calderón-Pacheco, 564 F.3d 55, 58 (1st Cir.2009); accord Citizens Awareness Network, Inc. v. United States, 391 F.3d 338, 354 (1st Cir.2004).

Lastly, Fitzpatrick argues that the judge should have warned the jury that the supplemental instruction was not a substitute for the original instruction. Perhaps such an instruction might be called for when a judge “amplifie[s] or explain[s]” the original instruction. See United States v. Parent, 954 F.2d 23, 27 (1st Cir.1992) (quoting Beardshall v. Minuteman Press Int’l, Inc., 664 F.2d 23, 29 (3d Cir.1981)). But even Fitzpatrick admits that the judge’s supplemental instruction essentially mimicked the original, unobjected-to “willfully” charge. And he cites no case — or any persuasive reason— requiring that a judge must give the pined-for warning in a situation like ours.

What this all means is that Fitzpatrick’s supplemental-instruction claim has no legs. But there is still work for us to do.

Sentencing Claim

Relying on USSG § 3B1.2, Fitzpatrick asked the judge at sentencing to give him a two-level “minor role” reduction when calculating his guidelines offense level. With that reduction, his recommended sentencing range would be 10-16 months, rather than 15-21 months. The government objected to Fitzpatrick’s request.

Ultimately, the judge declined to give the minor-participant discount — though he did call the issue “close,” saying “with a couple of different wrinkles maybe it would come out a little differently.” The judge’s reasoning was straightforward: Taking a “holistic” look at “the nature of the agreement” and what “this particular individual” had done “to further the criminal enterprise,” the judge concluded that Fitzpatrick “play[ed] the role of a high-level functionary” who took “relevant information” from Morosco “and pass[ed] it on.” The judge also called Fitzpatrick’s doings “necessary,” stressing that he “did not think that this undertaking could have been successful without” the “activity” Fitzpatrick “chose” to perform. Focusing on relative culpability, the judge found that “at least the people who were indicted, and perhaps others, shared the same level of significance, core significance to the activity, although some had more significant jobs than others.” And the judge suggested that “to the degree that this [analysis] needs to be refined further,” he would do it “in the context of variance or departure” — “[v]ariance, probably.”

Turning to the variance issue, the judge considered Fitzpatrick’s arguments as measured against the controlling criteria in 18 U.S.C. § 3553(a). That Fitzpatrick’s case “is a jail case seems to be” the sentencing guidelines’ “judgment,” the judge said. But, he added, “I am not sure that they adequately reflect the particulars of this case and the nature of this case.” The judge regarded Fitzpatrick’s crime as “very serious.” But he found that Fitzpatrick’s life story and family circumstances justified some leniency. And though convinced that he would not re-offend, the judge concluded that Fitzpatrick had to get some jail time to deter others from committing similar crimes. Using McLaughlin’s 12-month sentence as sort of a “lodestar,” the judge decided to vary downward from Fitzpatrick’s 15-21 month range and give him 3 months in prison — with a 1 year period of supervised release to follow.

Later, Fitzpatrick asked the judge to stay his sentence pending the outcome of his appeal, arguing (among other things) that a proposed amendment to section 3B1.2’s commentary supported a minor-role reduction. But the judge denied the motion, saying that Fitzpatrick’s request for a minor-participant adjustment “was essentially immaterial to the sentence imposed, which was in any event well below the guideline that would have resulted even had the reduction been granted”— and “[t]hat was because” the judge “did not view the applicable guidelines (which ... are likely to be changed by amendment shortly) as adequately capturing relative culpability in the unique circumstances of this case.” So, wrapping up, the judge emphasized that the issue Fitzpatrick “raise[d] about the guideline — even if close as a factual matter at the nisi prius level — did not affect the sentence ... imposed.”

After the judge’s ruling, the federal sentencing commission did amend the commentary to section 3B1.2. Pertinently, that amendment says that judges should not deny a minor-role adjustment “solely” because the defendant was “ ‘integral’ or ‘indispensable’ ” without considering whether he was “substantially less culpable than the average participant in the criminal activity.” See USSG, supp. app. C, amend. 794, at 118.

According to Fitzpatrick, this amendment is “critical” to his case, because in refusing to give him the minor-participant discount, the judge found that he had played a “necessary” role and that the conspiracy could not have been “successful” without him. And he asks us to remand so that the judge can reconsider giving him a minor-role discount in light of the amendment. The government opposes his remand request. 8

The parties skirmish over whether the amendment is “clarifying” or “substantive,” because only “clarifying amendments — amendments that are purely expository — may be applied retroactively.” See United States v. Cabrera-Polo, 376 F.3d 29, 32 (1st Cir.2004); see also United States v. Crudup, 375 F.3d 5, 8-10 (1st Cir.2004). We need not grapple with that question. Even assuming (without granting) that the amendment is clarifying, we think that Fitzpatrick’s remand argument is not a winner. Here is why: As we said a second ago, the amendment declares that judges should not reject a minor-participant reduction “solely” because the defendant’s conduct was “ ‘integral’ or ‘indispensable’ ” without pondering whether he was “substantially less culpable than the average participant in the criminal activity.” But the judge did consider (as the amendment requires) Fitzpatrick’s culpability in relation to his codefendants, finding that they all “shared the same level of significance, core significance to the activity.”

Seeking a way around the problem, Fitzpatrick argues that the judge focused more on the “significance” of his role in the conspiracy than on his culpability in relation to his comrades. But looking at the sentencing transcript, we see that for page after page the judge and the lawyers actually discussed the relative culpability among the eodefendants — which throws cold water on this argument. Also missing the mark is Fitzpatrick’s claim that the judge did not focus on factors like “the degree to which the defendant participated in planning and organizing and exercised decision-making authority” (one of the nonexhaustive list of factors for a judge to consider in deciding whether to make a minor-role adjustment). The parties spent significant time during sentencing on the planning, organizing, and decision-making issues, and we can infer the judge considered and rejected defense counsel’s points before selecting the sentence. Cf. generally United States v. Jiménez-Beltre, 440 F.3d 514, 519 (1st Cir.2006) (en banc) (indicating that we can infer that the judge considered a defendant’s sentencing claim “by comparing what was argued by the parties ... with what the judge did”).

Wrap Up

With that and at long last, we affirm Fitzpatrick’s conviction and sentence, and we affirm Morosco’s conviction too. Also, as we discussed above, we deny as moot Fitzpatrick’s earlier-filed motion asking us to stay his sentence pending appeal.

1 . We still need to .address their claim because, as Morosco writes, even though controlling precedent stands in their way, "[t]his does not mean that a second look is not warranted.”

2 . Even if the advance notice had zero impact on the REAC inspection's outcome, whether or not a conspiracy’s objective is actually achieved is irrelevant to a conviction for conspiracy, because “the essence of a conspiracy is ‘an agreement to commit an unlawful act.' " United States v. Jimenez Recio, 537 U.S. 270, 274, 123 S.Ct. 819, 154 L.Ed.2d 744 (2003) (quoting Iannelli v. United States, 420 U.S. 770, 777, 95 S.Ct. 1284, 43 L.Ed.2d 616 (1975)). The “agreement is a distinct evil which 'may exist and be punished whether or not the substantive crime ensues.' ” Id. at 274-75, 123 S.Ct. 819 (quoting Salinas v. United States, 522 U.S. 52, 65, 118 S.Ct. 469, 139 L.Ed.2d 352 (1998)).

3 . Hoping to show harmfulness, Fitzpatrick argues that stamping the error harmless "fails to appreciate the impact on the jury of knowledge” that prosecutors "had deemed Russell, but not Fitzpatrick, worthy of protection from criminal charges.” But the jurors learned about Russell’s immunity deal when he said "yes” to the unobjected-to question whether he was “testifying today pursuant to an immunity agreement.” So this argument is a no-go.

4 . Struggling to avoid the inevitable, Fitzpatrick suggests that a Ninth Circuit case— United States v. Bailey, 696 F.3d 794 (9th Cir.2012) — calls for a different result. A criminal case, Bailey held that "inconclusive allegations of prior similar behavior" found in a civil complaint are not admissible as prior bad acts under Fed.R.Evid. 404(b). See 696 F.3d at 799-802 & n. 6. Prosecutors in Bailey "used the [prior] complaint to prove intent.” Id. at 802. Fitzpatrick's prosecutors did no such thing. Consequently Bailey holds no sway here.

5 . Fyi: “Mens rea " is Latin for "guilty mind." Black’s Law Dictionary 1075 (9th ed. 2009).

6 . For support, the government cites out-of-circuit cases holding that a section-371 prosecution does not require "willful” intent, see United States v. Khalife, 106 F.3d 1300, 1303 (6th Cir.1997); United States v. Cyprian, 23 F.3d 1189, 1201-02 (7th Cir.1994); United States v. Derezinski, 945 F.2d 1006, 1012 (8th Cir.1991) — though the government acknowledges, at least implicitly, that the judge’s "[t]o act willfully’’ instruction essentially tracks the one we approved in another conspiracy-to-defraud case involving section 371. See United States v. Monteiro, 871 F.2d 204, 208 (1st Cir.1989).

7 . Here is what'Fitzpatrick’s lawyer proposed: An act or failure to act is, quote, willful, unquote, if done voluntarily and intentionally, and with the specific intent to do something the law forbids, or with specific intent to fail to do something the law required to be done; that is to say, with bad purpose either to disobey or to disregard the law. The burden to prove intent, as well as all other elements of the crime, rests with the government.

8 . We should say that no jurisdictional problem lurks here, because Fitzpatrick has not started his three-month prison stint. Fitzpatrick, you see, asked us early on to stay his sentence pending appeal. And we, in turn, entered orders staying his self-report date pending our decision on the motion and setting an expedited briefing schedule. His stay motion is still pending before us. So, for obvious reasons, we now deny it as moot.