Oral argument — Tesla v. NLRB

is the NLRB's authority to restrict speech about unions. The Supreme Court has recognized since the 1940s against the board's original position that such speech is a matter of public unions. Despite those rulings, Congress determined that the board was still overregulating speech, and so it added section 8c to the NLRA to make explicit that views, arguments, and opinions about speech aren't unfair labor practices if they're not threats of reprisal. And even after that instruction from Congress, courts have often had to rein in the board for reverting to its old favor. The board fell back into that habit here, and in doing so, it trained its sites not on workplace speeches, meetings, or letters to employees, but to a social media exchange on the digital public square. It ordered removal from Twitter of a public figure's casual statement of opinion in direct response from a question from a member of the public who is not a Tesla employee. According to the board, this offhand remark to a random person on the internet was a threat to Tesla employees if they decided to vote for the union. But on its face, the tweet affirmed employees complete freedom to vote for a union. Its first words were nothing stopping Tesla team from voting union, could do so tomorrow if they wanted. But the NLRB brushed over those assurances and distorted what followed, which was an opinion on why employees don't want union representation. The union will cost them money, and they'll get nothing in return. The board might not like that viewpoint, but it has no authority to prohibit it. No Tesla employee, as far as the record shows, construed the tweet the way that the board did, as a threat that Tesla would punish employees if they dare to vote union. Not only is that reading contradicted by the tweet's affirmation of employees' right to vote, it is also contradicted by what Tesla employees already knew about the company's strong commitment to employee stock ownership. For these reasons, both the content and the context of the tweet show that it isn't a threat of reprisal. And I think it's not merely a statement that voting for the union will lead to bad results. That's protected speech. It is a statement that conveys to a reasonable employee that voting for a union will cause the employer to punish the employee. That is what a threat of reprisal means. As Judge Friendly put it for the Second Circuit in the Golub case, it is a threat to return evil with evil. That is not what this tweet shows. Could I ask a quick, you have waived your time, right? No, I have waived it, yes. Go ahead, Your Honor. Okay. Is it just the tweet of Mr. Musk, that one tweet that has been ordered to be removed, or is it the entire, I know it was a thread, so is it the public member, the public person's tweet also, and then also the subsequent explanation a couple of days that talk about the auto union's views about management and ownership and how that works. Was that also removed, or is it just the single tweet that's been ordered removed? I'm sorry, I have that factual question still. No problem, Your Honor. So my understanding is that the order is to delete the May 20th tweet, nothing stopping unions, etc. A single tweet. Yes, although I think that the consequence of that, and the amicus brief from the Chamber of that tweet, not replied to it, but retweeted it, that would go away as well, just as a matter of how Twitter, the website formerly known as Twitter, works. But I don't think that that's dictated by the board's order, that's just a consequence of the board's order. I don't think there's any direction to remove the other tweets, although they would. Isn't that, that's other people's speech? Yes. That's third party speech. Right, and also Mr. Musk's other statements that are referring back to that speech. But I think that that other speech from other parties and from Mr. Musk makes a lot less sense if the thing that they're all responding to is wiped away off of the internet. I mean, I think that it's not possible to read the exchange on Twitter in this case, other than as a conversation that's going on over a period of a few days. And if you take away the seminal tweet in that thread, the rest of the conversation doesn't make much sense. Well, please get back to Judge Friendly or whatever it is that you were discussing. I didn't need to get you off track. So my point, Your Honor, was just that the content of the tweet doesn't meet the standard for threat of reprisal. How does the standard of review play into this? Aren't we reviewing the NLRB for substantial evidence? And if I may think the tweet was not a threat, if I think the context was broader or less broader, does that preclude the NLRB from finding the opposite? No, Your Honor. I think that there are five reasons that we've identified why deferring to the NLRB's view about what a reasonable employee would read the tweet as improper. And one of those is the First Amendment rule that generally applies, the Boe's rule. But what about the finding that the tweet is a threat? How much deference do we owe that finding? I think you view that de novo, because those cases from the First Amendment context make clear that the scope of the First Amendment's protection depends on the fact patterns that get classified as across the line or on the correct side of the line. And that is why the Court has said, you don't even apply the clear error standard when a jury makes determinations that are dispositive of that First Amendment question. That was the Boe's case itself involving defamation. And it's also applied the same logic to courts, both federal and state. Was Boe's an agency case, and does that matter? I think if it matters, it only matters in our favor, because obviously there's a— Was it? My first question— Oh, sorry. No, it was not an agency case. It was a case arising out of a federal bench trial. But there have been cases where the Supreme Court has applied this non-deferential fact review in First Amendment disputes to findings by juries. And juries are obviously an important part of our whole constitutional system. There's a— How about findings by other administrative agencies? I do not—I'm not aware of this issue coming up in other cases other than the FTC agency cases. What was the standard of review in those cases? In those cases, the courts of—this is not the Supreme Court now, but the courts of appeals, the Sixth Circuit and the D.C. Circuit in particular, applied substantial evidence review. But they did so noting that the speech at issue there was commercial speech, which, as the Supreme Court's cases make clear, is entitled to less First Amendment protection, and even Boe's said might not warrant the de novo standard of review. The Supreme Court has said, though, that advocacy about the benefits of unionism—that's the word it used in Edward J. DiBartolo—is not commercial speech. So there's no reason to extend that same carve-out on de novo review to the NLRB context. And the general rule, even stepping beyond Boe's, is that the administrative agency's interpretations of constitutional law are not entitled to deference. The court— Their interpretations of constitutional law might not be. I'm just grappling with the idea that here they're looking at a specific tweet, a message, that they found to be threatening. And that puts us in Gissel land, I think, doesn't it? And it doesn't—I mean, the parameters of the First Amendment—I mean, it's been pretty clearly held that threats are not protected. Yes, although Gissel itself makes clear at the very outset of its decision that neither the board nor the union may override speakers' or employers' right to free speech under the First Amendment. And so Gissel doesn't say, don't apply normal First Amendment rules that you would otherwise apply just because we're dealing with an administrative agency. And this court has said that it reviews constitutional objections to agency actions de novo. It said that most recently in the Alliance for Fair Board Recruitment case that we cited. So I don't think that it's at all unusual to apply the normal rules that apply to agencies when they get into constitutional areas and that apply across all sorts of triars of fact and are required by the First Amendment in free speech cases to the NLRB. That issue was not—yes, Your Honor. Mr. Keneally, suppose instead of posting on the site formerly known as Twitter, Mr. Musk had been in an Aspen Institute conference and they had been discussing free speech or whatever and exactly the same interchange had occurred between him and a member of the audience. Does that mean that he or any other corporate representative could be forced to zip their mouths by the NLRB? Well, yes, Your Honor. The NLRB's position, as I understand it, is that it doesn't matter what setting the speech takes place in. We're not dealing with Gissell where there were a series of speeches and letters to employees directly. We are dealing with, in this case, speech that's in the public, it's responding to a non-employee, and the NLRB says that the same standards apply in that setting that would apply in the workplace. Could he write a scholarly article about the role of ownership in companies and how the theory of employee stock options is contrary to the union's understanding of management because it banks you in intention with the management and that sort of thing? Could you write a scholarly article? Well, I suppose he could if he said the things that the NLRB thinks he needs to say about the given bargaining and all that, but the question here, obviously, is about a very casual brief remark. Right, no, but I'm just wondering if he had said, well, it doesn't make sense economically that the whole structure of power doesn't work if the employees own shares because that puts them in line with the management and that's contrary to the union's understanding of the role of the union and the management. Could you make those points, which are just points about management and government and economics that have been known for many years and that the union itself has positions on? I don't see any reason why not, Your Honor. My only position would be that he doesn't need to use that sort of detail, doesn't need to... Well, I was just thinking that he couldn't make those points because that's what he's doing in shorthand and that's what he's getting in trouble for, is calling it a threat. Well, it depends, I guess, on what he says. If he goes on and on at length and includes all the caveats and runs twice lawyers and so forth, I'm sure the board would think twice about declaring it a threat, but what we have here, obviously, isn't conducive to that. We're dealing with a platform where there weren't enough characters to fully spell out all the words to require that such speech also elaborates in detail about the given take in bargaining, but I do think that topic is covered in this tweet because... Can you win if substantial evidence... It's me again, sorry. Can you win if substantial evidence applies? If that standard applies? We do think we can win, and on that point, I would like to just emphasize that courts applying the substantial evidence that don't grapple with the arguments we've made against deference, but that apply substantial evidence, still emphasize that courts need to be vigilant in protecting the First Amendment values at stake. What are we to do with the idea that at least one viewer on Twitter saw his initial tweet as a threat? No, that's not true, Your Honor. Well, he asked, are you making a threat, didn't he? Or say something like that. Are you threatening to take away benefits? So what are we to make of that tweet? Well, I think the thing to make in that tweet is that within 11 minutes, he responded, no UAW does that. So I don't think that you can take for granted that every comment that's on social media, especially in response to a highly visible public figure, is sincerely conveying that person's views. And I think the more important point is that he immediately responded to say, no, no, no, no, you're misunderstanding, and then laid out his reasons for why that he was being misunderstood. And I know that the NLRB has argued that there's no guarantee that other employees saw that follow-up, but I will say that on this record, there's only one employee who said that they saw the original tweet, that's Mr. Moran, and he testified at record on appeal 840 to 842 that he did see the follow-up tweets. He did not say he viewed any of the tweets as a threat. Mr. Keneally, on the side. Could you just step back and tell me what, now that you're before the full court, is your primary argument as to why we should deny the petition a factual one that the context here disproves that it was coercive and proves that it was just predictive about what the union would do? Or is it a legal argument? And if it's a legal argument, is it primarily that for an employer's statement to be actionable, it's got to be an internal statement, a direct dialogue with employees. It can't be something uttered to the public. And if it's the latter, could you address the Laredo Coca-Cola case that's been cited several times, but I don't think in your briefs you've ever responded to it? Well, yes, we're making three arguments, Your Honor. The first, I would say the narrowest, I wouldn't call it our primary argument, but our narrowest argument that would avoid the court having to grapple with some of the broader legal questions is that even under the substantial evidence standard, this was not a threat of reprisal. And I think that that follows from cases like the Federalist case, which talks about the court's vigilance being necessary, and also the Crown Court Conceal case, which says that employers don't need to use the rhetoric of a Federal Trade Commissioner or a Federal Board member when unions are using the rhetoric of Mark Antony. So that's point one, narrow point. Then we have two primary legal points. I wouldn't say that the one Your Honor identified was a categorical claim that we're making about how you can't possibly have a threat outside the context of the workplace. Our point on that issue is just that it should be a little conclusion that a threat is threatening, that a statement is threatening when you're outside that and you're not talking to the employees directly. The Laredo case, Your Honor mentioned, I believe, was talking to the member of the press about how the company wasn't going to hire back workers who were on strike, which was itself an unfair labor practice. And so that's clearly a threat, because you're saying, I'm going to do- It was in the newspaper. Exactly. Yeah. So we're not saying there's a categorical ruling about that. Thank you. There are two main legal points that I think are broader. One is this point about what the court's role should be in reviewing NLRB determinations along these lines. And then the other is that the NLRB has failed to anticipate over many decades the Supreme Court's new test for true threats, which is not just that the content must be objectively threatening, but also that the speaker must have a wrongful mental state. That, of course, developed after the panel decision in this case, but we think that there's no basis for saying that that same logic doesn't extend- That's like countermeasure, correct? Exactly. But that would be the Supreme Court's prerogative to do the same with all these cases from the late 60s and 70s. Miller, Tinker, all these First Amendment cases that now were in the internet world, in the true threat case, it was the Supreme Court that did that in countermeasure. Are you asking us to do that ahead of them? Well, yes, Your Honor. The Supreme Court was responding to a circuit split on this issue, and there's nothing in the Supreme Court's case law that forecloses subjective mental state requirement. I just want to be very clear about that. There are Court of Appeals decisions that do, but obviously those don't trump countermeasure, and they don't justify a narrow reading of countermeasure. The closest that the Supreme Court has ever come to this question in its NLRB decisions is Gissel, of course, but Gissel talks repeatedly about the intent of the speaker and says that employers need to- It reversed Joy Silk, which was itself an intent case, right? The Joy Silk Doctrine, that's what Gissel- Well, that's a little bit of a different issue, Your Honor. Joy Silk- But it's state of mind. Right, but the Joy Silk question was about a bargaining order, the remedy for unfair labor practices that take place- Okay, thank you. That's very clear. Mr. Connealy, can I try to get us back to a question that Judge Douglas and Judge Wilson asked, and maybe this also goes to something Judge Higginson was getting at, which is how do we think about the First Amendment as to administrative agencies generally and the Board specifically? And another case that I'm confused about is Edward J. DePartolo, because that is an NLRA case, it is an NLRB case, and it involves an NLRB finding of coercion, right? The finding is that the hand-billing is coercive, and when it goes to the Supreme Court of the United States, the Supreme Court specifically says, we normally apply this deference that Judge Wilson was asking about, specifically apply the deference that Judge Douglas was asking about, and then says, but this is hand-billing, and it does not matter that the NLRB found it coercive, we are going to apply constitutional avoidance to get around this problem and to vitiate the order. So I'm wondering if this is any answer to Judge Higginson or to my colleagues on the other side of the bench. Well, I think that's exactly right, Your Honor. I think that the DePartolo case makes clear that that constitutional avoidance principle doesn't just extend to abstract principles derived from the NLRA, like perhaps the Counterman case, I think that that's a very good candidate for constitutional avoidance here, but it also applies to the NLRB's application of the NLRA to specific fact patterns, like the hand-billing of the mall in that case. And so I think that that's another reason why you can't just pretend that everything that the Supreme Court has said about review of, in First Amendment cases, and even with the NLRB, is irrelevant, because the Supreme Court said in that case specifically that constitutional avoidance is an important protection against even constitutional doubts about what the NLRB is deciding. But even if you meet the substantial evidence test, which you say you do, we shouldn't reach the constitutional issue, right? Well, you could certainly address the case, resolve the case on the narrow substantial evidence ground. I don't think that the Court needs to go further than that, although I think that it would be very beneficial to parties to have that kind of guidance, because this is not an issue that's going away any time soon, I think, especially as the line between workplace speech and public speech is getting increasingly blurred by social media and the like. Well, social media is just the facts here. The fact is also the Aspen Institute, an interview with a business magazine or the Wall Street Forum, in which a corporate executive voices some view about union activity. That's correct, Your Honor. That's absolutely true. For remedy, is there a version where this would go back for them to consider the Supreme Court's take on threats and let them look at that since it's the timing of it and that sort of thing, or is that not a possibility? Well, I think we're dealing now with a tweet that's about six years old. The Court hasn't asked for a remand to address countermeasures. It's just that we don't view countermeasures as relevant here, so I don't think that that is warranted, and I think it's impossible to read the record in this case as supporting a recklessness finding, and the reason why I say that is because of those follow-up tweets that were very prompt and that tried to dispel any misimpression. Even if you don't think that those shed light on the objective meaning of the original tweet, I do think they shed light on the mental state of this case. Assuming, arguendo, that someone disagreed or that the Court—and this is arguendo—assuming, arguendo, that the Court disagreed that substantial evidence that you want under that standard, but also was concerned that taking down speech is an odd remedy, is an odd thing to do, that normally the worst thing that happens is you have to read a little speech in front of all your employees and that sort of thing, and this is a unique approach to say we were going to take down your speech and then, as a result, also take down ripple effect, other people's speech, and so that the remedy is what the problem is. What happens then? Well, we do think that the remedy itself is problematic and that it's over broad given the remedies that are also in the order which say we won't threaten to take away your benefit. You know, that's a statement that the employer has to post, so we do think that that is an independent problem, but I do think that the broader problem is about trying to chill—well, not trying, but having the effect of chilling speech. But why couldn't this issue be addressed by what you argued around page 49 of your recent bench memo—I mean, sorry, your recent brief—is that Musk is not a party in this case and he was not on a Tesla—on a Tesla-operated account, and so why is this even in front of the NRLB in this case? Well, we don't have an objection to that, Your Honor. I think the only case where the NLRB has attempted to do this kind of thing with an executive or high-ranking officer's tweets that relate to the company, but on a personal account, was the Federalist Media case where the Third Circuit refused enforcement of that decision, so we wouldn't have an objection to— Doesn't that end a lot of these difficult issues that you've addressed? I think it would, Your Honor. Are you waiving the question about the discharge of Ortiz? No, Your Honor. I was going to turn to that if there's a few minutes left before my time expires. I think the key point on the Ortiz discharge is the ALJ's express finding at 6285 that the credited evidence shows that Respondent—i.e., Tesla—terminated Ortiz for lying during an investigation. The ALJ thought that this finding automatically established a violation because the lies were with protected concerted activity, but that's not the right analysis. The usual rule, and it's made explicit in Section 10C of the statute, is that the Board cannot second-guess an employer who discharges an employee for cause, and lying to the company is cause. This Court recognized in Federal Mogul, Miller-Brass, any employer has the right to demand that its employees be honest and truthful in every facet of their employment, and the Board has tried to make the questions in this case as completely unrelated to the business of the employer, but I don't think that's a fair portrayal at all, because we were dealing with a complaint from an employee who felt targeted and harassed that was then directed to HR, and also the alleged harassment involved use of company resources, and whether or not there was an advanced rule saying you can't use company resources this way, it's at least fair for the company to investigate exactly what happened so it can tell whether it needs to revise its rules or whether something improper happened here. So given all that, the employee didn't have a right to lie, and the Board can't order reinstatement. May it please the Court, my name is Micah Jost for the Labor Board. In my opening time, I would like to address three issues. First, deferential review of the Board's fact-finding. Second, the objective measure of coercion under Section 8A1, and third, Tesla's discharge of Richard Ortiz for union activity. On the first two points, Tesla is asking the Court to overrule Supreme Court precedent. Could you please be louder? I'll do my best, Your Honor. I'm sorry, I had my appendix removed last week, and I'm doing my best. Sorry to hear that. I will try to speak up. On the first two points, Tesla asks the Court to overrule Supreme Court precedent, and on all three points, Tesla is asking the Court to depart from its own precedent and that of its sister circuits. First, on the standard of review, the Supreme Court said in Gissel that the courts must defer to the Board's judgment as to the impact of utterances in the context of the employer-employee relationship. For 55 years, this Court and every other circuit have taken the Supreme Court at soaring with disdain out of expression. Only now, though, is it calling to account what the Supreme Court had said in 1973, where it said the Court is required to give great deference to the board's expertise in this highly sensitive area. Tesla now argues that Gissel meant the exact expressly. That is the review that the board asked the Supreme Court to affirm. Meanwhile, the employer, like Tesla, asked for an independent examination of the facts, reciting word for word the standard that Tesla now repeats, which comes from New York Times v. Sullivan in 1964. The Supreme Court rejected that standard. It asked whether the board's finding was reasonable, and to eliminate any possible doubt, the Court cited Virginia Electric and Power Company from 1941, reaffirming that precedent, and that case emphatically said, in a case dealing with the interpretation of written speech, that Section 10e of the Act, quote, precludes an independent consideration of the facts. That deference the Court held in Gissel does not offend the First Amendment, and the Supreme Court has never questioned or revisited that holding. Meanwhile, the courts of appeals have uniformly refused to extend the Sullivan and Bowes independent review standard into the agency context. The FTC cases that we cite have recognized that to do so would be to usurp the Supreme Court's prerogative to overrule its own precedent. The FTC cases also help explain why it would not make sense to extend Sullivan and Bowes as Tesla requests. A board like the FTC has expertise in a complex area, labor relations, and the board is in the best position to draw factual inferences about what will or will not coerce employees in context under the totality of the circumstances. Unlike in Bowes, there is no judge-made common law standard here. There is a statutory standard that the agency has always administered, and that judges have no particular expertise in applying. Tesla's response is simply that the FTC deals with commercial speech, and employer speech is not precisely the same as commercial speech. But I note that the Supreme Court has repeatedly drawn a close parallel between the two for these purposes. As I'll discuss more in a moment, in Virginia State Board of Pharmacy in 1976 and Orlick v. Ohio State Bar Association in 78, the court drew on Gissel in formulating the constitutional framework for regulating commercial speech. In short, Gissel rejected Tesla's arguments for de novo review, and Gissel remains the law of the Long before Gissel, it was settled that the standard for coercion under Section 8A1 is strictly objective. To cite just two examples, in 1959 in American Freightways, which is cited in the union's brief, the board said that it was already well settled at that time that the employer's motive is irrelevant. The test is whether conduct reasonably tends to coerce. In 1963, this court in Hendricks said that it was irrelevant that employees had testified they were not threatened and that the employer testified that it had no such purpose. All of that, quote, misses the whole point in this court's words. The question is what employees could reasonably conclude. I'd also note this court's recent decision in UNF West in 2016 that says the very same thing. That was the well-established law that was before the court in Gissel when it expressly endorsed, quote, the standards used below for evaluating the impact of an employer's the test was whether the communications reasonably tended to convey a threatening message that the employer could close the plant if employees unionized. As the Supreme Court put it in Gissel, the question is not the employer's sincerity, but rather whether a prediction is capable of proof. Absent from that test is any mens rea requirement. The Supreme Court's recent decision in Counterman is in complete harmony with Gissel in this respect. Counterman imposed a mens rea requirement for criminal true threats of violence and in doing so, it followed Sullivan which said that to prevent the chill of free speech, there must be a mens rea requirement for defamation of public figures. But the Supreme Court has regularly recognized in Bose and again in Counterman that certain kinds of speech such as commercial speech do not implicate those chill concerns because the speaker is charged with a and can thus easily avoid overstepping the line of protected speech. That is where Gissel fits in. Gissel also cited Sullivan and distinguished it. While Sullivan was concerned with the chill of free speech in the political arena and the sort of comments that might be closer to comments that an employer might make at a conference or in a law review article, instead what Gissel was talking about was an employer's speech in a narrow setting in which the speaker is in complete control and has superior knowledge of the effect its words will have. That is what the court is talking about in Gissel when it says that an employer can avoid conscious overstatements and that is why the board can reasonably infer what the speaker intended based on the presence or absence of objective factual support. When an employer predicts negative consequences of unionization with no objective basis, it knows perfectly well the statements will have on economically dependent employees. And again, all of this is confirmed in Virginia State Board of Pharmacy and Oralik. Those cases reaffirm that the board can regulate the impact of utterances without offending the First Amendment for this reason. Chill is not a concern here in the same way as it is in cases like Counterman. Once again, Gissel remains the of Tesla's misconceptions. First, the board did not apply a novel standard here. The standard is well settled. I refer the court in particular to the board's Fresenius decision. I think that even Tesla recognizes that that decision lays out the controlling test. There is simply no daylight between that test and this court's test in Cordua and the D.C. Circuit's test in United Services Auto Association and SELCO. The only real dispute is over the board's application of that standard to the facts, and that is unquestionably a question for substantial evidence review. Second, Tesla continues to argue that the board made an express finding that Ortiz's lie was its real motivation for discharging him. That is simply wrong. At ROA 6283, the administrative law judge discusses competing testimony about the reason that Ortiz was given for his discharge. There was testimony that it was because of the lie, and there was testimony that it was because of violating Tesla's conceitedly unlawful confidentiality policy. At ROA 6285, the board through the ALJ simply resolved that credibility determination in finding that Ortiz was told he was discharged for lying. But two pages later at 6287, the board found as a matter of fact that the lie was really a pretext, and Ortiz was in fact discharged for his union activity. That is an alternative additional basis for finding the discharge unlawful even aside from the Fresenius analysis. Third and finally, Tesla misunderstands Section 10c of the Act, which bars reinstatement of employees fired for cause. Our brief cites two Supreme Court decisions that dispose of this argument. In Fiberboard, the Supreme Court explained that reinstatement necessarily complies with Section 10c if a discharge was an unfair labor practice in the first place. By definition, such a discharge is not for cause. Transportation Management at footnote 6 of that decision also explains that Section 10c does not bar reinstatement if union animus played a role in a discharge. Here, Ortiz was not fired for cause as the board found he was fired for union activity, his discharge was an unfair labor practice, and it has remained unremitied for six years now. We respectfully ask the court to reinstate him now. I now wave the remainder of my uninterrupted time and welcome the court's counsel. I didn't hear you say anything about Edward J. DiBartolo. So in that case, the board said the hand-billing is coercive, made a finding of fact, and I'm looking at the decision right now, and I do not see a single site to just jizzle, however pronounced the case, and the court specifically said we're not deferring at all. Not only did you not get substantial evidence, you got zero deference in that case, and that's obviously post-jizzle, 1988 case, unanimous decision by the Supreme Court of the United States. Why does that not end the entire reliance you have on jizzle in this case? Your Honor, jizzle is dealing with a—jizzle, the Supreme Court was interpreting Section 8A1 and affirming the board's interpretation of that provision. In DiBartolo, the board and the Supreme Court were dealing with a different provision, and the question was whether hand-billing is sufficiently akin to picketing to make it automatically coercive. It's a finding of fact of coercion, is it not? That was the entire basis of the board's decision in that case, the 1988 case. I think ultimately, Your Honor, the Supreme Court's decision regarding hand-billing dealt with a category of conduct. It dealt with not just the specific facts of that decision, but rather with a category of union conduct that the Supreme Court was saying, we are unwilling to bring this new type of conduct into the same realm as picketing. That's different from this case where, in gizzle, the Supreme Court already addressed this type of comment. It addressed threats from an employer that are encountered or heard by its employees, and the Supreme Court did not find that there was any first amendment— What was the remedy in gizzle? The remedy in gizzle was a cease and desist order and also an order to recognize the union because there had been an election. What about the pamphlets? I'm not aware of any remedy pertaining to pamphlets. I'm not aware of any remedy with the pamphlets either. It's not like the Supreme Court said, burn the pamphlets, right? No, that's right, Your Honor. Would it matter if Elon Musk's tweet in this case wasn't in a text? He instead made a video and put it on Twitter and just said the same content? You're asking in terms of the remedy, Your Honor? Yeah. Could you order him to delete the video? I'm not sure if the board has ever confronted that circumstance. I'm asking you—I mean, your point is that Twitter speech is no different than in gizzle. So my question is, what if—it's still on Twitter. It's the same stuff. Instead of putting it in text, he just makes a video. He's on his phone and he posts it on Twitter. I think it's possible, Your Honor, that the board would have to confront what is the full content of that video if there's only a small offending. It's identical. It's simply that statement on video. I imagine that under the rationale here, if it remains posted, the order would be to take— What if he wrote it in a book? If he wrote it in a book, I think we would have to look at the kind of remedies that the board has imposed historically. The board has never required someone to go around and round up all the physical copies of something that exists. And there's no such order here. This tweet, for example, obviously it appears in the briefs. It's been republished in media. It's all over the place. And the board is not ordering Elon Musk to go hunt down every copy. It's simply attempting to remedy the unfair labor practice to the best of its ability by taking down the most prominent display. So could they just say, you don't have to round up all the books, you just have to round up some of the books? It's exactly the same speech. I'm trying to understand. The way I understand the board's position in this case is that it doesn't matter that it was on Twitter. It doesn't matter that Twitter is a public forum. It just matters that it was a threat. We have found, as a matter of fact, that it was a threat. And we have the ability to make you take down the threat. And so my question is, what if he had put it in a book? Exactly the same speech. Don't fight the hypo. I'm simply saying, Your Honor, based on what the board has done in past cases, I don't think that the board would order any destruction of books. So the position of the board is that Twitter enjoys less protection than a book? I wouldn't say less protection. I would say that the way that it's displayed is distinct in that it's entirely within Musk's control to simply delete this thing. It's more akin to a billboard that he's displaying at his workplace. If he was displaying a billboard, when he tells his employees in the remedial order, I will cease threatening you, but then continues to display a billboard that threatens, I think that the board would say that that is an inadequate remedy. Counsel, I have two questions. You told us that we should defer because Gissel says that we look at the impact of utterance in the context of the employer-employee relationship. But as Judge Gissel would not seem to be applicable when it wasn't uttered in the context of the employer-employee relationship. So that's the first question, if you could address that. If it's not in that context, then that deference seems to be misplaced perhaps, but maybe it is in that context. And the second question is related. You also said Gissel says that you decide if the speaker is predicting negative consequences with no objective basis. Isn't that what you said? That's right, Your Honor. Okay. Well, isn't there an objective basis to predicting negative consequences here? Management labor theory in the UAW's position at numerous past employers, so there is an objective basis for the comments that were made by Mr. Musk, weren't there? So there wouldn't be deference under Gissel because it's not negative consequences with no objective basis. So those are my two questions, please. Yes, Your Honor. Beginning with the second question first, there is simply no objective basis for the initial tweet on May 20 where Elon Musk said that employees would give up stock options if they unionized. Elon Musk doesn't know what the union's demands are. The union cannot force an employer to give up stock options. There's no evidence that the UAW has ever given up stock options. Wait, is an objective basis, it doesn't have to be correct necessarily. I mean, he gave an objective basis at least two days later by saying, no, no, no, no, the union makes you give up the stock options. That's happened in all these other plants. So, Your Honor, he didn't provide that objective basis and employees would not have inferred that there was such an objective basis from the initial tweet on May 20. And we are looking at the initial tweet on May 20. There's no finding that any subsequent tweet was an unfair labor practice and there's no order with regard to those things. So doesn't it just have to be there is no objective basis? I don't think you have to objective basis. That's not part of the Gissel rule. It just has to be no objective basis. And history and management theory would both be objective bases, wouldn't it? Wouldn't they? I disagree, Your Honor, because his initial statement was not the UAW will demand that I get rid of stock options. Had he made that statement, then that would be the question. Was there such an objective basis? But the board's finding is that a reasonable employee would read his initial statement to mean I will take away your stock options if you unionize. A reasonable employee recognizes that Elon Musk and Tesla conveyed the stock options and they have the power to take them away. A union simply does not have that power. Where do you get the, I mean, are you just inferring from the fact that Musk runs Twitter as well as Tesla that the employees necessarily were apprised of these comments? The board found, and it's undisputed, that at least one employee viewed them. But also, how do we know that was an employee? I'm not talking about the person who responded. I'm talking about testimony from Jose Moran on the record where he was asked, did you see this tweet? And he stated that he had. But we're also dealing with only one employee is sufficient to invoke not only the claim that this is a company wide threat, but also censorship. One employee is sufficient, Your Honor. But beyond that, we're dealing with What's your best case for that proposition? I would point the case, well, many of our cases I guess it's consistent with the idea that even if all the employees aren't threatened, it can still be a threat. But it has to be brought, the employees have to know about it, right? Your Honor, employees do have to know about it. And we cite many cases where a single employee, the Crown Stationers Give me your best case, please. The Miller Electric case is sort of a quintessential case that the Board cites for the standard, that's a case where an employer spoke to a single employee. But I think in Hendricks, I think there were also threats that were delivered to individual employees. But I don't want to get away from Judge Elrod's question about the medium here. We are again dealing with Elon Musk with a following at that time of over 22 million people. He has 40,000 employees. And I think it's a reasonable inference that many of them follow him on Twitter. He was tweeting about his employment situation. He was talking about his benefits, his employees, their union campaign. This was not a broader philosophical discussion. And in that context, I want to point to the stipulation at 4756 of the record, where Tesla stipulated that Musk uses his account to tweet about official business and about employment policies at Tesla. Employees would have reasonably understood that. I also point to testimony from his second in command, his head of HR, Gabby Toledano. Her testimony is discussed at 6089. She said that she understood that when Elon Musk tweets about Tesla, he is speaking for the company. And that is the standard. What was the record site for that? The record site is in the 700s, Your Honor. The transcript page would appear in the board's discussion there, which is at 6289. Counsel, going back to Judge Oldham's line of questioning, why isn't Packingham just dispositive of the treatment of Twitter as a public square entity as opposed to something that is regulated differently by NLRB as a medium? Your Honor, I don't think that there's any argument from the board that Twitter is being regulated differently. Twitter is being regulated in the same way as a press release or as a statement as in Laredo Coca-Cola from a, you know, responding to a cold call from a newspaper reporter. An employer spoke off the cuff and said something that was going to be reasonably coercive to employees. And that's exactly the kind of scenario that we have here, where a statement is made online in a context where employees are sure to encounter it and do, in fact, encounter it. We have to look at the perspective of the employees here. But there's a distinction between, if I understood your answers correctly to Judge Oldham's questions, the ability to delete information or to retract or censor a comment that the NLRB considers to be violative. You're treating Twitter, in this instance, to take something down, I think is the word we've been using in the briefs, differently than something I think Judge Oldham mentioned, like something in a book or something that was put in a journal or the Wall Street Journal or something like that. That's not treating it differently? No, I think I understand your question, Your Honor. If you're speaking about the remedy in terms of the deletion— The remedy, yes. I want to emphasize that that's separate from the question of whether there was an unfair labor practice and whether there can be— Understood. But at that stage, I would just point to the board's customary remedies when you're dealing with unlawful discipline, for example, even in this case and in cases like In-N-Out that this Court has enforced. The employer is ordered to expunge the records to get rid of content, and it's getting rid of content that memorializes and documents an unfair labor practice. Similarly— Could you turn to Ortiz and the— He's right. Oh, I'm sorry. Could you turn to the issue of Ortiz and particularly in the stability legally in the case law of the position you're taking that there's a right to dissemble or to lie if it's involving union organization? Yes, Your Honor. I think the key takeaway with regard to Ortiz is I think I would distill the test as follows based on Fresenius, based on Cordua, and other cases that we cited. The question is essentially, does the employee reasonably understand the employer is prying into union activity? If so, then the board asks, does the employer have a legitimate interest in the questions it's asking, and are those questions reasonably tailored to those interests? And if those two are answered in the employee's favor as they would be here, first of all, Ortiz reasonably understood he was being asked about union activity, which is to say sharing this information, and Tesla had no legitimate interest because this was a policy that did not exist prohibiting employees from sharing amongst themselves workday photos that they all independently had access to. Well, doesn't the employer have an obligation to investigate a facially valid complaint of the release of non-public information? The non-public information here was Pratt's salary, which is never mentioned in your brief, nor by the ALJ in her conclusions, and also that Pratt, and I know the ALJ has some throwaway line about how this was a disingenuous complaint, but I don't think HR is entitled to say that any complaint of employee harassment is disingenuous. Your Honor, I want to first clarify the salary was public information because Travis Pratt testified to it at a public hearing in front of legislators. That's absurd. That's absurd because it was not, the fact is even Ortiz and Moran did not know who had testified and whether the people who testified in Sacramento were employees. Your Honor. It's not as if, and then posting it on an internal Facebook page for employees of the company gives it a circulation that it clearly would not have had just because he testified in Sacramento. Your Honor, that information was not drawn from workday. It did not come from the employer's proprietary systems. Of course, it came from the workday. No, Your Honor. That's your inference. There's nothing in the ALJ report about that. Your Honor, there is nothing in the record. Moran said that he looked at the workday profiles in order to find out how much other people would make so he could compare those. That is the only thing that I read in the 10 pages of very small tape that refers to the origin of Pratt. The ALJ never says that this information about Pratt's salary came only from Sacramento and was therefore waived. You're making that up. Your Honor, I simply disagree. There's no evidence in the record that that information was obtainable by Moran through workday. It's uncontested that the only evidence . . . Moran said he looked at workday. They asked Moran, why does he look at workday? Why does he access workday? And he says, I access workday to compare my job duties with my peers and see who's making what. Just the same as people at the law firms. You know, people at the law firms don't have access to other associates' pay scales ordinarily. Your Honor, Moran's testimony was that he was trying to look at the levels, the numbers associated with people's names, nothing about the actual pay that they made. That information was not available to him through workday. But in any event, I want to point to the final consideration in the test that I was laying out in response to Judge Southwick, which is simply, even if the employee reasonably understands that this was prying into union activity, even if the employer has no legitimate interest and the questions are not reasonably tailored, there's a final question. Did the lie actually have any impact on the employer's business or on the employee's job performance? And here, all three of those elements are in the employee's favor for the finding that this was an unlawful discharge. With that, I would simply ask for enforcement of the Board's order in full. Thank you, Your Honors. Good afternoon, Your Honors. I'd like to start by briefly discussing the context for the Board's findings, which is the uncontested unfair labor practices in this case. In addition to the matters argued here today, the Board's 2021 order found that Tesla committed the following violations. Prohibiting workers from speaking to the media about working conditions. Repeatedly blocking workers from distributing union flyers criticizing those working conditions. Announcing workers could be fired for distributing union materials without company authorization. Interrogating Mr. Ortiz about flyers concerning workplace injuries. And disciplining an employee for union activity. Tesla does not contest these Board findings. These violations form the context for the incidents under review today. And the Board properly considered this context, both when evaluating the threatening nature of Must's statement and the motivation for investigating and discharging Mr. Ortiz. As far back as Virginia Electric, the Supreme Court has said that the Board's findings on isolation from other conduct of the company. And here, there's an extreme amount of other conduct that forms the basis for the Board's finding that this was a coercive statement. How do we know that's an extreme amount? Do you have a comparison? Or, I'm sorry, did you waive time? I did not. I apologize, sir. I'd just like to add that to the issue that just was discussed prior. I want to bring up citing to the record. It was at the public hearing that Pratt testified both his name and his salary. And that's at ROA 4832. And then Must watched that video and then learned what his salary was. And that's where this all came from. And I'll point you to the screenshots that we're talking about that were transferred from one worker to another. Those screenshots are in the record. They do not contain this worker's salary. The only place this worker's salary was put out there was at this public hearing in Sacramento, which was an issue about union, critical in the union campaign of the plant. But that, there's not a word about that in the ALJ opinion, nor is there a word about the salary being public knowledge as to any employee, as far as that goes, in the Board's initial briefing to us. I think what they talk about is fact, you know, factual errors and guess which is report to Graminger. That's what they're focusing on. And then Moran did testify that he looks at workday in order to compare his salary with those of others. As my colleague said, he was talking about the levels, Technician 1, Technician 2, Technician 3 to see who's been there longer than him and has the different levels. Turning to the constitutional issues in this case. Is that a lot? Is the amount of incidents that you read, you said that was an extreme amount. Do you have some context for that comment? I was asking you that at a wrong time, but it's still my question. Yes, of course. No, I believe I said, well, they're severe and pervasive. They may also be extreme. I'm not sure I said that word. That's what you said. Okay. Well, yes, this is a lot. This coercion was a heavy level of coercion. These workers did not end up forming a union. These unfair labor practices create a strong understanding that when workers are listening to what the employer says, they're listening for any implication that affects their own livelihood. Their economic dependence is on this employer. And so those, they don't have a disinterested ear. They hear any implication. And that's why I'd like to say that unlike the public arena, the workplace is not a neutral ground where employers and employees can debate workplace issues in a vacuum. The employer has immense power to affect every aspect of its employees' lives, whether by terminating them, reducing their benefits, or threatening to do so. And the Gissel court emphasized this power difference. Well, wait, you said in the workplace. Yes. Mr. Musk's tweet was not in the workplace. Yes. And I think it's very important to realize that, and there's a stipulation that's a fact, Mr. Musk makes company announcements through Twitter. Still not in the workplace. It's a workplace. And still not directed to employees. He is speaking to work. He specifically talked about workers giving up their stock options. So it's very much a workplace discussion. Well, won't you agree, though, that it is a leap to go from the 1960s, 70s cases in the Twitter in the 2020s? I think it's new technology, but it's not different than a press release, which we had in the 60s. It's not different than a television interview. He's choosing his medium to speak, but it's getting through to the workers, and that is the workers are the most interested in what he has to say. He's announcing new products. He's announcing new policies. He's constantly talking about his company on his Twitter. In the public arena, you said unlike in the public arena. Are you saying Twitter is not a public arena? I'm saying that as the Justice Council said, the public arena and the workplace arena have been distinguished in cases like Gissel, where we're talking about the workers' economic dependence on the employer. And so what the employer says, that sort of threat has an added impact. It's not like we're talking about an election where there's no kind of economic dependence and where the worker can be pressured, coercively pressured, based on what the employer says. Counsel, if I may ask you one final question. Given what you just said about Twitter and the workplace, is there anything that Elon Musk could ever say on Twitter that you would not consider in the workplace context? Of course. If he's talking about his views of unions in general, that is different. If he's talking about a debate about unions in America or a law, that is public debate. Well, here he's talking about his own workplace, the conditions of his workers, the benefits he provides his workers. It's very much a workplace discussion. I'm sorry to ask one quick question. A lot of this is over stock options. Can you name a UAW-organized company where there are stock options for the employees? Aren't those a subject of bargaining? Stock options would be a subject of bargaining. As many know, Ford and GM had contracts with UAW for profit sharing. So I'd say company performance-related pay is definitely part of UAW's philosophy. You're not saying—but he said stock options. Yes, and I can say there are some auto parts plants that have employee stock option plans. I believe in Missouri. I can't see the exact plans, but it's not totally unique. And there, Musk was just wrong. Thank you, Your Honors. Thank you. Quickly, on the Ortiz point, before turning back to the First Amendment issues, Board counsel said that the express finding that the accredited evidence showed that Tesla terminated Ortiz for lying during an investigation doesn't mean what it says, because later in the analysis that the ALJ said it shouldn't even be engaging in, it found that the rationale that Tesla gave was pretext for discrimination. On that point, despite the inconsistency, I really want to focus on the D.C. Circuit's decision in Circus Circus Casinos, which noted in another right-line case that the Board can't simply use the label pretext to assume the conclusion that it needs to establish through substantial evidence. It needs to show actually that the decision-maker wasn't acting based on the stated reason, or that other employees who engaged in similar misconduct were treated dissimilarly. And there's no such showing here. To turn back to the First Amendment speech issue, I think that the DiBartolo case may provide a good way forward for the courts thinking about this, because even Board counsel conceded that it does apply, that rule of constitutional avoidance as to the NLRB does apply when we're dealing with certain categories of speech. Now, I actually think that the text of that decision also applies to the facts of the case. At page 578 of the decision, that's what the Supreme Court said, that the section is open to a construction that obviates deciding whether a congressional prohibition of hand billing on the facts of this case would violate the First Amendment. So I do think that constitutional avoidance extends to facts, not just categories. But even assuming I'm wrong about that, there are several categories of speech that are at issue here where the Board has not previously succeeded in finding a speech violation that it is asking this Court to endorse. One of those, as the discussion this morning has made clear, is the category of speech that's in the digital public square and that is involving not just employees, but predominantly, in fact, public figures and members of the public. And the Board counsel tries to compare this exchange on Twitter to, you know, a poster that's unlawful in the workplace, or the UAW's counsel talks about, you know, how this is no different than a press release. But it is because it involves members of the public engaging about a topic of public concern and deleting the tweet, or even just saying, public figures, you shouldn't engage in that kind of conversation at all, is offensive to the First Amendment and at least is a novel application of the Board's authority. Oh, counsel, you're certainly right. This is a novel application of it. It seems to me if we uphold a categorical barrier where the NLRB cannot consider what's on Twitter, it just creates potential wild west of unfair labor practices, of threats, and whatever else that could never be addressed. And so, rather than dealing it in a more traditional 1960s way, clear threats are put on Twitter. Well, and that's why, Your Honor, in response to Judge Higginson, I didn't say that we are saying you have to endorse that categorical rule. But I'm saying that there are. But isn't that the effect of a categorical rule? Were we to adopt it? Well, I think there are ways of providing guardrails that weren't thought of or even relevant in Gissel that would avoid the complete extension, a whole cloth of Gissel's principles to this very different context, but that would avoid a categorical rule. We've given a few, but I don't think that the Board has given any reason to be completely complacent or uncritically accepting of its proposed expansion. Board counsel couldn't identify a single principle that would prevent the Board from requiring the collection of written materials and destruction of written materials. He just said that the Board hasn't yet ordered that, and he doesn't think they would. But that's not a good safeguard for First Amendment speech. If the Board is claiming the authority or doesn't disavow the authority to destroy speech that is, in its view, overly critical of unions, it's asking for a special First Amendment rule where we don't apply normal safeguards against chilling of unlawful speech. We don't resolve ambiguity in favor of the speaker rather than the censor. We don't have courts apply their normal sort of review to ensure respect for constitutional rights. Gissel doesn't endorse any of that, and this Court shouldn't either. The First Amendment applies to the NLRB as it does any other governmental actor, and under those principles, the speech here, it was not unlawful. Thank you. That will conclude the arguments. This case is under submission.