Last week, I published a lengthy commentary on the recent explosive reports from the Intercept regarding the Lava Jato operation in Brazil—reports that were based primarily on text messages provided by a source who apparently hacked (or otherwise gained unauthorized access to) the cell phone of Deltan Dallagnol, the lead prosecutor in the case. Because I am unable to read Portuguese, my discussion was based exclusively on the two substantive English-language reports, here and here. (There are more reports in the series, but so far they’ve not been translated into English; if and when they are, I may update my commentary.) The Intercept’s reports argued that these leaked text messages indicate: (1) that Judge Moro engaged in unethical and possibly illegal coordinating with and coaching of the prosecutors; (2) that the prosecutors recognized that their case against former President Lula was without solid legal or evidentiary foundation; and (3) that the prosecutors were motivated by political/ideological bias against Lula and his party, the PT.
In last week’s commentary, based on my preliminary analysis of the Intercept stories, and what I knew about the background context, I reached the following tentative conclusions:
- First, I thought that the evidence of extensive text communications between the lead prosecutor and the presiding judge was (or at least should be) per se impermissible. I used very strong language in making this point, describing the fact that the two were in regular text contact as “the height of impropriety,” and a “shocking and inexcusable breach of judicial ethics.”
- Second, though, I thought that the specific text exchanges reported by the Intercept—the ones that allegedly showed the coaching and collaboration—were largely innocuous, and didn’t seem to contain much problematic material over and above the fact of the communications themselves.
- Third, I did not think that the text messages reported by the Intercept provided any reason to call into question the legal and evidentiary basis for Lula’s conviction. That conviction was and remains controversial, but the leaked text messages don’t show anything other than a prosecutor preparing appropriately for his case.
- Fourth, I concluded that although texts exchanged among prosecutors in late September 2018 did indeed indicate that the prosecutors did not want the PT candidate to win the election, this didn’t necessarily show that the prosecutors were biased against the PT back in 2015-2016 (when the decision to investigate and prosecute Lula took place), nor was there any evidence that the prosecutors had taken any concrete action that could be ascribed to partisan bias.
Much to my surprise, last week’s post seems to have attracted a lot of attention, particularly in Brazil. As a result, I’ve had the opportunity to engage in substantive exchanges with multiple Brazilian experts from across the political spectrum, who hold a wide range of views on Lava Jato, Lula, and related matters. Some of these exchanges can be found in the comment section of last week’s post, which I highly recommend that interested readers check out (particularly those who might have read that post the day it came out, before the comment thread included over 60 separate entries); others have communicated with my privately. (To be clear, though, I have not communicated about the post, publicly or privately, with Mr. Dallagnol or anyone else named or discussed in the Intercept story.)
Based on these conversations, and on further reflection, my views on the Intercept’s reporting have shifted somewhat, mainly in the direction of thinking that this “scandal” is considerably less scandalous than the Intercept reported, or that I’d originally believed.
Before I explain why, though, let me note three important caveats: First, as I disclosed in the original post, I have a friendly professional acquaintance with Mr. Dallagnol, and while I don’t think this biases my analysis, readers have a right to know. Second, my discussion here is still based only on the Intercept’s English reports. There have been a number of subsequent stories published in Portuguese, and it may well be that my views might shift based on what’s in these reports once I’m able to read a translation. Third, I’m not an expert on Brazilian law, so it’s entirely possible that my analysis may be flawed due to misunderstandings of the relevant Brazilian rules. (Indeed, it seems this was to some extent true of my original post.)
OK, with those caveats out of the way, let me say a bit more about my current perspective on the Lava Jato Leaks.
First, my conviction that the leaks do not undermine Lula’s conviction is as strong, or stronger, than it was last week.
In last week’s post, I chastised the Intercept for its sensationalist suggestion that the leaks showed that Mr. Dallagnol and his team knew that their case against Lula was baseless. I called this piece of the Intercept’s reporting frivolous. I stand by that. Indeed, I find it telling that among the numerous comments I received, many of them suggesting that I’d understated the extent of the ethical lapses and asserting clear bias against Lula, none offered any serious challenge to my conclusion that the allegedly damning texts regarding the legal and evidentiary weaknesses of the case against Lula showed anything other than a lawyer doing a good job prepping for a hard case. Yes, yes, I know there are still arguments that Lula shouldn’t have been prosecuted or convicted for the alleged corruption involving this beachfront apartment, but my point is not to re-litigate that issue, but simply to stress that the leaks don’t add anything.
Second, I remain unconvinced that the September 2018 texts “lend obvious credibility” (as the Intercept put it) to accusations that the Lava Jato prosecutors are “right-wing ideologues whose overriding mission was to destroy the PT and prevent Lula’s return to power in the 2018 election.”
That is a grave accusation, one that, if lodged by serious journalists, ought to be supported by convincing evidence. But while the stolen text messages do indeed show that by September 2018 the Lava Jato prosecutors (or at least some of them) were bitterly hostile to the PT, they do not show that such hostility (1) existed back in 2015-2016, when the investigation and prosecution of Lula began, (2) influenced any actual prosecutorial decision, or (3) was based on right-wing ideological motivations (rather than, for example, the fact that the PT had been attacking the Lava Jato team in aggressive, often personal terms for years, and had overtly threatened to shut Lava Jato down and undo much of its work).
Now, on that last point, I think it’s worth emphasizing a distinction that much of the current commentary seems to overlook. The question whether the prosecutors (and Judge Moro) treated Lava Jato defendants (including Lula and other PT members) unfairly is different from the question whether the Lava Jato team was politically or ideologically biased against the PT, or the Left more generally. I say this because in response to my claim that texts sent in September 2018 don’t show that there was partisan bias back in 2016, several commenters responded with some version of, “But what about the evidence that Dallagnol and Moro were collaborating back in 2016? Doesn’t that show they were out to get Lula?” The answer is no. Let’s assume for the moment, just for the sake of argument, that it’s indeed the case that the prosecutors and Judge Moro engaged in impermissible collaborations in the case of Lula and other defendants, and that the goal of these impermissible collaborations was to secure convictions. (I’m not convinced that this is in fact true, for reasons I’ll elaborate below, but for now let’s assume it’s true.) This would be really bad, for sure, but it would not be evidence of political/ideological bias. The prosecutor and judge might have been (over-)anxious to convict because they genuinely believed these defendants to be guilty. Indeed, most instances of prosecutorial misconduct, at least in the US (and I suspect elsewhere), arise due to a desire to convict at all costs, not because of the political ideology of the defendants.
Is it possible that the prosecutors and the judge were ideologically biased against Lula and the PT? Sure, it’s possible. But there’s at least one piece of prima facie evidence against that conclusion: The fact that Lava Jato has gone after lots of people from a range of parties across the political spectrum, including the PT’s political opponents (like Michel Temer and Eduardo Cunha). Furthermore, the Intercept’s reports to date actually may imply another piece of evidence against the right-wing-conspiracy view of Lava Jato: the absence of any incriminating text messages (so far) that show a strong anti-PT or anti-Left bias back in the early days of the operation, before the Lula prosecution caused the PT and its supporters to launch their scorched-earth, no-holds-barred attack on Lava Jato. It’s plain that the Intercept reporters have it in for the Lava Jato operation, so I think it’s fair to assume that they’re doing their best to publicize the worst, most incriminating texts they can find. If they’d found texts from 2015-2017 that adopted the same anti-PT tone as the September 2018 texts, I’m sure the Intercept would publish them right away. So, the more time passes without stories revealing such texts, the more I’m inclined to believe that they don’t exist, which would be an exoneration of a sort for the Lava Jato team from charges that their prosecutions were motivated by anti-Left bias.
Third, I think I may have been too quick to condemn the mere existence of the text exchanges between Judge Moro and Mr. Dallagnol as per se illegitimate.
Although I thought that the Intercept had gone overboard in some of its accusations, my original post did agree that Judge Moro likely did engage in a serious breach of judicial ethics in communicating with Mr. Dallagnol privately by text, and that Mr. Dallagnol had committed, at the very least, a lapse of judgment in engaging in these conversations. Indeed, I used very strong language in condemning these conversations.
I now think I was too hasty in reaching conclusion that the text communications were clearly unethical, regardless of their content. To be clear, I am still deeply troubled by the fact that a prosecutor and a presiding judge were regularly texting each other, and the tone of some of the messages does suggest an overly collaborative tenor to the relationship. But in my original post, I’d written (among other things) that “for a judge to engage in secret, ex parte communications with a prosecutor … regarding a pending case is the height of impropriety—full stop,” and I now think that that may not be quite right, at least in this context, for two reasons.
- First, and more importantly, on closer review it appears that most, perhaps all, of the private communications in question were in the context not of the trial phase, but rather the investigation phase—when the prosecutor was doing things like applying for warrants, subpoenaing witnesses, etc., which the judge has to approve. In this context, the prosecutor and the judge may (indeed must) engage in secret ex parte communications. And I gather, based on some of the comments I received, that these exchanges are likely to be more frequent in Brazil than in other jurisdictions due to certain features of Brazilian procedural law that require the judge to be more closely involved in overseeing the investigation phase than is true in other countries. It’s a bit unusual for these communications to take place via text message, to be sure. But it’s not unheard of for US prosecutors to apply for warrants over the telephone, so this doesn’t seem like that much more of a leap. That doesn’t mean that any communications that take place in this context are ethically fine, and I’ll say a bit more on this in a moment. It does, however, mean that I was wrong to say there is or ought to be per se bar on secret private communications between a prosecutor and judge. Even if that were true at the trial/litigation phase, it isn’t and couldn’t be true at the investigation phase.
- Second, Brazilian law doesn’t seem to prohibit substantive ex parte communications even at the litigation phase, so long as both sides have an equal opportunity to engage in such communications. In the US, the American Bar Association’s model code of judicial ethics flatly forbids such contacts except under a narrow range of circumstances, and I wrongly assumed that something like that standard also applies in Brazil. But Brazil is apparently more liberal regarding ex parte contacts between judges and lawyers (see Federal Law 8.906/94, Article 7, Section VIII). I don’t think that section applies to these communications, in part because Judge Moro was an active participant rather than simply letting Mr. Dallagnol make his case. But I feel like I should correct the record, since I’d wrongly assumed that the fact of the text contact about the case, even without looking at the content, would establish an ethical violation, but that seems not to be true in Brazil.
Now, the fact that extensive ex parte contacts between the prosecutor and judge at the investigation phase are unavoidable raises a tricky legal problem, which I don’t think most of the existing commentary (including my prior post) really grapples with adequately. It goes like this:
- Brazil, section 254 of its criminal procedure code, sensibly prohibits a judge from presiding over a trial if the judge has given legal advice to any party to the case.
- However, in the context of overseeing a criminal investigation, it may not make sense for the judge to be wholly passive; rather, the judge should give feedback to the prosecutor, including warning the prosecutor when his or her investigative plan may run afoul of legal rules, or explaining to the prosecutor the sort of evidence that must be produced to justify the authorization of intrusive investigative measures. Think about some of the things Judge Moro appears to have said to Mr. Dallagnol, along the lines of (and here I’m paraphrasing, but I think fairly): “Make sure the police doing these investigations follow the legal rules,” “Be careful about filing that request unless you have substantial evidence to support it,” “If you want to talk to that witness, make a formal request through the regular channels.”
- The problem is that it’s conceivable that the latter sort of guidance might be construed as “legal advice.” But maybe it doesn’t, or shouldn’t, given how prosecutors and judges interact in the context of conducting a complex investigation. As I can’t emphasize enough, I’m not an expert on this subject. I’d very much welcome any commentary from those who can explain how Brazilian law has typically resolved this tension. At the very least, it seems that the case that Judge Moro overstepped ethical lines is far from clear.
- The more general problem here is that by involving the judge so heavily in overseeing the investigation, the judge may start to identify overmuch with the prosecution, and develop an overly collaborative relationship with the prosecutors. That may well be a problem, but if so it’s not so much an ethical failing by this particular judge (or these particular prosecutors), but rather a structural problem with Brazilian criminal procedure.
Fourth, the most potentially problematic text exchange between Judge Moro and Mr. Dallagnol is sufficiently ambiguous that I’m not prepared to say that it was unethical.
In my original post, I went through the content of the Moro-Dallagnol text messages quoted in the Intercept’s English reporting and concluded that “none of them seem to show … procedural irregularities that might undermine the fairness of the trials, other than the fact of the communications themselves.” Several commenters argued that I’d failed to fully comprehend the allegedly most serious (and certainly most confusing) of the exchanges at issue.
Because this is so hard to understand (at least to me), here’s the key passage from the Intercept story, reproduced in full:
Another example of Moro crossing the line separating prosecutor and judge is in a conversation with Dallagnol on December 7, 2015, when he informally passed on a tip about Lula’s case to the prosecutors. “So. The following. Source informed me that the contact person is annoyed at having been asked to issue draft property transfer deeds for one of the ex-president’s children. Apparently the person would be willing to provide the information. I’m therefore passing it along. The source is serious,” wrote Moro.
“Thank you!! We’ll make contact,” Dallagnol promptly replied. Moro added, “And it would be dozens of properties.” Dallagnol later advised Moro that he called the source, but she would not talk: “I’m thinking of drafting a subpoena, based on apocryphal news,” the prosecutor said. While it is not entirely clear what this means, it appears that Dallagnol was floating the idea of inventing an anonymous complaint that could be used to compel the source to testify. Moro, rather than chastise the prosecutor or remain silent, appears to endorse the proposal: “Better to formalize then,” the judge replied.
The Intercept’s interpretation thus seems to be that Mr. Dallagnol proposed inventing a (fake) anonymous complaint that could be used to compel a witness to testify, and that Judge Moro went along with this, telling him to “formalize” the request. That sounds outrageous. But it’s not clear this is the right interpretation of the exchange. One of the commenters on my original post rendered the entire exchange (which I gather is in the Portuguese version) into English, as follows:
Moro (17:42:56): So. The following. Source informed me that the contact person is annoyed at having been asked to issue draft property transfer deeds for one of the ex-president’s children. Apparently the person would be willing to provide the information. I’m therefore passing it along. The source is serious.
Deltan (17:44:00): Thank you!! We’ll make contact
Moro (17:45:00): And it would be dozens of properties.
Deltan (18:08:08) I called him, but he walked back on his remarks. He said he had nothing to say etc… when I applied some pressure, he hung up on me… I’m thinking of drafting a subpoena, based on apocryphal news.
Moro (18:09:38): That’s weird, since it was him who would have alerted the persons who talked to me. Better to formalize then.
Moro (18:15:04): Supposedly he would have spoken with [REDACTED], who then passed the information ahead until it arrived here.
Deltan (18:16:29): May I identify this intermediary source?
Moro (18:59:39): I’m unsure about this now.
Moro (19:00:22): Perhaps you should talk with [REDACTED] beforehand.
Deltan (20:03:00): Ok
Deltan (20:03:32): Ok, thanks, I’ll call
It’s hard to figure out what to make of this exchange, not least because the most crucial phrases–“apocryphal news” and “better to formalize”–are ambiguous, at least to me. It’s possible that the Intercept’s interpretation is correct. As I noted in my original post, it’s strange (at least so it seems to me) that a judge is passing on a witness tip to a prosecutor in the first place. But putting that aside, it’s possible to interpret this exchange as follows: Person X tells Intermediary Y that he has relevant evidence against Lula and might be willing to testify. Intermediary Y tells this to Judge Moro. Judge Moro can’t do anything with this information himself, so tells Mr. Dallagnol. Mr. Dallagnol calls Person X, who refuses to cooperate. Mr. Dallagnol relays this to Judge Moro, and says he’s thinking of subpoenaing Person X’s testimony, based on an anonymous tip that Person X has information. This is not fully accurate, since Mr. Dallagnol knows the identity of Intermediary Y. But it’s not the case that the tip is “fabricated,” because there really is a source—Intermediary Y. It seems Mr. Dallagnol is contemplating drafting the application for a subpoena without using Intermediary Y’s name. Judge Moro then tells Mr. Dallagnol to proceed through formal channels (which I assume must mean drafting a subpoena rather than making an informal phone call). But Mr. Dallagnol then asks if it might be better to use Intermediary Y’s name in the subpoena, rather than treating his tip as anonymous. (That is, to the extent that Mr. Dallagnol’s initial instinct to say the tip was anonymous might have been improper, he almost immediately backs away from that idea and asks Judge Moro about that.) Judge Moro advises him to get in touch with Intermediary Y directly, which seems like a sensible thing to do under the circumstances. And then, so far as we can tell, nothing ever happened. Neither the Intercept nor any other commenter has pointed to any evidence that Mr. Dallagnol ever subpoenaed this witness, with the tipster’s name or otherwise.
Was this unethical? Honestly, I have no idea. The only thing that’s clear to me is that the answer isn’t clear. Some questions to which I don’t know the answers:
- Is it ethical in Brazil for a judge who receives a tip that a witness may have evidence relevant to the investigation to pass that information on to the prosecutor? Or would the judge be ethically obligated to tell the source he/she must contact the prosecutor directly?
- Is it ethical in Brazil for a prosecutor to file a subpoena for a witness based on a genuine, credible tip that the witness has information, but stating the source of this tip is anonymous even when the prosecutor (and the judge) knows the tipster’s identity?
- Assuming the phrase “apocryphal news” does mean anonymous/unidentified in this context, does a subpoena based on “apocryphal news” imply that the identity of the source is in fact unknown to the prosecutor (and judge)? Or is it just a way of saying that the formal filing will not list the informant’s name (that is, it’s more like a “confidential informant” than an “anonymous informant”)?
If anyone out there knows the answers to these questions, it would be extremely helpful!
Oh, also, one more thing: While the Intercept says that, when presented with Mr. Dallagnol’s allegedly unethical plan, Judge Moro, “rather than chastise the prosecutor or remain silent, … appears to endorse the proposal,” this is a bit unfair. True, Judge Moro initially responds with the ambiguous phrase “better to formalize,” but less than an hour later he tells Mr. Dallagnol that he should talk with the intermediary first. And, again, so far as we know nothing happens after that (or at least the Intercept hasn’t yet reported it).
This brings me to an additional, more general observation here: While I have great respect for investigative journalists, and I agree with the Intercept’s argument that in this case these leaked materials are of sufficient public importance that they can be disclosed in the context of this sort of story, I am disappointed in a number of aspects of the Intercept’s coverage of this issue. I’ll put to one side the fact that the Intercept, disregarding the usual practice of professional journalists, did not contact the subjects of the story to give them an opportunity to respond prior to publication—which the Intercept claims was because they had a credible fear that the Brazilian government would react by enjoining publication. Even if we give the Intercept the benefit of the doubt on that one, there’s also the fact that the Intercept appears not to have consulted with any experts on Brazilian law or legal/judicial ethics, instead appearing to believe that the reporters’ own judgments regarding what is permissible under Brazilian law are correct and do not require substantiation. So, the stories say over and over, in conclusory fashion, that Moro’s communication with Mr. Dallagnol “is unethical for a judge” and “violates the Judiciary’s Code of Ethics,” that the leaked texts “reveal serious ethical violations,” etc. There is no hedging language (“possibly,” “arguably,” “probably,” etc.), nor are there any quotes from legal experts supporting the claims of unlawful conduct. The Intercept stories don’t even bother to cite or quote the relevant provisions of the law or ethics code that were supposedly violated. With all due respect to the Intercept’s journalists, this seems to me rather arrogant.
And that’s of a piece with the more general editorializing, and occasional gratuitous potshots, scattered throughout the pieces. Just to take one trivial but grating example,when quoting texts in which Mr. Dallagnol discusses praying for a certain result, the Intercept declares that Mr. Dallagnol “often boasts of his religious piety,” an unnecessarily snide way of saying that Mr. Dallagnol often speaks openly about his religious faith. (And while Mr. Dallagnol may have his faults, as do we all, I’ve never known him to be boastful about anything.) That’s but one example among many.
Two final points here before closing:
- First, I regret to say that this story and its reception has demonstrated that those of us who thought Judge Moro was making a huge mistake—and threatening to undermine the reputation of the Lava Jato operation as a whole—when he accepted the post in Bolsonaro’s cabinet were right in their fears. Both the original Intercept story and virtually all of the subsequent commentary have emphasized that this appointment casts doubt on Moro’s impartiality from the beginning, and might even suggest a quid pro quo. Imagine how different the impact of the Intercept revelations would be if Moro had not only declined the nomination, but had done so with some sort of public statement that it would be inappropriate for him to accept the appointment because of the importance of maintaining the neutrality of Lava Jato. Had that happened, the Intercept leaks would still have made Judge Moro look bad, but his defenders would have found it easier to excuse the possible improprieties as the result of excessive zeal in the anticorruption cause (much as many, to the PT’s chagrin, were willing to forgive Judge Moro’s release of the secret recordings of the Dilma-Lula conversations). In that alternative universe, Judge Moro’s conduct in the Lava Jato cases wouldn’t have looked like it was motivated by ideology or desire for advancement—now it does, whether or not that’s actually the truth.
- Second, though the focus right now is understandably on the content of the leaks, at some point people need to start worrying about how the Intercept’s source got hold of all that data on Mr. Dallagnol’s cell phone in the first place, and what their motivations are. This is a big-time data theft, a serious felony. Pause for a moment and think about how you’d feel if you know that someone out there had all of the information on your phone, going back at least five years. Now think how you’d feel if the people who had that information had a powerful ideological/political interest in discrediting you. And though the Intercept reporters are honorable journalists who are careful not to disclose private material simply for the sake of embarrassing people, we have no way of knowing what else the original source might do with the data. Is this an attempt to shut down Lava Jato via a veiled blackmail threat, if the initial adverse publicity from the Intercept stories doesn’t suffice? Is the goal specifically to shut down Lava Jato, or is it to further polarize and destabilize Brazilian politics more generally? Whatever the public interest in writing stories based on these leaks, and I agree with the Intercept that there is such an interest, the fact that sophisticated hackers are targeting the personal cell phones of public figures and releasing the information in order to either undermine an anticorruption investigation or destabilize a country is quite scary.
In closing, and circling back to the main focus of the discussion, I’m beginning to wonder whether theere’s much of a scandal here after all. Yes, I’m concerned about what looks like an overly chummy relationship between the lead prosecutor and the presiding judge. And yes, the leaked material disclosed so far (at least in the English versions) contains some troubling material, which might suggest ethical violations. But even the most problematic text exchanges contain considerable ambiguities, such that the case for an ethical violation, while plausible, is not nearly the slam-dunk that the Intercept made it out to be. Indeed, when you strip away the rhetoric, insinuations, and rehash of material already in the public record, it’s not at all clear that the leaked texts reported so far (in English) actually demonstrate serious wrongful conduct.
That view is highly tentative. I know that many readers will likely disagree, and I invite, indeed request, that those of you who think I’m wrong share your critique of my analysis in the comment section below. I’ve learned a lot from the feedback on the earlier post, and I appreciate the substantive, respectful, informative discussion that seems to be underway. It should also go without saying that, as more evidence comes out (and/or is translated into English), I will update my views accordingly. Nobody should rush to judgment in this case.