A Primer on How to Read: So the NYT Can Stop Telling Paul Manafort’s Lies

NYT Continues to Tell Paul Manafort’s Lies for Him

It has been two and a half days since I pointed out that their single anonymous source — described as “a person knowledgeable about the situation” — lied to the NYT last month when it reported that Paul Manafort and Rick Gates shared poll data, “most of which was public,” “in the spring” with Konstantin Kilimnik.

Both Mr. Manafort and Rick Gates, the deputy campaign manager, transferred the data to Mr. Kilimnik in the spring of 2016 as Mr. Trump clinched the Republican presidential nomination, according to a person knowledgeable about the situation. Most of the data was public, but some of it was developed by a private polling firm working for the campaign, according to the person.

The NYT has not corrected the error and identified who turned them into a vehicle for significant propaganda.

Instead, two of the same journalists, plus Scott Shane, wrote a story they say is based on “A closer look at the transcript” focusing on the Ukrainian stuff that had already been revealed in significant detail last month.

In it, they correctly identify roughly where the beginning of the poll sharing discussion starts, but describe a lie that Manafort corrected — the same lie the NYT continues to tell — as the final testimony of Manafort.

The transcript suggests that Mr. Manafort claims that he wanted only public data transferred. But Mr. Weissmann told the judge that the question of whether any American, wittingly or unwittingly, engaged with Russians who were interfering in the election relates to “the core” of the special counsel’s inquiry.

They don’t mention that the judge, Amy Berman Jackson, Andrew Weissmann, and even Manafort’s lawyer Richard Westling, all acknowledge this was not just public polling data.

And then they present a comment by ABJ that was about Manafort’s poll sharing lies and suggest (in a story focused on the Ukraine peace deal) it generally relates to Manafort’s comments on Kilimnik.

Judge Amy Berman Jackson seemed to agree with prosecutors that whether Mr. Manafort lied about his contacts with Mr. Kilimnik was important, saying at one point, “I am, actually, particularly concerned about this particular alleged false statement.”

Because the NYT is struggling so much, as a service to them (and in hopes they expose whoever lied to them), I’m going to provide a primer on how to read redacted documents so they don’t have to continue to be a mouthpiece for Paul Manafort.

Identify How this Document Fits into the Pattern

We have a lot of tools with which to read Manafort’s breach hearing transcript, largely because it is part of a series. As I noted in my post laying out the lies NYT continues to tell, there are four prior versions of this discussion.

The hearing is a discussion about the arguments made in all these earlier documents, particularly the last two, which means we can look to them to understand what we’re seeing in the transcript.

All four discuss the same five topics. Though, as Judge Amy Berman Jackson notes, the lawyers have not remained consistent in the order in which they discuss them. (Note to Judge ABJ: I was also annoyed by that. Thanks for razzing them about it!)

This morning I’m going to organize myself by the issues the way they were numbered in the initial declaration. It was great because in every pleading, you all numbered the five issues into different orders. So I can’t really call them Issue No. 1 and Issue No. 2, but that’s the template I’m going to use. And what I’m going to do is, I’m going to hear from both sides on each issue before I move on to the next issue.

Thankfully, ABJ is more helpful at providing guideposts than the lawyers, though to clarify, when ABJ says she’s using the “initial declaration,” she’s referring to this FBI declaration, not Mueller’s original breach filing. You can tell that’s the case because it’s called a “declaration” and because it starts the same way ABJ does, with Manafort’s lies about the kickback payment.

The Structure and Content of Past Filings

To read a document that is the fifth in a series it’s helpful to map its structure and understand how that structure compares to previous iterations in the series.

I laid out the structure of the declaration ABJ says she’s following in this post. Here’s an updated version in which I’ve included some of what past documents refer to as proof and timing.

I) Kickback to/from Rebuilding America Now (0-series exhibits)

Firm A to receive 6% commission from Firm B (12/7)

After a break, it became clear that the government’s facts were incorrect – it was a $125,000 payment. (1/8 filing)

Manafort explained that it was unclear to him how this payment was recorded by his accountants and he believed the original plan was to report the payment as a loan, but that it had actually been reported as income on his 2017 tax return. The Government has indicated that Mr. Manafort’s statements about this payment are inconsistent with those of others, but the defense has not received any witness statements to support this contention. (1/8 filing)

Three false statements, first that the other people had paid him, then conflicting statements from the others (but Manafort recording the payment as income), finally that it was a loan (supported by loan documents provided at that time) (1/15 filing)

Three false statements, the last being that it was a loan

II) Konstantin Kilimnik’s role in witness tampering (100-series exhibits)

During a proffer session with the OSC on October 16, 2018, Mr. Manafort acknowledged that he and Mr. Kilimnik agreed to reach out to the witnesses. Mr. Manafort was asked to agree that Mr. Kilimnik, too, possessed the requisite state of mind to legally establish his guilt. Mr. Manafort balked at this characterization, because he did not believe he could confirm what another person’s internal thoughts or understandings were, i.e., another individual’s state of mind. (1/8 filing)

Kilimnik didn’t think he had exerted pressure (1/15 filing)

Manafort expressing Kilimnik’s views (1/23)

III) Interactions with Kilimnik (200-series exhibits)

a) Discussions of the Ukraine Peace Deal

Manafort “conceded” that he discussed or may have discussed a Ukraine peace plan with Mr. Kilimnik on more than one occasion (1/8 filing citing 12/7 one)

Issues and communications related to Ukrainian political events simply were not at the forefront of Mr. Manafort’s mind during the period at issue and it is not surprising at all that Mr. Manafort was unable to recall specific details prior to having his recollection refreshed. (1/8 filing)

Beginning August 2 and continuing until March 2018 Kilimnik and Manafort communicated about Ukraine peace plan. Three discussions were in person (1/15 filing)

Manafort freely brought up August 2 meeting, didn’t think one plan would work

1) August 2 meeting

Manafort would have given the Ukrainian peace plan more thought, had the issue not been raised during the period he was engaged with work related to the presidential (1/8)

Discussed at September 11 and 12 debriefings, then in grand jury on October 26, he admitted he saw the email (1/15 filing)

2) December 2016 meeting

Discussed September 11, September 21, October 26 (1/15)

3) Madrid meeting

Admitted it after shown evidence (12/7)

After being told that Mr. Kilimnik had traveled to Madrid on the same day that Mr. Manafort was in Madrid, Mr. Manafort “acknowledged” that he and Mr. Kilimnik met while they were both in Madrid (1/8 filing, citing 12/7 one)

September 11, 12, 13, October 26 (1/15)

Manafort lied, claimed it was about a business investment, then was shown something, and then admitted it (1/23)

4) A 2018 proposal

Not brought up prior to GJ (1/23)

b) Manafort’s false statements about sharing polling data

Email and testimonial evidence (12/7)

The same is true [he needed his memory refreshed] with regard to the Government’s allegation that Mr. Manafort lied about sharing polling data with Mr. Kilimnik related to the 2016 presidential campaign.

Evidence = interviews (plural) with gates, Gates on his access, multiple emails, on the morning of the meeting (at which Gates came late) (1/15 filing, ¶¶53-36)

Whether he told anyone to do something, SCO relies on Gates’ testimony (1/23)

IV) Another DOJ investigation (possibly that of PsyGroup or the hush payments) (300-series exhibits)

Another district (12/7)

Manafort provided the government with information pertinent to an investigation in another district prior to entering into the plea agreement in this case but then, in post-plea proffer meetings with other prosecutors not associated with the OSC, provided a different version of the same events. (1/8 filing)

One version on September 13, another on October 5, largely retracted second version; series of text messages, prior to leaving campaign (1/15)

Corrected in same interview (1/23)

V) Manafort’s contact with the Administration (400-series exhibits)

Text May 26, 2018 (12/7)

There is no support for the proposition that Mr. Manafort intentionally lied to the Government. The first alleged misstatement identified in the Special Counsel’s submission (regarding a text exchange on May 26, 2018) related to a text message from a third-party asking permission to use Mr. Manafort’s name as an introduction in the event the third-party met the President. This does not constitute outreach by Mr. Manafort to the President. The second example identified by the Special Counsel is hearsay purportedly offered by an undisclosed third party and the defense has not been provided with the statement (or any witness statements that form the basis for alleging intentional falsehoods). (1/8 filing)

May 2018 effort, targets (1/15)

Misread text messages, Gates claim (1/23)

By mapping out what the prior versions of the series look like, we can put this transcript into the structure ABJ has told us we’re dealing with, to identify with certainty which discussion is which.

The Structure of the Breach Hearing Transcript

Now we can identify the structure of this document, which will help us identify the boundaries between these parts of the discussion.

After some introductory legal discussions, ABJ teaching Andrew Weissmann how to use a microphone, and then Weissman framing why they think Manafort is a lying turd trying to get a pardon but honestly they did engage in a good faith effort to get him to cooperate, the substantive discussion starts.

ABJ tells us she is using the “way they were numbered in the initial declaration” and she’ll hear from both sides before she moves on to the next issue.

So here’s what the structure of the breach hearing looks like. Importantly, while Weissmann addresses a few issues at the beginning (which are noted), otherwise the discussions have clear start and end points, meaning we know that what appears between those start and end points pertain to the topic at hand.

I. Kickback to/from Rebuilding America Now

Start: Page 25, line 18: “With respect to the $125,000 payment by”

End: Page 47, lines 6-9

THE COURT: All right. I think you made that clear. And I think I understand everybody’s point of view about this, and what the evidence is. But, there’s some aspects of the evidence I’m going to need to re-review.

Also page 14-15, 20

II. Konstantin Kilimnik’s role in witness tampering

Start: Page 47, lines 10-12

All right. So let’s go on to what is II, or the second subject touched upon in the declaration, which is Mr. Kilimnik’s role in the obstruction conspiracy.

End: Page 63, lines 3-5

THE COURT: All right. Well, I don’t think I need any more of your telling me what it says because I’m going to read it again.

Also pages 14, 20

III. Interactions with Kilimnik

Start: Page 63, lines 5-8

So let’s go on to III, the interactions with Kilimnik, which I think I’m going to break up a little bit into the Ukraine stuff and the [polling] stuff.

a. Discussions of the Ukraine Peace Deal

Start: Page 63, lines 9-10:

With respect to the first, sort of, subtopic here, the discussions concerning the [redacted] Ukraine

End: Page 82, lines 9-13:

THE COURT: Right. But, I think what gives them cause to be theorizing is the fact that it’s described differently on different occasions, and described inconsistently with the communications between Mr. Kilimnik and Mr. Manafort, and that leads them to wonder.

b. Manafort’s false statements about sharing polling data

Start: Page 82, line 14-15:

But, I think we can go on to the question of the [polling]

End: Page 110, lines 13-17:

THE COURT: All right. I mean, when I asked you, do you want to hear from him, you said you wanted to file something. I just want to make sure you’re saying we’re done; when this record is concluded, we’re done with the record.

MR. DOWNING: Correct.

Also, pages 18-19,

IV: Another DOJ investigation

Start: Page 110, lines 20-21:

Okay. I think we can go on to category IV, the other DOJ investigation.

End: Page 121, line 18:

THE COURT: All right.

V: Manafort’s contact with the Administration

Start: Page 121, line 18-19:

Well, that leads me into No. 5, the contacts with the administration.

End: Page 132, line 5-6:

THE COURT: All right. That covers all the subject matter areas.

Validate the Model

Now it helps to make sure this model does match the prior model.

Unfortunately, the issue that NYT is perpetuating Manafort’s lies about — the sharing of polling data — is one for which we don’t have that many signposts in past filings (because this discussion is so heavily redacted). But the key dispute is clear from past filings. The government maintains the evidence includes Gates’ testimony and email evidence, while Manafort would like ABJ to believe the government is relying exclusively on Gates’ testimony.

The same is true [he needed his memory refreshed] with regard to the Government’s allegation that Mr. Manafort lied about sharing polling data with Mr. Kilimnik related to the 2016 presidential campaign.

Evidence = interviews (plural) with Gates, Gates on his access, multiple emails, on the morning of the meeting (at which Gates came late) (1/15 filing, ¶¶53-36)

Whether he told anyone to do something, SCO relies on Gates’ testimony (1/23)

Thankfully, this is a place where Weissmann’s earlier comments provide another point of validation. At the beginning, he uses the polling data as an example to refute the defense claim they had engaged in a “gotcha,” by not providing them Gates’ prior statements on the issue.

But, I wanted to address that that’s not — this is an unusual case. This is an unusual case. Not because we did that, it’s an unusual case because of the volume of evidence that the defendant had. As the Court knows, there was a trial in the Eastern District of Virginia. And as the Court knows, there was a discovery order in this case. There, the vast, vast majority of information was available to the defendant. And as one of the submissions having to do with bail conditions and — or, prison location, what’s in the record is that the defendant, on tape, in prison, says yes, he has been through all of that discovery. So, for one example of that, all of the Gates 302s that were extant in September of last year were something that had been disclosed to the defendant. So, the defendant was very well aware of what Mr. Gates had said about sharing of polling data, and that it was something that was not — not simply a matter of [redacted]. And it sort of [redacted].

In the section devoted to the topic, we see several of the things that show up earlier: Manafort ordering Gates to do something, and the reliance on Gates’ testimony.

Whether he told anyone to do something

Manafort asking Mr. Gates


THE COURT: And because Mr. Manafort told Mr. Gates to do it?

MR. WESTLING: That’s what Mr. Gates says, yes.

THE COURT: In an e-mail.

MR. WESTLING: But I think that the e-mail says, Please print this. That’s all it says.

THE COURT: Doesn’t it say bring it to the meeting?

MR. WESTLING: I’m sorry?

THE COURT: Doesn’t it say bring it to the meeting?

MR. WESTLING: It says related to a scheduling meeting. Doesn’t say anything about a meeting with Mr. Kilimnik, it doesn’t say anything about — just on the same date.


THE COURT: The only thing I said that corroborated his testimony about this matter was the e-mail within — related to on this date. Is that correct?


THE COURT: And you’re saying read more carefully, Judge, because it doesn’t say [redacted] to the meeting. So I will do that, but —

MR. DOWNING: I doesn’t say that, Your Honor —

THE COURT: — I do believe that that is corroborative.

Reliance on Gates’ testimony

they believe because Mr. Gates says so and because it’s referred to in Mr. Kilimnik’s various emails

Multiple references to whether they’ve gotten the 302s in question

Kevin Downing’s repeated attempt to suggest Gates couldn’t be credible because the jury didn’t find him credible (even while being careful to avoid having Gates testify to refute that).

Weissmann’s description of the earlier 302s they had in time for the EDVA trial.

Process New Information

Having now validated that that discussion pertains to the sharing of polling data question, we can now turn to what else new we learn in it.

There’s Weissmann’s description of Manafort telling the grand jury he understood someone was going to be sharing the data with some entity and some individual and that  considered that a win-win for himself (which is why I say Manafort sold Trump out, because he figured even if this didn’t help Trump win, he would still curry favor with his Ukrainian and Russian paymasters).

which he admitted at that point was with — he understood that it was going to be given by [redacted] to the [redacted] and to Mr. [redacted], both. That from his perspective, it was — there was no downside — I’m paraphrasing — it was sort of a win-win. That there was nothing — there was no negatives.

There’s ABJ’s question about why the pollster was getting paid so much if this was no big deal.

And if that’s true, then why was [redacted] being paid so much

Westling responds by trying to argue that it’s no big deal because the data is so detailed it would be so incomprehensible to him.

This is very detailed [redacted] on a level that is very focused

Which ABJ says is why the polling data is so important.

THE COURT: But if I determine that it is established by the record and in his statement — but that’s what makes it significant and unusual.

Whereupon Westling (again, this is Manafort’s defense attorney!!!!) says that sharing data would be beneficial if it were something more public, effectively refuting the claim Manafort tried to make, which is the claim the NYT refuses to correct.

if the goal were to help Mr. Manafort’s fortunes, that some other kind of [redacted] something more public,

Then, in an effort to suggest this was just about the campaign meeting that morning, Westling says this was the most recent data.

it was the most recent, from what we can tell, the most recent

Then ABJ corrects Westling’s claims about timing, noting that Gates specifically tied this to the Havana Club meeting.

THE COURT: Didn’t he say it happened at the meeting where they had to leave by different doors and all that? Doesn’t he connect [redacted] to the meeting and the Havana Club and the coming and going

Weissmann explicitly supports this timing later in the hearing (with what seems to be a description of Manafort walking Kilimnik through what the data showed).

And then Mr. Gates, in — I think I referred you to 236, on page 3, Mr. Gates talks about the August. 2nd meeting and actually has Mr. Manafort walking Mr. Kilimnik through

And Weissmann returns to this, once again making it clear the data sharing happened on August 2.

Both of those refer to [redacted] and also refer to the discussions of the — discussions of [redacted] at the August. 2nd, 2016 meeting.

ABJ also notes there is some kind of ex parte information that she has seen that the government can’t share with the defense.

THE COURT: I need to ask the Office of Special Counsel about something ex parte because — and so I apologize for that, but I need to do that. And it may be after I talk to them, they tell me there’s no problem with sharing it with you.

This information seems to give ABJ further confidence that the government is telling the truth here.

The NYT Continues to Tell Paul Manafort’s Lies

So to repeat: both ABJ and Andrew Weissmann make it clear that on the morning of August 2, 2016, Manafort told Gates to print out some polling data. Later that day, they clandestinely meet with Konstantin Kilimnik, where they discuss both a “peace” deal in Ukraine — which Manafort admits amounts to sanctions relief — and the polling data. Indeed, Weissmann claims that Gates said Manafort walked this guy, with ties to the same Russian intelligence agency that was still hacking Hillary Clinton, through that very complex and recent polling data.

And the fact that the data was so complex, according to ABJ, is “what makes it significant and unusual.” Indeed, even Manafort’s own lawyer suggests this is not public information, which is one of the things he tries to argue would suggest Manafort wasn’t trying to benefit himself.

When the NYT says this:

The transcript suggests that Mr. Manafort claims that he wanted only public data transferred. But Mr. Weissmann told the judge that the question of whether any American, wittingly or unwittingly, engaged with Russians who were interfering in the election relates to “the core” of the special counsel’s inquiry.

They are not telling their readers that Richard Westling, in an attempt to defend Manafort, made it very clear this was not public data.

And when the NYT suggests that this comment pertains to Manafort’s interactions with Kilimnik generally,

Judge Amy Berman Jackson seemed to agree with prosecutors that whether Mr. Manafort lied about his contacts with Mr. Kilimnik was important, saying at one point, “I am, actually, particularly concerned about this particular alleged false statement.”

They need to acknowledge that the comment comes from a paragraph (on page 103, the section exclusively dedicated to a discussion of the polling data) that focuses on the defense effort to discredit Gates’ polling data testimony by claiming they hadn’t gotten his 302s from January 2018.

THE COURT: All right. So, whether we need to have a hearing on that because I am, actually, particularly concerned about this particular alleged false statement. But I also think we need to think about what the purposes I’m being asked to find whether or not this is, what the burdens are, etcetera. So, you’re entitled to think about it, although I don’t think this has come as a surprise, that this was the issue, since this was the only evidence they pointed to as the fact that this fact was false, was Mr. Gates’s 302s and the e-mail.

Manafort doesn’t want this public because he knows it’ll kill his chance for a pardon

Here’s why I just wasted so much time trying to teach the NYT to read (aside from the fact that the NYT probably “corrected” a story that was initially correct, that this data got shared with Oleg Deripaska, which is made more obvious once you stop telling the lie that the data got shared in the spring).

This is an area where Weissmann specifically suggests Manafort was lying last fall to sustain his chance for a pardon.

the other motive that Mr. Manafort could have, which is to at least augment his chances for a pardon.

Paul Manafort doesn’t want the public to know he gave highly detailed polling data to a GRU-tied Russian, Konstantin Kilimnik, at a clandestine meeting he may have flown home from on Oleg Deripaska’s plane. He doesn’t want the public to know that because it’ll kill his chance for a pardon.

And for some unfathomable reason, the NYT doesn’t appear to want the public to know that, either.

As I disclosed last July, I provided information to the FBI on issues related to the Mueller investigation, so I’m going to include disclosure statements on Mueller investigation posts from here on out. I will include the disclosure whether or not the stuff I shared with the FBI pertains to the subject of the post. 

68 replies
  1. pseudonymous in nc says:

    Dear god, it’s frustrating. I can’t imagine what’s going on behind the scenes: whether it’s a stubborn stick-up-ass refusal to correct their own reporting until it’s completely fucking obvious that they fucked up, or because if they do correct their reporting, their “person knowledgeable about the situation” will be offended and stop talking to them, even though what that person is telling them isn’t true. (Burn shit sources.)

    As I said in another thread, I’m still wondering why Paulie’s lawyers wanted to raise Gates’s campaign activities in the EDVA cross-examination — where Andres’s objection/sidebar was specifically about the August 2 polling data exchange — if Manafort’s telling the truth about it would damage his chance of a pardon. Perhaps the aim was to implicate Gates without tainting Paulie, but that’s a narrow needle to thread given that Gates’s 302s were locked in. Maybe the intention was just to blow up the case entirely. [ETA: Downing suggested he should cross-examine Gates in front of ABJ, so he’s clearly up for it.]

    That said, it’s sort of funny that Manafort may have had the world’s biggest burner iPod touch collection but appears to be as shit with printers as he was with doctoring PDFs and having Gates serve as his IT person got him in deep shit.

    • emptywheel says:

      My read is that Downing was calling ABJ’s bluff on Gates, pretending he would like to cross examine him in front of her, but ultimately balking when she said, “is the record done or do I need to hear from him.” He knows Gates would come off more credibly than Manafort.

      I assume you remember I wrote about that objection in real time?

      • pseudonymous in nc says:

        Oh yes. I immediately thought of that post when I saw Weismann refer to the sidebar. But I’m not sure how your speculation then–

        Had he succeeded, perhaps Trump would have recognized the jeopardy that put Manafort (and, presumably, himself) in. Perhaps he would have taken that moment to pardon Manafort, and save him from that jeopardy.

        squares off with Weissmann saying that Manafort’s bullshit narrative after taking the plea (vs whatever Gates might have testified under cross-examination) was to protect his chances of a pardon.

        The EDVA transcripts aren’t on RECAP any more, so I can’t go back and check the exact language, but looking at Zoe Tillman’s account, perhaps the intent was to get Gates on record about the polling data and then try to attack his credibility on that subject. But again, that seems like a high-stakes gamble.

        And yeah, Downing was bluffing. The whole spiel about Gates and the EDVA jury was pound-the-table lawyering, and ABJ was very not amused.

  2. AndTheSlithyToves says:

    “And for some unfathomable reason, the NYT doesn’t appear to want the public to know that, either.”

    Maybe the Washington Post wants the public to know. If you’re listening, @Jeff Bezos, get your eyeballs over to emptywheel.net and read this post by Marcy Wheeler!

  3. Anvil Leucippus says:

    You’re like an inference engine. So using those steps on Mifsud popping up again in the news:


    I infer that Mifsud was an unwitting co-conspirator in Russia’s influence over the 2016 election; who went so deep underground that his lawyer has to setup a shell company for them to communicate with each other.

    He thought he was upgrading his career, campaign team gets an offer too good to be true (which it is) and just gobbles it up because they are greedy people with the common sense of a rock. This also sounds an awful lot like plans to build a giant phallus with Trump’s name on it.

  4. Avattoir says:

    Now I feel something akin to memories of watching teacher upgrade a particularly selfish, stubborn willful cheat of a classmate before the entire class, not just dismissively condemning the little cheat as ‘I know what you did’, but walking the little shit and every witness to the process thru each twisted things said, written & done,

    Would I want now to be a fly on the wall listening to the discussion of this post in the national news chief’s office, or Baquet’s corner suite? I’m sure I’m not alone here in having watched sometime last year post-Aspen the video of an interview conducted there with several NYT reporters, – AND BAQUET – connected to covering – in the Times’ special fashion – this Trump-Rus horror show, including in particular that utterly wrong, Trump-serving piece last in the 2016 campaign that blared out the Grey Lady’s considered view that Trump was free of Putin cooties. Having watched the snooty, imperious, dismissive tone & several self-justifications from each of Baquet, Haberman and Schmidt, I don’t have much difficulty imaging a revival of that dismaying performance in house tonight.

    I find myself thinking now of what LeCarre had Control saying to his aides as he worked thru the betrayals in Tinker, Tailor: ‘There are 3 of them and Alleline’.

    There are 3 of them and Sulzberger.

    • Eureka says:

      Well the troika’ve been having too many tete-a-tete-a-tetes, apparently; I’ve noticed of late that Schmidt’s face, perma-furrowed into some type of haughty consternation, is about to collapse into itself along the center line.

      I don’t think the popular botulinum shot helps with that.  Subsidized disdain, helluva drug.

      ETA: tho perhaps that is ‘seriousness,’ and I am misreading.

    • BobCon says:

      I think the thing that saddens me the most about Jill Abramson’s implosion is that this will only empower Baquet. He must get endless joy that the story of how he screwed up Eric Lichtblau’s reporting and created the “No Clear Link” fiasco is now tainted by association.

      She would have saved herself a lot of grief by skipping the lame stuff about challenges of the media in a digital environment and stuck to what she could have done best — digging up the bodies that Dean Baquet had buried. Now it’s even less likely we’ll ever get that public service.

        • harpie says:

          Here’s Franklin Foer last October about Baquet:

          4:58 AM – 8 Oct 2018 Dexter Filkins does a meticulous job revisiting the Trump/Alfa Bank server connection–and lands at about the same conclusion I did a few years back: This wasn’t random. / […] / Dean Baquet didn’t run the Times story on the subject–and he slammed my piece on Alfa to @ErikWemple . But he now says, “It felt like there was something there.” 

  5. Rusharuse says:

    Trump goes to El Paso and lies his ass off, meanwhile Mueller is sitting on this dynamite. Hey Mueller, do your #2 or get off the pot!
    FFS – Schiff (or someone with authority) . . DO SOMETHING!!!

  6. Jibe-Ho says:

    I’m pretty much shaking here. So NYT is not “all the news that’s fit to print.” How in the world have we come to that? The great takeaway from this phenomenal break-down of the chronology is that ABJ is really on top of the case. She is no ordinary judge, and Marcy parsed everything so succinctly from her comments. Thank you, Marcy; you are a national treasure. I agree with other more learned folks above: get this story to people in power and to we the people as well!

  7. Eureka says:

    As an appreciator of details, I like that you gave them VERY LARGE THICKLY-FONTED CATEGORY HEADINGS to point the way.

      • Eureka says:

        Well, it’s the BIGGEST THICKEST font I’ve ever seen- maybe there were more subordinate categories in this particular post to trigger it or something.

        Just seeing the news and it appears that WaPo, at least, has caught up on their reading.  (Snarks aside, it’s nice that ew laid-out and linked everything such that any diligent citizen reader could verify for themselves as needed.)

        (Adding:  I just realized your current post is an open thread, so I cut an add-on para. out of here to stick over there—>)

        • Rayne says:

          LOL She did whip out both the monster H2 as well as the H3 this time. They can’t say they can’t read it because of poor visibility.

  8. progressiveandsane says:

    Marcy — I so appreciate what you are doing here. It is very important. In general, the public has no idea how frequently journalists — especially from premier marquee entities such as the NYT — fuck up majorly. And often, their mistakes are carried downstream, and repeated by others at smaller, less resourced entities. On the one hand, it is a peculiar behavioral response of some reporters to avoid corrections at all cost. On the other, was my response (when I did daily journalism) to the discovery that a source lied to me: Make it my express mission to show the readers that this person was a liar — in every instance that was legitimately (by facts and newsworthiness) possible. Every. Single. Time. I wanted them to rue the day they made that very bad decision to get me to put a lie in the newspaper. I know it is frustrating to watch a news source that many rely upon for the facts repeatedly fuck up something this big — when you know better –much better. Fortunately, I see you and your analyses being relied upon more frequently by other, smarter reporters and given greater prominence. You, go!

  9. Missy says:

    The NYT (well these jornos) has been Trump’’s mouth piece since forever. They choose not to report on his crimes and helped get him elected.

    The question is why? It’s not $$$. My god, the truth about Trumps crimes is even more juicy than their lies. Who has what to gain with NYT complicity? And why was the public editor let go?

  10. Jenny says:

    Thanks Marcy. Your in-depth post is outstanding. “Lies lies everywhere lies” came to me when I read this. A song? Not sure, just seems to fit.

  11. brewstate says:

    I think the polling data is the lynchpin that is holding at least one arms of Mueller’s investigation up because it so clearly directly links a quid-pro-quo between Russia and Trump’s campaign and must be more than just public polling. As a side note and sheer conjecture, I revisited the seemingly strange decision for Trump to attack the states that eventually won him the election (Michigan, Wisconson, Penn) https://www.newyorker.com/news/john-cassidy/why-is-donald-trump-in-michigan-and-wisconsin and have continually wondered if the Kilimnik data related to those states? I have no answers, only questions. 1. What changes did Trump make in his campaign after the August meeting that could be related to polling? As the article above shows, sometime around September he had an odd but prescient vision of a midwestern win against the advice of most political pundits who suggested Florida or other swing states. Public polls on Fox news etc showed no path there at all. Did he have special polling that showed him there might be a path there? Was it shared and coordinated with Russian intelligence on the off chance he could pull it off? I don’t know, but if it was sheer coincidence, it seemed to be a very profitable one for the campaign.

    • bmaz says:

      Hi Brewstate – Good handle, Marcy is going to assume that means Michigan. We like beer here, not like Kavanaugh, but in a good way. Anyway, this seems to be your first comment, so welcome to Emptywheel, jump in often, it is a good community.

    • MattyG says:

      My assumption has been that ‘polling data’ refers to the complex analysis (performed by Trump contractors CA etc.) of where best to focus Russian voter suppression operations – the targets, message and timing – with specific regard to electoral college effects – rather than a more open field discovery of potential enclaves DT might have hidden advantage in voter preference that national pollsters had not yet identified.

      The close contests in the Midwest apparently presented a suitablly narrow focus tactical battlefield to apply these tools.

      • Herringbone says:

        Ohhhh. That makes sense. I was going on the theory that Manafort was arguing, based on polling data that showed a closer than expected race in key states, that Russia should go all in—deploy the resources they’d used in Ukrainian elections—rather than play to lose.

        And also arguing that if Deripaska would just hold off on terminating Manafort’s outstanding account for a few months longer, he could get the president of the United States on the payroll.

      • pseudonymous in nc says:

        Westling talks about the poll data being prepared for a “scheduling meeting”, i.e. where to send the candidate and surrogates. Now look at this:


        The polling data probably was used for scheduling, and that Manafort read Kilimnik into that. So where the campaign held its rallies — often in small cities — would indicate areas to target. But the data probably also showed where voter suppression messaging would work.

  12. OrionATL says:

    Emptywheel has begun teaching journalism as she has practiced it for more than a decade. 😉 In addition to the techniques mentioned here see the list of resources at the end of this post:

    A journalist who studies and follows this advice and some of the specific techniques will be a more reliable, more complete journalist. Of course editors can be significant roadblocks to good journalism. I suspect that is almost always a part of the nytimes’ endemic trustworthiness and good judgement problems.

  13. BobCon says:

    Another way to think about this fiasco is to realize that the Times reporters clearly have put no thought into what “polling data” means. They clearly think of it in the most superficial way, as indicated by the credibility they offer to the claim that it was public information.

    It would have helped them immensely if they had bothered to talk to a professional data hound and asked them what might fit into the boundaries of what is described in the court documents.It would have been obvious almost from the get go that it is not, as they clearly seem to think, just a small set of numbers. It’s like writing about a “gun shipment” and never once trying to puzzle out from available clues whether this was a shipment of two handguns or an entire container ship full of heavy weapons.

    There is simply no way, for example, that Manafort would have felt the need to personally hand over some pages of weekly Gallup poll results that had been printed in the NY Times. Nor, for that matter, was it something a little more sophisticated, like a printout of fivethirtyeight.com’s state polls.

    They completely skip over the telling phrase “This is very detailed [redacted] on a level that is very focused” and miss the obvious implication that this is not a hundred or so lines of text. How much data is it? Megs? Gigs? How was it structured?

    And probably most intriguing, how were the Russians going to use it? You can’t simply hand over a bunch of pure polling data three months before the election and expect it to be of much use. The amount of time to put unfamiliar data into shape and create models for using it would A) put them too close to the election to have much impact and B) run the risk of it being too old for use at the level they would want.

    My suspicion is that this was linked with sharing that had already been going on. Manafort knew how the Russians were targeting Americans, including the outreach methodologies and the formatting of their data. What Manafort was supplying was the refined fuel needed to run their machines.

    Maybe I’m wrong, but one thing that is clear from the Times reporting is that they have no sense of what “polling data” is referring to, and they never bothered to think it through. They think they know, and that’s led them to miss what’s right in front of them.

    • Rayne says:

      Manafort knew how the Russians were targeting Americans, including the outreach methodologies and the formatting of their data.

      Likely the same way Manafort helped Russia manipulate Ukrainian voters about Yanukovych.

      • North Jersey John says:

        I’m hazarding a guess at these crucial redactions, as well as piling onto EW’s great analysis of the centrality of Manafort’s polling data lies.  Let’s suppose that the crucial term being redacted is Modeling (or Segmentation, or Cross-tabs).  If I was seeking to illegally coordinate my digitally targeted influence operations, I’d love to know exactly how the campaign was analyzing and subdividing the electorate.  What issues had they polled, and how did they subdivide the respondents?  What issues or statements had greatest/least salience with each sub group.  Which subdivides seem to have strongest correlation with specific target groups.   There are tons of fascinating details hiding underneath the bland label of “private polling data”.

        snip.   Note how neatly Modeling fits into both redacted brackets.

        And Weissmann returns to this, once again making it clear the data sharing happened on August 2.

        Both of those refer to [redacted] and also refer to the discussions of the — discussions of [redacted] at the August. 2nd, 2016 meeting.

        • Rayne says:

          Mmm-hmm. To be really effective at microtargeting — enough to focus on a few thousand voters at a time — I don’t see how the campaign could have done anything but use multiple data sources and model anticipated responses. Like so:

          1 — Public polling data = how a voter behaved at the polls over time, sort out for [X]
          2 — Opposition polling data = identify Lean GOP/Neutral/Dem but not HRC-committed and sort out of [X] => [Y]
          2 — Social media data = what topics a [Y] voter responds to most often and the nature of their response, sort for [Z]
          3 — Online polling data = A/B switch data collected over a narrow window of time from [Z] voters, sort for most common volatile response inducing most action (a final A/B switch)
          4a — Craft messages and push to [Z] voters, modified message to [Y] voters.
          4b — Identify key node (influencer) voters on each dispersion and earmark for increased effort.
          5 — Rinse, repeat until Step 3 shows a solid anticipated response rate.

          Yeah, I could see Modeling fitting in context. They had about 4-6 weeks at that point in time to perform a couple iterations on whatever they’d already collected and processed before it became really serious. They’d have little wiggle room after conditioning the audience during initial A/B tests.

          But whatever the two [redacted] fields are, it’s not the word Modeling alone. Microtargeting doesn’t, either. Monospace Courier font is a blessing and a curse.

        • pseudonymous in nc says:

          I agree with this. It’s state models, locality models, turnout models, demographic models.

          ABJ: “Doesn’t it reflect particular [REDACTED]?” (p.89)

          Westling’s “it’s meaningless to me” argument is that the polling for a scheduling meeting, which implies that it would influence where to send the candidate and surrogates, which in turn implies sufficient local granularity to pick out places like West Bend, WI and Dimondale, MI for rallies.

          But passing on turnout models also fits into Parscale’s “voter suppression” strategy.

        • Justin Huff says:

          Seriously? – you realize we’re talking well over 60 days for pre-election voter suppression with a dataset generated by the resources of Kushner’s Project Alamo (aka Cambridge Analytica) given to an asynch influence op up and running for well over a year? No way it’s the first one…  (just a hypothetical guess – I could be wrong… ;-).

      • Leila512 says:

        Maybe off-track, but was it ever established what the Alpha Bank/Trump Tower bursts of server activity were about? Could it have been related to how they were using data for targeting and messaging?

      • Joe says:

        There is public polling data, private Republican polling data, and private Democratic polling data.  (The Democratic data was already stolen.). Of the many ways to “transfer data,” sending a Manafort-type campaign manager on a plane ranks orders of magnitude below other methods such as FTP or FedEx.  However, given different data sets, problems occur with the join of those data sets.  That is, the question “hey we just got our last batch of Democratic Party data (also possible to define as “public data” because it had been stolen and made public), how do we integrate it with Republic party data?” is important. Or, there could be different programs used on the data sets by the Republicans and the Russians.  Further, there could be clarity in computer programs and tactics, but lack of clarity about strategy.  A strategy discussion for the last month of the campaign would suit a Manafort visit.  The mundane topics of tactics, data normalization, and data transfer are less demanding of a man wearing an ostrich jacket.  However, it’s also possible that Manafort had to advise on the question of which data set to privilege over the other in unique circumstances:  in Ohio, was the Democratic Party data better than the Republican?  Micro targeting can’t all be done with generic algos.  But yes, not only, “What data?” but “What’s wrong with FedEx?”

        • BobCon says:

          You raise a valid point about the question of why not use Fedex or FTP. I wonder whether Manafort’s squirrely behavior regarding his iPods is related, although it’s always possible it was something like personal financial data on them.

          However there does seem to be concern about eavesdropping, and if Manafort’s going there anyway, maybe use him as the courier? Maybe not, if you’re worried he’ll lose his luggage…

          Regardless, I agree his main purpose was probably to provide the keys to the data and how to use them.

    • Silence Hand says:

      If they’re ignorant about “polling data”, they’re willfully so.  My sneaking suspicion is that the rot of Access Journalism gnaws deep in the Grey Lady’s bones.  I think they just can’t stop crafting stories that maintain their good graces with insiders giving them the hot poop.  In fairness, it’s how a lot of journalism has functioned since time immemorial.

      The rationalization, of course, is that Access Journalism is in service of the greater good: and, after all, how many people whose opinions are actually malleable understand the distinction between the spin and the actual fact?  Plus, doing actual research (as Marcy does) is difficult, time consuming, and unglamorous.  Not for the lazy, the starstruck, or the lazy starstruck.

  14. viget says:

    Reply to Eureka @0243

    As one of my mentors in the lab said to me once about powerpoint presentations of my work, “You’ve basically got to dumb it down and tell a story like you’re reading to your 5 year old.” Hard to believe that reasonably intelligent people need this, but when people are bombarded by conflicting messages all day long, getting their attention and keeping it is a valuable commodity.

    Basically, people like to be led down the primrose path. Honestly, as Marcy and the rest of us start to finally see the whole Matrix on this, we need to write more of these succinct articles to get the rest of the journalist-class’s (limited) attention. Yeah, they’ll be plagiarizing the work done here without credit, but hey, at least this important story gets told.

  15. Trevanion says:

    I hope that when (if) the dust finally settles someone will methodically unpack and analyze the very weird symbiosis that has developed between modus-NYT (2015-2019) and Trump.

  16. Drew says:

    Great post. And it clarifies some things I was puzzled about in previous posts, like the Manafort selling out Trump thing.

    On the composition of the polling data passed on on August 2. It stands to reason that it wasn’t just clips from 538-and it wasn’t a core dump of the entire polling data on the Trump campaign & related consultants’ servers [because it was something Gates could print out on his office laserjet & pass along in physical form]. The discussion above prompted me to think–perhaps Manafort had received instructions on what to have extracted that would be of particular use to the Russians?

    Who knows how much of the underlying data had been gleaned from public/semi-public sources? –the capability of U.S. political consulting/polling firms to aggregate such data is something that the Russians would want to exploit rather than replicate–but the GRU would have its own questions to be asked of that data. In other words, as detailed as it was, Manafort may have been passing on data that was already trimmed and targeted for usefulness.

    The representations of the “person knowledgeable of the situation” are, to me, very reminiscent of the representations about Gates, Manafort, etc. by Kevin Downing in the hearing transcript. Since he seemed to be happy to mouth off pretty aggressively to ABJ about Gates’ credibility & the judgement of the jury, I’m sure he would be similarly aggressive in presenting his version of the facts to the New York Times.

  17. Mark Ospeck says:

    Great stuff EW.  I’ve been trying to figure Baquet’s motivations for pushing dezinformatsiya for a while now.  vg that you are calling he and NYT on it.  Maybe a list and timeline that summarizes all the substantive facts that NYT has gotten wrong about Trump-Russia and the SCO investigation over the last couple years would prove useful. Then run a correlogram on it against then-current events.

  18. orionATL says:

    i thought the most important info in the lafreniere, et al article was that the discussion between the trump campaign manager, manafort, and the russian intelligence operative kilimnik included not just polling data, but discussion of a settlement of the 4-year old ukrainian conflict between russia and the ukraine as well as discusion of economic sanctions the u.s, had levied against russia, both of which were extremely important to the kremlin. polling data was just one of the technical means to insure that the guy who would relieve russia of onerous u.s. conduct would be elected prez.

    given the alleged range of the discussions, i doubt much technical detail about polling was discussed. i would guess this meeting served the purpose of manafort delivering to a messenger to putin some specific bona fides from the trump campaign (and the candidate himself) to the russian leadership and putin.

  19. Pete says:

    Any chance of getting this published as an op-ed in the NYT?

    I didn’t think so, but other than at EW where else?

    Most news outlets, televised especially with their 7 minute segments, cannot work.  Plus, a lot of people could not follow the incredible complexity even with the hell-of-a-job does by Marcy to connect the dots.

    Maybe a televised PodSaveAmerica segment.

  20. MattyG says:

    Great walk-thru. Not sure if Mike and Maggie attain Judith Miller status but NYT reportage has been painfully muzzled often to the point of irrelevancy in these scandals – though as illustrated here hiding behind a source which as reporters they must assume at very least is or may be duplicitous, without the rigorous research and analysis and careful parsing (thanks EW!) to dissect source accounts in print is a service worse than irrelevant.

    • Silence Hand says:

      Yes.  It baffles me that Mike/Maggie et al. apparently haven’t realized that “Spin the Grey Lady” is a top-five action item for politicians engaged in getting away with figurative (and, who knows, maybe literal) murder.

      They learned nothing from the Miller fiasco.

  21. tinao says:

    I love that you are making it known how nyt maintains liar’s lies. They lost my readership years ago when they screwed ACORN out of existence.

  22. Vinnie Gambone says:

    Based on what I saw in my own blue collar neighborhood on Facebook more than a year before the election there was what seemed like operation Surrogate. One, then two, then 10 then 20 known people in the community who relentlessly pilloried Obama with messages that were slightly beyond their ability to create, especially the graphics and photos. One individual in particular took great delight in delivering what seemed a torrent of negative messages. This solidly pro union pro democrat neighborhood,and others then went for trump. I believe the russians had a hand in testing where anti Obama messages were liked or shared and it served as a barometer for where to ply similar messages later. It was like chumming when you go fishing. Somehow they were testing what would work physiological and where;- it wasn’t always about politics. It was about mood, and world view. People i knew my entire life and previously liked became imbeciles, fanatical imbeciles, right before my eyes. So disturbing. A seething bitterness seem to swell up in people who had not shown that trait before. It’s enough to make you want to smack them in the mouth for being so stupid. Fortunately I have a secret zen Jiminy Cricket that reminds me that their punishment is they are stuck in their own skins and nothing I can do could make them more miserable than they already are. There’s no helping them, but there are plenty of other people who I can help, and so I deliberately shift my focus there, and banish the morons from my thought process. More and more americans are turning into the rats in John Calhoun’ s utopia experiments, eating each other. ( Mass shootings? ) I am so afraid for out children and the world they will inherit from us.

  23. Mulder says:

    This at the end of the post

    …at a clandestine meeting he may have flown home from on Oleg Deripaska’s plane.

    Is this reported anywhere?

    Great analysis, Marcy. And so well written I read it twice.

  24. maestro says:

    From reading the whole transcript, my strong impression is that the polling data at issue concerned testing public opinion on possible sanctions relief plans. Curious if anyone else got that impression as well.

  25. Reader 21 says:

    @ Mulder—Today’s WAPO, article linked to upthread.

    @ Maestro—definitely not. And positing such, TBH, reads like a distraction.

    Great work, EW—glad someone is calling Maggie and mike on their bullshit. Maggie Duranty, indeed.

  26. Silence Hand says:

    Not to be boring and redundant and boring, but re-upping my comment from above. I suspect that Marcy is calling out the NYT on some especially greasy Access Journalism.  As noted, the rot that gnaws deep in the Grey Lady’s bones.  I doubt that they’re so lazy they don’t understand the difference between the spun information and actual fact, but I bet the bastards consider it an abstruse point that won’t matter to Joe Reader.  Best to continue parroting the source to maintain access – sweet, sweet access.

    I’m forwarding this EW post to everyone I think should care, because here Marcy has exposed them for the lazy, duplicitous tools they’ve become in this process.  As a scientist, I can attest that actual research is hard.  If you’re doing it right, most of your analysis is boring.  Put another way, think of it like mining gold the way it’s actually done.  Nobody strikes huge veins of the metal anymore.  Rather, they process hundreds of tons of ore to get ounces of it.  As in the business of news affecting the politics of the Republic, that insider telling you about a massive strike is actually selling you on digging up iron pyrite.

    They’ve learned nothing from l’affaire Miller. Could this be the public whipping that finally weans them from the icky stenography of Access Journalism?

  27. Silence Hand says:

    upon reflection, shorter Marcy:


    Or, at least that’s what I take from it.

  28. Democritus says:

    Let them have it Marcy. It’s also driving me nuts watching you two steps ahead of MSM, I imagine you have excelled at the art of headdesks for a while now.

    Also thank you for a bit of a framework :)

    Kudos, you rock!

Comments are closed.