The indefatigable Doug Cassel is back with a pro-Chevron look at Ecuador’s forensic evidence regarding the judgment ghostwriting allegation in the Lago Agrio case. It’s fair to say the evidence is equivocal—both sides have a case to make. Here is Ecuador’s new brief addressing Chevron’s arguments: I hope to have an Ecuador advocate give Letters Blogatory readers a post about Ecuador’s position soon. My basic perspective is this: perhaps the evidence that seems to suggest that Judge Zambrano actually wrote the judgment can be challenged; but the fact that there is such evidence gives me reason to think that he did write the judgment. Otherwise, we have to think that Judge Zambrano created false evidence at the time of the judgment to make it appear that he wrote the judgment. How can that be if, as Chevron suggests, Judge Zambrano is a chucklehead? I don’t believe he could have been that devious. I am also influenced by the presumption of regularity, which I do not think we can discard in this case without making the argument profoundly circular. And as long-time readers know, I found Judge Guerra’s testimony deeply unconvincing.

Update: Adam Klasfeld has responded to Doug Cassel’s criticism.

A Courthouse News reporter recently penned a purported bombshell: according to a leaked brief written by Ecuador’s lawyers in its international arbitration with Chevron, Ecuador’s computer expert found no evidence in Judge Nicolas Zambrano’s computers that the Lago Agrio judgment was ghostwritten. Instead, the expert claimed, a file containing the judgment was created in a Zambrano computer and then saved and edited hundreds of times. Contrary to the findings of US federal Judge Lewis Kaplan in Chevron’s suit against Lago Agrio lawyer Steven Donziger, Ecuador’s expert supposedly confirmed that the computer evidence was “consistent” with Zambrano having written the judgment.

The story was shaky from the start. The Courthouse News reporter had not seen the actual report by Ecuador’s expert, J. Christopher Racich. The story quoted only from an unredacted copy of a brief written by Ecuador’s lawyers, to which the story linked. Even though a computer amateur can easily ascertain from the metadata that the “author” of the linked copy is “Karen Hinton”—Donziger’s PR person—the reporter did not let on that his source was an interested party (although he did disclose the admission by his source that the leaked document came from the office of Ecuador’s Attorney General).

The reporter clearly had not seen either Chevron’s briefs or the expert reports filed in the arbitration on its behalf by Spencer Lynch, Director of Digital Forensics in the London office of the forensics firm Stroz Friedberg. The reporter did not even know that Lynch had filed a rebuttal to Ecuador’s expert.

The reporter also evidently saw nothing amiss in the oddity of Ecuador’s leaking its lawyers’ characterization of a report which Ecuador itself—over Chevron’s objections—had asked the arbitral tribunal to keep confidential. Because of the confidentiality order, when the reporter asked for comment, Chevron could not disclose what the expert reports actually say.

Following the leak, Chevron asked the arbitral tribunal to lift the confidentiality order. The tribunal only partly agreed. It allowed Chevron to make public its own briefs. But the underlying forensic analyses of Zambrano’s hard drives, by Chevron’s expert Lynch and by Ecuador’s expert Racich, remain confidential—at Ecuador’s request and over Chevron’s objection—except to the extent they are quoted in the lawyers’ briefs.

Even this limited disclosure, however, demonstrates that the Courthouse News reporter was suckered by Hinton. The new information shows that the reporter got the story backwards: Instead of confirming Zambrano’s claim that he wrote the Lago Agrio judgment, the forensic evidence from his computer hard drives contradicts his claim.

Testifying in Donziger’s trial, Zambrano repeatedly insisted that he spent long hours daily, from October or November 2010, until the nearly 200-page Lago Agrio judgment was issued on February 14, 2011, singlehandedly drafting the judgment. He claimed he did so entirely on his “new” computer. He testified that he dictated most of the draft to his 18-year-old typist, and typed the rest himself. No documents or text, he claimed, were ever inserted into the draft from any other source.

Chevron’s expert Lynch examined the hard drives on Zambrano’s two computers (both his “new” and his “old” computer) to see whether they were consistent with Zambrano’s testimony. They were not. As quoted in Chevron’s newly disclosed brief, Lynch concluded, “In summary, the totality of available forensic evidence is inconsistent in all material respects with Mr. Zambrano’s testimony describing how the drafting of the Ecuadorian Judgment occurred on the New Computer.”

Lynch also found that even the fallback story hypothesized by Ecuador’s computer expert – that Zambrano, contrary to his testimony, in fact saved successive drafts of the judgment on his “old” computer—is shot through with holes.

In contrast, the forensic evidence in Zambrano’s hard drives is consistent with Judge Guerra’s testimony for Chevron, namely that the Lago Agrio judgment was drafted on plaintiffs’ lawyer Pablo Fajardo’s computer, and that Guerra, at Zambrano’s request, “fine-tune” and “polish” the draft—on Fajardo’s computer—making “very few changes”.

Zambrano’s “New” Computer

Zambrano testified that he drafted the entire judgment on his “new” computer. This was no casual recollection. He pressed his claim repeatedly and insistently. He used his “new” computer instead of the “old” one “because it was the more modern computer.”

Forensic analysis belies Zambrano’s claim. The only copy of the full judgment text on his new computer was added in July 2011—five months after the judgment was issued. During the six weeks preceding the issuance of the judgment, from December 28, 2010 to February 14, 2011, when Zambrano was supposedly working overtime on the judgment, the Microsoft Word program he would have used to draft the judgment on his “new” computer was open for a grand total of only four hours. He could not possibly have dictated or typed the judgment on his new computer as he claimed.

Ecuador’s Fallback Story

Ecuador’s expert tries to explain away the absence of any judgment on Zambrano’s new computer during the drafting period. He raises the possibility that the judgment was repeatedly saved on Zambrano’s “old” computer—even if Zambrano did not realize it.

The basis for this surmise is a document in Zambrano’s old computer called “Providencias” (which means, in legal Spanish, court orders in matters of little importance). Providencias was repeatedly opened and saved on the old computer during the period when Zambrano claims to have written the judgment. It contained progressively increasing amounts of the Lago Agrio judgment text. Ecuador’s expert assumed that Providencias was the document that became the judgment.

There are multiple problems with this theory. First, how could Zambrano have believed he was drafting the judgment on his new computer, when in fact he was saving it on his old computer?

Ecuador’s expert explains that Zambrano’s old computer could be accessed via the new computer. He conjectures that, even though Zambrano thought he was editing and saving Providencias on his new computer, he was inadvertently saving it on the old computer.

Unfortunately for Zambrano, this theory is contradicted by the hard drives. Chevron’s brief quotes the finding by its expert Lynch: all versions of judgment text found in Providencias during the relevant period “were saved by the Old Computer, and not through … the New Computer.”

How does Lynch know that? As quoted in Chevron’s brief, he explains, “Had any of those versions been saved using the New Computer the metadata for those versions would reflect that they had been saved using the New Computer.”

In short, Zambrano neither saved Providencias on his new computer, nor mistakenly thought he was doing so while in fact saving it on the old computer.

Ecuador’s expert also assumes a drafting schedule that renders his theory a physical impossibility. He posits that Zambrano began drafting on October 11, 2010 (the date Providencias was opened). But on that date the new computer had not even been shipped by its manufacturer. The new computer was not purchased by Ecuador’s judiciary until November 26. It was not used for the first time until December 7. So for the first two months of the expert’s assumed drafting schedule, Zambrano could not possibly have used a computer he did not yet have. He could not have been drafting Providencias on his new computer, while unknowingly saving it on his old computer.

Did Zambrano Forget?

So Zambrano did not draft the judgment on his new computer. But maybe his memory was mistaken. Might he have drafted it on his old computer?

Even aside from contradicting his emphatic testimony, this second fallback theory, like the first, suffers from multiple credibility problems. To begin with, would we not expect a judge drafting a judgment (let alone a multi-billion dollar judgment) to save it as a separate document? Yet the Providencias document on the old computer contains the texts of various documents. The Lago Agrio judgment is only one among them.

There are also unexplained discrepancies between the judgment on Providencias and the judgment as issued. The latest judgment text in Providencias on Zambrano’s old computer was saved after the real judgment was issued. But the headers in Providencias do not match those of the real judgment. Ecuador’s lawyers try to explain this away. They argue that when judgments are issued, Ecuador’s computerized judicial records system automatically inserts a header. That could explain a discrepancy if Providencias had no header. But in fact Providencias had a header—only the wrong one.

Ecuador’s computerized judicial records database casts further doubt on Ecuador’s supposition that Providencias was the document that became the judgment. Chevron’s brief notes the finding by Lynch, based on records from Ecuador’s database, that the judgment does not appear to have been uploaded into the database from either Zambrano computer.

Finally, Excel program data confirms that Providencias could not have been the document that became the final judgment. As quoted in Chevron’s brief, Lynch found that the real judgment “contains data from two different Excel spreadsheets, including statistics calculated across thousands of rows of data.” Yet during the period the judgment was drafted, Excel was opened for only four minutes on Zambrano’s computers—far too little time to input that volume of data and calculate the statistics.

Lynch concludes that “based on the recorded activity of the Excel program, it would not have been possible, in the amount of time Excel was recorded as having been used, to use either of the Zambrano computers” to derive the statistics appearing in the judgment, or even to copy the Microsoft Excel data into Providencias.

Who Really Drafted The Judgment?

Zambrano’s hard drives do not, by themselves, reveal who drafted the judgment or on what computer. What they evidence is that Zambrano’s claim—that he drafted the judgment on his new computer—is inaccurate.

In contrast, the evidence from Zambrano’s hard drives is consistent with Guerra’s testimony. If, as Guerra testified, plaintiffs drafted the judgment on Fajardo’s computer, and Guerra then fine-tuned it on Fajardo’s computer, that would explain why the judgment cannot be found on either Zambrano computer, and why Ecuador’s judicial database appears to show no uploading of the judgment from Zambrano’s computers.

But does the forensic evidence admit the possibility that plaintiffs drafted the judgment on Fajardo’s computer, as Guerra says, and then periodically inserted portions of the text into Providencias?

Even though Ecuador’s expert unavailingly tries to discount this possibility, the answer is yes. Ecuador’s expert argues that no document listing plaintiffs’ counsel as author can be found on Zambrano’s computers. But that proves nothing: if plaintiffs merely inserted excerpts of other documents into Providencias, no author listing would appear for the inserts.

Could plaintiffs have done so? Again, the answer is yes. Both experts agree that the computer evidence suggests that external texts—not found on either Zambrano computer—were inserted into Providencias. In addition, the forensic evidence shows that multiple USB devices (typically called “thumb drives” or “flash drives”) were inserted into both Zambrano computers during the judgment drafting period.

Ecuador’s expert argues that no Lago Agrio documents were inserted via the USB devices. He notes that none of the USB document names refer to the Lago Agrio case. But several USB documents have generic names such as “KKKK.doc” and “Documento 1.doc”. Neither expert knows their contents. And if plaintiffs—who used code names even in their internal emails—were secretly passing ghostwritten texts, they would hardly label their documents with incriminating names.

In addition, the citations to foreign cases in Providencias—from the US, England, France and Australia — almost certainly came from an external source. Zambrano knows neither English nor French. In Donziger’s trial he explained that his 18-year-old assistant found these foreign language cases through internet searches and then translated them through internet translation services.

In an effort to support this whopper, Ecuador’s expert reported that Zambrano’s computers show internet searches on Google, Bing, Lexis Ecuador and other search engines. But in fact no search of a legal research site—with one exception—was done during the period the judgment was supposedly drafted. That one exception was a search on the Spanish language search engine “” Yet as a Spanish-speaking expert reported for Chevron, the foreign language cases cited in the judgment cannot be found on fielweb.

The foreign citations, then, came from an external source. Again, the most plausible explanation—and one consistent with Guerra’s testimony—is that the citations came from plaintiffs’ multi-national, multi-lingual legal team. But wherever they came from, they apparently did not come from Zambrano or his young assistant.


The foregoing highlights do not exhaustively summarize the briefs on the lengthy forensic reports. Readers are welcome to review the full briefs here and here. Given the passage of time and possible destruction of documents, the forensic evidence in Zambrano’s hard drives, by itself, might not be conclusive on all points. But in the words of forensic expert Lynch, the “totality of available forensic evidence is inconsistent in all material respects” with Zambrano’s tale of how he drafted the judgment. The new evidence also shreds Ecuador’s fallback story.

In contrast, Zambrano’s hard drives are consistent with Guerra’s testimony, which was found to be credible by US Judge Lewis Kaplan after hearing and observing both Zambrano and Guerra on the stand. Far from supporting Zambrano, or validating the leak to Courthouse News by Donziger’s PR person, the forensic analyses of Zambrano’s computers provide strong corroboration of the other extensive evidence—forensic, physical, testimonial and circumstantial—that the Lago Agrio judgment was ghostwritten.