David Margolis Is Wrong

The Justice Department’s ethics investigation shouldn’t leave John Yoo and Jay Bybee home free.

Also in Slate, Dahlia Lithwick writes that the overturning of the OPR report means there are no legal limits for torture.

Jay Bybee, former assistant attorney general for the Office of Legal Counsel

After more than a year’s delay, the Department of Justice released the Office of Professional Responsibility’s report on whether government lawyers who wrote two notorious torture memos violated professional ethics. OPR is the Justice Department’s internal watchdog, and it has the authority to recommend referring errant DoJ lawyers for professional discipline or even criminal prosecution. The 260-page OPR report, plus two earlier drafts released at the same time, harshly condemn the torture memos and recommend discipline for the lawyers who wrote them. But Friday’s document dump also included a 69-page memo by David Margolis, the senior DoJ attorney who resolves challenges to negative OPR findings like this one. Margolis ripped into OPR and rejected the ethics office’s recommendations that Jay S. Bybee (now a federal judge) and John Yoo (now a law professor) should be referred to their state bars for discipline. Margolis’ decision means that Judge Bybee and Professor Yoo will be spared further ethics investigations. They are home free.

The blogosphere has been crowded with initial reactions. (Two pieces on Balkinization are particularly good: here and here. And James Fallows’ comment here is indispensable; he rightly notices how heavily Margolis leans on post-9/11 panic to excuse the lawyers for shredding the law against torture.)

I’d like to delve deeper. Margolis’ critique got Yoo and Bybee off the hook. So how strong is his position compared with OPR’s? The short answer: weak.

The focus of OPR’s investigation is two memos from Aug. 1, 2002. They were written in secrecy by John Yoo and an assistant in the Office of Legal Counsel whose name is redacted. Bybee, as head of OLC, signed off on the memos. One memo analyzes torture, executive power, and criminal defenses for interrogators accused of torture; the other approves 10 techniques for tormenting Abu Zubaydah in a secret CIA prison, including water-boarding him. The question for OPR was whether Bybee and Yoo had violated ethics rules by twisting the law to the breaking point to give CIA operatives maximum assurance and leeway.

Both the OPR report and Margolis agree that (in Margolis’ understatement) “these memos contain some significant flaws.” There they part company. The OPR report finds that Yoo and Bybee violated two rules of professional conduct: Rule 1.1, requiring competence, and Rule 2.1, requiring lawyers to “exercise independent professional judgment and render candid advice.” Margolis rejects OPR’s analysis and concludes that “poor judgment” rather than professional misconduct “accounts for the entirety of Yoo’s work” on the torture memos.

But that’s not the right characterization for memos that used extravagant legal reasoning to approve torture. It’s like saying that Iago’s advice to Othello showed poor judgment. OPR made a powerful case against Bybee and Yoo. In response, Margolis went after OPR like a defense lawyer, upped the burden of proof beyond what the ethics rules require, and minimized the liberties that Yoo and Bybee had taken with the law.

OPR’s analysis changed between the drafts and the final report, and Margolis goes on for pages about that, quoting liberally and uncritically from Bybee’s and Yoo’s objections and insinuating that OPR’s efforts to respond are worrying signs of “a shift in OPR’s reasoning”—although he admits that he was the one who recommended that OPR solicit and review the objections. Originally, OPR looked to the rules of ethics of the D.C. bar instead of OPR’s own more lawyer-protective framework. The OPR standard requires not just an ethics violation, but an ethics violation that the lawyer committed intentionally or in reckless disregard of the rules of conduct. In other words, OPR’s framework requires proof of a guilty mental state over and above what the ethics rules themselves require.

This difference is not simply an arcane technicality. Proving mental states is always hard for prosecutors to do, and the more you ask them to prove, the less likely a finding of guilt becomes. Even so, in the final report OPR concluded that Yoo and Bybee should be referred for discipline under the OPR framework. At which point Margolis slams OPR because “its misconduct findings do not identify a violation of a specific bar rule”—although, in fact, OPR cites the same rule its earlier drafts accused Bybee and Yoo of violating. In a catch-22, Margolis faults OPR for switching to the framework he insists is the proper one.

To be sure, one problem for OPR is that very little case law exists interpreting the rule requiring lawyers to give candid, independent advice, and none of it deals with facts remotely like this case. In its early drafts, OPR spelled out standards for when an opinion clearly violates candor and independence:

 1. Exaggerating or misstating the significance of the authority that supported the desired result; 2. Ignoring adverse authority or failing to discuss it accurately and fairly; 3. Using convoluted and counterintuitive arguments to support the desired result, while ignoring more straightforward and reasonable arguments contrary to the desired result; 4. Adopting inconsistent reasoning or arguments to favor the desired result; 5. Advancing frivolous or erroneous arguments to support the desired result.

If this sounds perfectly obvious, that’s because it is. Bybee and Yoo objected that these standards do not come from the case law on Rule 2.1. That is true, because the case law simply has never dealt with lawyers tailoring their advice to yield the client’s desired result: the lawyer as absolver or indulgence-seller. Margolis concludes that the absence of case law on standards of candor means that the standard is ambiguous. But that certainly does not follow. Otherwise, any law that has never been interpreted by a court would automatically be ambiguous.

Margolis also isn’t even sure that standards of candor apply to OLC lawyers. He approvingly quotes Jack Goldsmith’s testimony that it’s an unsettled question whether OLC should offer “neutral, independent, court-like advice” or something “more like … an attorney’s advice to a client about what you can get away with. …” Wow. What about the president’s constitutional obligation to faithfully execute the law, which OLC is supposed to help the president discharge? Whatever you think “faithful execution” means, it surely isn’t “what you can get away with.”

Still, proving a guilty mental state for the OLC lawyers has always been the biggest challenge for the ethics investigation, and I have long thought that the case against the torture lawyers could founder. To prove that legal advice isn’t candid, as the ethics standard requires, you need to prove that the lawyers themselves didn’t believe it and therefore were offering it in bad faith. How do you do that?

Basically, there were two roads OPR could take. It took them both: First, OPR looked to the internal evidence of the memos themselves. The Office of Legal Counsel in the Justice Department is arguably the most elite law firm in the country. Yoo and Bybee are very capable, intelligent, and well-trained lawyers. If they produced an opinion riddled with weird arguments, cherry-picked quotations, and inexplicable omissions, the natural inference is that they weren’t being candid. The famous “empty head, pure heart” defense simply doesn’t wash when you are talking about OLC. Much of the OPR report tries to show—at elaborate length—that the arguments in the torture memos are so bad and so tendentious that lawyers of this caliber could not have produced them in good faith. That is what the five standards in the earlier drafts were about. Margolis’ response is that the arguments in the torture memos are bad, but not that bad. This may look like a debate about the law, but really it is a debate about the mental state of the lawyers who wrote the memos: Did they twist the law so much that they simply couldn’t have done it in good faith?

Yoo cited legal authorities (often with dubious interpretations) to support his conclusions. Yet somehow he managed to omit all the authorities on the other side—dissenting judicial opinions, later opinions by the same courts he did cite, and even Supreme Court decisions. This is first-year law-school stuff. OPR discusses one instance in which Yoo failed to mention some executive branch statements that undermined his argument—even after his assistant brought them to his attention. Margolis reads all this as merely a failure to be “thorough” and responds that “the requirement to be thorough does not necessarily require that any memorandum setting forth the attorney’s opinion communicate to the client every countervailing argument and every non-controlling fact.” This response completely misses the point. Cherry-picking authorities so that you mention those on your side and leave out the rest is not a failure of “thoroughness.” It is evidence of bias.

Sometimes Margolis uses a divide-and-conquer strategy to downplay instances in which the Yoo-Bybee memo states a position one-sidedly or omits opposing authority or (in one instance) falsifies what a source actually says. Thus, of a tendentious misreading of a leading foreign case: “Even if the analysis represented a serious deficiency, it was not likely to have resulted in prejudice to the client, and therefore, in and of itself, does not constitute a violation” (italics are mine). Yoo’s falsification of what a law review article said (it stated that the law of self-defense does not work for torture; Yoo cited it for the opposite proposition) is “too inconsequential to support a finding of misconduct in and of itself.” The problem with Margolis’ pooh-poohing is that all these separate grains of sand really do add up to a heap.

Margolis also makes short work of the Yoo-Bybee memo’s strategic omissions. United States v. Lee is a 1981 case prosecuted by the Reagan Justice Department dealing with a Texas sheriff and his deputies who were convicted for water-boarding men they arrested to make them confess. The U.S. appeals court called the technique “torture” a dozen times, and the case comes up instantly if you type “water torture” into a legal database, as any competent lawyer would do if asked to research whether water-boarding is torture. Neither Bybee memo mentions the case, but this does not trouble Margolis, who blows off the omission of Lee because “the opinion does not describe the technique.” True enough, but the government’s brief in the Lee case does, and the brief is readily available on legal databases. It turns out that the only difference between the technique described in Lee and water-boarding is that the Texas sheriffs used a tilted chair instead of a tilted board.

Margolis also speculates “that Congress might have adopted a definition of torture that differed from the colloquial use of the term,” in which case Lee could be dismissed on the basis that it wasn’t using the word torture in a legal sense. The trouble is that the speculation is wrong, as Margolis would have discovered by looking at dictionaries from the time of Lee. (In fact, the legal definition of torture is so close in wording to that of the 1971 Oxford English Dictionary that it seems likely that the law and the treaty were following the dictionary.) Examining briefs and dictionaries is exactly the kind of research lawyers do, and we know from the torture memo itself that Yoo consulted several dictionaries, including older ones. Margolis’s cavalier treatment of Lee is especially noteworthy because Senator Sheldon Whitehouse, D-R.I., closely questioned former Attorney General Michael Mukasey about the case in a July 2008 hearing and focused attention on it again at a Senate judiciary committee hearing in May 2009. The Senator actually displayed a poster board highlighting the word torture in the Lee opinion.

The second road OPR took to establish that Yoo and Bybee gave bad-faith legal advice allowing torture was to look at the circumstantial evidence. If Yoo and Bybee were under pressure from the CIA and the White House to produce an opinion that is as permissive and reassuring to interrogators as possible, the natural inference is that the opinions are shoddily one-sided because the lawyers needed to reach a desired result.

OPR investigators asked John Yoo if he added his analysis giving enormous latitude to the commander-in-chief in response to the CIA. The usually smooth and articulate Yoo fumbled for an answer:

Well, I know—well, I mean, they wanted, you know, this declination from the Criminal Division which we couldn’t provide. So it wouldn’t be—I mean, I just don’t remember whether that was a response to a specific—it kind of makes sense that it would have been that we could have said, look, you know, we can talk about what happens if you go over the line, but we’re not saying we would approve what went—you know, anything that happened, but I don’t have any real—like for example, I don’t remember sitting in a meeting and saying, oh, well, we can’t provide a declination, but we could do this. But it makes sense, although I don’t have any memory of it.

Translated: The CIA asked the Justice Department for an advance assurance (the “declination”) that government lawyers would decline to prosecute interrogators for torture. When DoJ said no, the CIA came back to ask for new material in the torture memo that would do the job. The result was the notorious argument that the commander-in-chief can authorize torture or anything else, as well as another argument that it would be unconstitutional for Congress to try to force the executive branch to prosecute torturers.

Now, the commander-in-chief override is the one argument in the torture memo we can be pretty sure John Yoo believes, based on his published scholarship. (Check out Page 64 of the OPR final report, where Yoo imperturbably testifies that if the president orders “a village of resistants to be massacred” it falls within his legal powers. The questioner repeats the question: “Q: To order a village to be [exterminated]? A: Sure.”) But that is not the point. The question is whether the CIA was asking for changes to the memo. And Yoo’s answer is that “it makes sense” to suppose they were. Another OLC lawyer recalled that Yoo told him the commander-in-chief sections were included because “they want them in there.”

Margolis objects to OPR’s inquiry into whether the OLC lawyers were being told what result their client wanted—after all, lawyers usually know what result their clients when they go to draft a legal opinion. Again, though, this misses the point. If the OLC lawyers were being pushed by the White House or CIA to reach a certain result, that would be evidence that the contorted lawyering in the memos was deliberate.

Of course, the best evidence of what the lawyers were thinking, who they were talking to, and what pressures they faced, might be their e-mail traffic. But the OPR report informs us that “most of Yoo’s emails had been deleted and were not recoverable.” The same is true of the e-mail of Patrick Philbin, another OLC lawyer who had read the Yoo-Bybee memo and advised Bybee to sign it. (This itself is troubling; I have been told by government lawyers that their deleted e-mails are always stored and recoverable.)

Without the e-mails, it is harder to know John Yoo’s state of mind as he composed the torture memos. And yet the other evidence amply justifies OPR’s conclusions. What does Margolis’ rejection of OPR’s finding of professional misconduct mean? In practical terms, very little. DoJ will not refer Bybee and Yoo for bar discipline, but state bar authorities are badly equipped to take on an ethics investigation of this magnitude. They tend to be cautious and politically timid, and experts have told me that the probability of action from either the D.C. bar counsel or his Pennsylvania counterpart was close to zero. (The Pennsylvania bar has already refused once to investigate John Yoo.) Even if Margolis had followed OPR’s recommendation, the cases would almost certainly have wound up in the dead letter box. Nonetheless, this is a bitter outcome for those who think that torture devised at the highest levels of government disgraces us as a nation.

At the same time, the outcome isn’t entirely happy for Bybee and Yoo, whose handiwork OPR excoriates and even Margolis rejects. For the last year, newspapers reported that DoJ was withholding the OPR report so it could be weakened and watered down. In the end, Attorney General Eric Holder released all three versions of the report—redacted, to be sure—along with Margolis’ memo. This helps, even as Margolis strikes a blow against accountability. Margolis gets a lot wrong in his memo, but he did get one thing right. “OPR’s findings and my decision are less important than the public’s ability to make its own judgments about these documents and to learn lessons for the future.” One lesson from this sorry episode is that in America we don’t do accountability for government officials who approve torture. But perhaps that won’t be the only lesson we learn.

Become a fan of Slate on Facebook. Follow us on Twitter.