CommunityEmptywheel Archive

Did DOD Have ANY Authorization for Torture after 2004?

There are a couple of things that have been bugging me about the authorizations DOD got for interrogations.  It’s not clear what kind of authorization DOD used to justify detainee interrogations after the Yoo memo was withdrawn in 2003-2004–they had no overall interrogation approval from OLC. While it’s possible they were just relying on already-existing DOD documents, there are hints that DOD was either relying exclusively on the CIA’s more expansive authorizations (that included waterboarding), or they had some alternative approval that may not have involved OLC at all.

As I’ve shown (here and here), in March 2004, DOD requested approval to use–at the least–extended isolation with detainees. In response, Jack Goldsmith and Steven Bradbury started trying to replace the 2003 Yoo memo.

At precisely the same time, Goldsmith was working through the mess created by the Legal Principles document. As you recall, faced with clearly illegal conduct and with the opportunity to investigate that conduct themselves in 2003, CIA worked back channel with Jennifer Koester and John Yoo to summarize the legal advice given on torture, going so far as to claim certain techniques (like abdominal slap and diapers) had been approved when they hadn’t been. During that period, Koester and Yoo gave CIA an opportunity to review and provide input on the 2003 Yoo memo. Then, Koester and Yoo relied on the Yoo memo for several of the claims they made in the Legal Principles. That raises the possibility that one reason the Yoo memo was so bad (it was even more permissive than the Bybee One memo) was to help CIA avoid criminal liability for crimes already committed.

At the very least, this is proof that CIA and DOD were both relying on advice given to the other agency to justify their own agency’s actions. We know DOD used the Bybee memos (and oral authorization from Yoo based on that analysis) to authorize its treatment of Mohammed al-Qahtani in 2002-2003. And the Legal Principles show CIA was using the Yoo memo, written for DOD, to authorize its treatment of multiple detainees in anticipation of the CIA IG Report. In other words, though DOJ liked to maintain the fiction that the approval tracks for CIA and DOD were separate, they weren’t, at least not when John Yoo was involved.

And that was becoming crystal clear in spring of 2004. (In the same phone conversation in which Goldsmith confirmed that the Legal Principles weren’t an official OLC document, he also asked Yoo for details of his verbal authorizations to Jim Haynes leading up to the al-Qahtani torture, so he clearly pursued these issues in tandem.)

Yet after that, CIA’s memos got withdrawn and replaced. DOD’s Yoo memo reportedly was withdrawn. But no formal guidance from OLC ever replaced it.

So what happened after that point?

The Daniel Levin Memo

My concerns about DOD’s later authorizations stem partly from a memo Daniel Levin wrote John Ashcroft and Jim Comey in September 2004 to summarize all the advice OLC had given on torture. It shows the state of affairs as it existed in September 2004–the way in which DOJ was transitioning from authorizations based on Yoo’s crappy memos to more arguably defensible authorizations. I wrote a detailed post on Levin’s memo here, but here’s how Levin described that state of affairs (the following is my summary of his summary, except in one direct quote from his DOD section).


1. Previously Given: Bybee One memo

2. Current/Pending: This is redacted, but must describe what became Levin’s own December 30, 2004 memo that replaced Bybee One.


1. Previously Given: Bybee Two memo

2. Current/Pending:

a. Ashcroft’s July 22, 2004 letter reauthorizing 9 techniques from Bybee Two

b. Four letters pertaining to three individual detainees: An August 6 letter authorizing waterboarding for use with one detainee and an August 26 letter adding four new techniques, including water dousing, for use with that same detainee; a September 6 letter authorizing 12 total techniques including water dousing but not waterboarding; and a September 20 letter authorizing those same 12 techniques with a third detainee.

c. This is redacted but must describe the memo that would become the May 10, 2005 Techniques memo, and possibly also the May 10, 2005 Combined memo.

d. Mostly redacted, but this bullet describes the May 30, 2005 CAT memo.


1. Previously given

a. Yoo memo

b.Levin described OLC approval of 24 DOD techniques this way:

In addition, we approved 24 specific techniques the use of which the Secretary of Defense approved. Although it is not entirely clear to me when that was done it was reaffirmed, for example, in a July 7, 2004 letter from Jack Goldsmith to Scott Muller (referring to approval of both CIA and DOD techniques) and also in a July 17, 2004 fax by Jack [Goldsmith].

2. Current/Pending; [entirely redacted]

The general and the CIA approval is everything we’d expect to see, with the possibly interesting detail that in September 2004, Daniel Levin (who drafted most of the Techniques memo but who was ousted before Bradbury wrote the Combined memo) saw just one memo to be forthcoming addressing individual torture techniques.

But Levin’s understanding of DOD’s approval is interesting for two reasons. First, he confesses that he has no idea how and when OLC approved the 24 DOD techniques (these were techniques Rummy approved in April 2003), only that somewhere along the way Goldsmith sanctioned them (this accords with what we know, but it also means OLC conducted no independent review of them; this is further important because it’s where some things–like isolation–got approved). And, Levin believes as of September 2004 that OLC is actively working on a memo for DOD analyzing individual torture techniques, one which, according to the OPR Report, was never completed.

There’s one more notable detail. Levin references the July 7, 2004 Goldsmith letter to CIA authorizing the 24 techniques. But he also references a July 17, 2004 fax also authorizing the 24 techniques. We’ve seen the former, but not (to the best of my knowledge) the latter. And that July 17 fax was written on Goldsmith’s last day at OLC, after having moved up his last day from August 6 to July 17 for some reason.

The OPR Report

Which brings me to what the OPR Report says about DOD Authorization.

As I said before, it reports that in April, Goldsmith started working on a replacement for the Yoo memo with Steven Bradbury.

Sometime in April 2004, Goldsmith began working on a replacement draft for the Yoo Memo, assisted by then Principal Deputy AAG Steve Bradbury and several OLC line attorneys. [this statement is followed by a redacted paragraph, suggesting something about the replacement remains highly sensitive]

After a discussion of the Legal Principles/Bullet Points controversy (which I mentioned above) the OPR Report returns to a discussion about the drafting of the replacement for the Yoo memo.

The first draft of the replacement memorandum was produced in mid-May 2004, and at least 14 additional drafts followed, with the last one dated July 17, 2004. Beginning with the sixth draft, dated June 15, 2004, specific criticisms of the Yoo Memo were discussed in footnotes. Although the criticism was removed from later drafts, Goldsmith told OPR that it was not removed because of any doubts about its accuracy. Rather, Goldsmith ultimately concluded that it was unnecessary to specifically address the errors.

A couple of notes about this timing. Goldsmith told Jim Haynes not to rely on the Yoo memo in December 2003. But he didn’t start his efforts on replacing the Yoo memo until after DOD asked to use–at a minimum–extended isolation in March 2004 (and after DOD, but according to the Terror Presidency, not Goldsmith, knew about Abu Ghraib). The first draft was not completed until after the Abu Ghraib scandal had broken (remember that Goldsmith was very busy trying to salvage Cheney’s illegal wiretap program between March 10 and May 6, 2004). Then, after writing six drafts between the time he started this process and mid-June, he started attacking the Yoo memo directly.

Significantly, the very day Goldsmith told Ashcroft he would withdraw the Bybee One memo, he also had Bradbury put this into a footnote on the first page of a draft memo purportedly replacing the Yoo memo:

The Yoo Memo “is flawed in so many important respects that it must be withdrawn.” June 15, 2004 draft at 1, n.l.

This language speaks of withdrawing the memo as something that had yet to be done, suggesting that the Yoo Memo was not, in fact, operationally withdrawn yet in June 2004. Goldsmith was still making the case to do so in the footnotes of its replacement! More importantly, he started making the case in those footnotes on the same day he finally decided to withdraw the Bybee One memo. That’s not an accident. As Goldsmith admits in Torture Presidency, he ultimately withdrew the Bybee One memo because of his thinking on the Yoo memo.

In the end I withdrew the August 2002 opinion even though I had not yet been able to prepare a replacement. I simply could not defend the opinion. I had rejected its reasoning in the March 2003 opinion, and I knew that the August 2002 opinion would eventually suffer the same fate.

In addition to the footnote above, the OPR includes several others dated June 15 that could also be applied directly to the Bybee One memo, notably those addressing Commander-in-Chief power and possible defenses.

In other words, what Jack Goldsmith did with his footnotes on June 15 was withdraw the Get Out of Jail Free Card from both DOD and CIA at the same time. And the next day, he tendered his resignation, to go into effect on August 6.

The OPR Report is very opaque about how this process related to Goldsmith’s departure and why he left three weeks early. The Report includes the following details peppered throughout the discussion of the withdrawal of the Yoo and the Bybee One memos.

Goldsmith left the Justice Department on July 17, 2004, before he was able to finalize a replacement for the Yoo Memo. On July 14, 2004, then Associate Deputy AG Patrick Philbin testified before the House Permanent Select Committee on Intelligence as to the legality of the 24 interrogation methods that had been approved for use by the Defense Department. Sometime thereafter, the Defense Department reportedly informed OLC that it no longer needed a replacement for the Yoo Memo.


Shortly after the Bybee Memo was leaked, Goldsmith was asked by the White House if he could reaffirm the legal advice contained in the Bybee Memo. Because the analysis in that document was essentially the same as the Yoo Memo, which he had already withdrawn, Goldsmith concluded that he could not affirm the Bybee Memo. He consulted with Comey and Philbin, who agreed with his decision, and on June 15, 2004, Goldsmith informed Attorney General Ashcroft that he had concluded that the Department should withdraw the Bybee Memo. Although Ashcroft was “not happy about it,” according to Goldsmith, he supported the decision. The following day, June 16, 2004, Goldsmith submitted a letter of resignation to become effective August 6, 2004.


Goldsmith was determined to complete his replacement for the Yoo Memo before he left the Department, and he also assigned an OLC line attorney to prepare a replacement for the Bybee Memo.93 At some point during the summer,
however, it became apparent that the Yoo Memo could not be replaced by August, and Goldsmith decided to advance his departure date to July 17, 2004. [Note, this last passage is followed by a long redaction.]

93 Several replacement drafts for the Bybee Memo were prepared under Goldsmith’s direction, the last of which was dated July 16, 2004.

So the story is that at some point Goldsmith decided he couldn’t finish the Yoo memo–at least not by August–and so he just left.

And while Levin seems to have believed two months later that it was still in the works, the OPR Report says “sometime thereafter [after July 17], the Defense Department reportedly informed OLC that it no longer needed a replacement for the Yoo Memo.” There’s a lot that stinks about that statement: the term “reportedly,” which suggests that OPR saw no documentation about that decision, the vagueness regarding the timing, and the apparent disinterest in explaining why DOD would no longer need the memo after it had been deemed so important in earlier periods.

Now, the OPR Report, at least, appears to believe the 24 techniques originally approved by Rummy was what governed DOD interrogations after the withdrawal of the Yoo memo (suggesting that DOD was satisfied with those 24 techniques). It mentions Philbin’s statement to HPSCI on July 14, as if that were definitive. Though Levin’s comment–noting that he has no idea when and how OLC authorized those techniques–suggests some doubt.

I’m particularly intrigued, though, by Levin’s mention of a July 17 Goldsmith fax reiterating approval of the 24 techniques.

One explanation for that fax is that it was actually draft number 14 reported in the OPR Report–that that reiteration of approval for the 24 techniques was a draft OLC memo. That would be significant because it would suggest that Goldsmith was combining the general authorization for DOD torture with specific techniques.  He may have tried to do that. The OPR Report describes his confusion as to why Bybee One and Two were dated with the same date, suggesting he thought a generalized memo distinct from a specific one might be particularly dangerous. If so, would that suggest that one reason Goldsmith realized he couldn’t finish Yoo replacement before August might be because someone objected to including actual techniques in the more generalized authorization?

Then again, it’s possible that Goldsmith just sent a fax on his way out the door in an attempt to make sure DOD stuck to the limits of the 24 techniques. Significantly, the problem that both Goldsmith and Philbin had with the Yoo memo was that it would serve as a blank check for new torture techniques. Goldsmith even complained that Philbin had given an oral caution–but no written one–that DOD should stick to the 24 techniques in 2003, when Yoo issued the DOD memo.

The broad nature of the memorandum’s legal advice troubled [Goldsmith] because it could have been used to justify many additional interrogation techniques.. As he later explained in an email to other OLC attorneys, he saw the Yoo Memo as a “blank check” to create new interrogation procedures without further DOJ review or approval.

Philbin responded to that email as follows:

John’s March memorandum was not a blank check at least as of the time [Jennifer Koester] started work at DoD OGC (Summer 2003) because I told her to make sure they did not go beyond the Rumsfeld approved procedures and did not rely on the memo. This was only an oral caution but please do not sell us short by ignoring it.

Goldsmith answered as follows: “I’m not selling anyone short – It’s just that Haynes said he heard nothing about that advice.”

Goldsmith’s memorialization that the 24 techniques (but presumably only the 24 techniques) had been authorized was one of the last things he did at DOJ. That he made the effort suggests that he believed such a written reminder was necessary to ensure DOD stick within limits authorized by OLC (though, as Levin reminds, they really hadn’t been, not formally). That he made the effort also suggests he thought such a warning would work to make DOD stick within the limits of those 24 techniques. Given that the White House issued some kind of memo on torture just 4 days after Goldsmith left (see document 63 at PDF 81), that might be overly optimistic.

In any case, it’s not clear what authorization DOD relied on after Goldsmith left.

Previous post

Republicans Offer Five Amendments to Kill Individual Mandate

Next post

Obama and Congress are the real healthcare advocates



Marcy Wheeler aka Emptywheel is an American journalist whose reporting specializes in security and civil liberties.