Back on September 30, 2004, I blogged about press miscoverage of a New York court decision, Doe v. Ashcroft, that invalidated the national security letter provision of the 1986 Electronic Communications Privacy Act codified at 18 U.S.C. 2709. The MSM reported that the decision had found a key part of the Patriot Act unconstitutional, a description apparently resulting from a misleading ACLU press release announcing that the decision had “struck down an entire Patriot Act provision.” This wasn’t quite right: Section 2709 was a 1986 law expanded by the Patriot Act, but the Doe decision invalidated 2709 for reasons apparently unrelated to the Patriot Act’s amendments. The key mechanism found unconstitutional was from 1986, not 2001. Recognition that the press had relied on the ACLU’s misleading press release led to much media attention and corrections at most major newspapers.
Today Slate offers a defense of the ACLU’s position in an essay by Robert Poe. According to Poe, those attempting to delink the Doe decision from the Patriot Act are engaging in a “spin campaign,” while the ACLU’s characterization is essentially right. According to Poe, the ACLU’s decription
. . . was a reasonable description because the NSL law was one of a number of measures the Patriot Act souped up to drive the most powerful vehicle ever assembled for gathering private information about American citizens. Conservatives such as Sen. John Cornyn of Texas, on the other hand, argued that the ruling had nothing to do with the Patriot Act. They noted that the Electronic Communication Privacy Act first authorized NSLs in 1986, and they’ve been in use, complete with gag provision, ever since. They claim the Marrero decision did nothing but overturn a nearly two-decades-old law, meaning the basic principles that inspired the Patriot Act remained above reproach, or at least undamaged.
Why was the ACLU’s decription “reasonable” in Poe’s view while “the conservatives” were engaging in a “spin campaign”? Poe’s key argument seems to be this:
Marrero ruled that NSLs [permitted by 18 U.S.C. 2709], at the time of the lawsuit, presented a serious enough threat to citizens’ constitutional rights that the lack of access to judicial review was a fatal flaw. That means those who want to claim that the ruling had nothing to do with the Patriot Act have their work cut out for them. To make their case, they have to penetrate the mind of the judge and discern that he would have made the same decision about pre-Patriot Act NSLs that he did about post-Patriot Act ones. And that would be a hard sell, given that the earlier NSLs carefully targeted only spies and terrorists.
As I understand it, Poe’s point is that no one really knows whether Judge Marrero’s decision has anything to do with the Patriot Act. We can’t tell whether Judge Marrero would have struck down Section 2709 before the Patriot Act, he concludes, so those trying to say that the decision had nothing to do with the Patriot Act are in effect guilty of expressing false certainty. This seems a bit inconsistent with his characterization of the ACLU’s position as “reasonable” and Cornyn’s as “spin,” in that it would seem to apply equally to the ACLU’s claims. But that seems to be Poe’s argument.
I don’t think Poe’s argument ultimately works, though. To be sure, Doe v. Ashcroft is a meandering and often incoherent 122-page opinion; the reasoning in the case is very difficult to follow, and “penetrat[ing] the mind” of Judge Marrero could be helpful. But as I read the opinion, the key to the unconstitutionality of Section 2709 according to Marrero is a lack of statutory clarity as to whether NSLs can be challenged by recipients, combined with the breadth of its gag order provision. The key language in both cases comes from the 1986 statute. Marrero reads the 1986 language of Section 2709 as not allowing recipients to challenge NSLs and the gag order as very broad. He then concludes that (a) the inability to challenge the order violates the Fourth Amendment, (b) it may also violate the First Amendent in some cases, threatening speech, and (c) the 1986 gag order is too broad to survive First Amendment challenge.
As I understand Poe’s argument, his point is that Marrero might have reached a different result on argument (b) under the more limited, pre-Patriot Act version of NSLs. That’s possible: Marrero’s argument in this section is unclear, and as Poe recognizes we just don’t know the answer here. But (b) is only an alternative holding, and Poe’s point does not seem applicable to the Fourth Amendment argument in (a) or the gag order section in (c). Poe has recognized uncertainty about (b), but I don’t think that’s quite enough to make the ACLU’s characterization seem correct.
Ultimately, I think it’s fair to say that Doe v. Ashcroft struck down a 1986 law expanded by the Patriot Act; that most of the reasoning of the decision is focused on the 1986 law; but that there is uncertainty as to whether the outcome of one of the arguments would have changed had the Patriot Act never been enacted. Does that mean that the ACLU was right, and that others have been merely spinning? Others can disagree, but it doesn’t look that way to me.
(Thanks to my friend Dan Markel for the link. Oh, and my apologies to VC readers who might think it weird to devote such a long post to this question. It’s of interest to me, and it’s quite difficult to address the issues well without being rather detailed.)
Comments are closed.