U.T. v. Longhornsingles.com Part 2 – state email filtering
White Buffalo v. U.Texas, Austin, No. 04-50362-CV0 (5th Circuit, August 2nd, 2005)
This second post of a two-part posting continues looking at the 5th Circuit’s opinion in this unique case. The first post focused on the CAN-SPAM Act and how the court applied it in this case. This second post will talk about 1st Amendment complications when a government agency or entity (in this case the University of Texas) adopts a filtering policy or blocks email. Part One concluded that the 5th Circuit rather bungled the application of CAN-SPAM, even if eventually got to the correct conclusion. In contrast, in Part Two, the 5th Circuit seems quite aware of the more nuanced 1st Amendment issues and applies them quite well, with an eye to the unique concerns that email technology can raise. Oh, and the 5th Circuit couldn’t resist dropping a slightly hidden filtering time bomb into the opinion as well…
Background
Part One gives the details of the case, so we’ll jump right on in to the 1st Amendment issues.
The 1st Amendment
White Buffalo asserted that the University of Texas was restricting its speech in violation of the First Amendment. We’ve heard about the 1st Amendment complications of regulating spam before, and even seen laws similar to CAN-SPAM survive 1st Amendment challenges, but a quick primer might help focus things.
The 1st Amendment says “Congress shall make no law…abridging the freedom of speech.”
It is worth pausing to figure out whether the 1st Amendment even applies – the text does say “Congress,” and U.T. is certainly not they. Originally, as the Supreme Court ruled in Barron v. Baltimore, the Bill of Rights restricted only the federal government. But we cannot go home just yet. In 1925, the Supreme Court, in Gitlow v. New York, recognized the 14th Amendment might have changed things and gave birth to the doctrine of incorporation, under which provisions of the Bill of Rights were deemed to restrict the states as well as Congress.
So the 1st Amendment applies to state actors and state actions.
Question #2 – is the spam filtering policy at U.T. state action? U.T. did try to say no, it wasn’t a state actor, but the 5th circuit dismissed this argument as “meritless.” It’s beyond the scope of this post, but just when subdivisions of a state are or aren’t state actors is a contentious question that often gets litigated, so U.T. was probably making a plausible, if weak, argument. CAN-SPAM, for instance, tries to avoid this considerable question by defining both states AND their subdivisions into the law.
Public or Private Fora? The 5th takes a pass.
1st Amendment jurisprudence distinguishes between ‘public’ and ‘private’ fora, and places different levels of scrutiny on government speech restrictions depending on whether they restrict speech in public or private. For example, government prohibitions on employees posting “for sale” ads on an agency’s office bulletin board may be allowed with less scrutiny than government restrictions on street corner pamphleting. The 5th Circuit recognizes that email has elements of both, and rather than answer what it calls a “dicey but admittedly important” question, it takes a clever pass. Recognizing that the ‘public forum’ test is the more stringent of the two tests, it assumes without deciding that email filtering restricts public forum activity and analyzes it under that tougher test. Strategic opinion-writing perhaps, but you can almost hear the disappointment dripping from campus speech & 1st amendment organizations’ pens as they report on the decision.
Central Hudson and commercial speech
Diving into the mechanics of analyzing government restrictions on commercial speech, the 5th Circuit pulls out the relevant Supreme Court case on the subject, Central Hudson. This 1980 case not only recognized that commercial speech was different, and subject to lesser protections, than other types of expressive speech, but gave a detailed analytical framework for reviewing government restrictions on commercial speech.
Where the rubber meets the road: the 4 Central Hudson tests.
1) Is it misleading or speech about illegal acts? The Court whizzes through this – no one alleged misrepresentation, or that the ads were for illicit substances, etc. – they even comply with CAN-SPAM.
2) Does U.T. promote a substantial interest in adopting a filtering policy? The 5th circuit recognizes that U.T. has a legitimate interest in adopting a policy to benefit its users, especially after users have complained about the unsolicited emails. No one is alleging the filtering is pretextual or discriminatory. Sounds like a legitimate interest, not that hard of a question. U.T. also says it has an interest in not stressing or clogging its mail servers. The court correctly pauses a bit at this argument. Mail servers are designed to deliver mail, complaints about overloading shouldn’t be rubber stamped, but require inspection, the court says, taking a shot at the opinions in eBay, Inc. v. Bidder’s Edge and other court cases that seem too ready to find wrongdoing in any undesired computer loads. The court follows up on this criticism in considerable detail under the final test.
3) Does filtering address the problem? Here the court basically says ‘well, duh!’, stating “One can hardly imagine a more direct
means of preventing commercial spam from appearing in account-holders’ inboxes and occupying server space than promulgating a
policy that excludes such material…” It is worth noting, although the court doesn’t, that spam filters aren’t perfect, they don’t block all spam. The Central Hudson test doesn’t require perfection, however, but only that the policy work.
4) Is the solution tailored to prevent or minimize collateral damage? Here’s where the 5th Circuit spends the meat of its energy, finding that, by blocking email from a sender’s IP address for known bulk spammers, U.T. is helping its users. More specifically, the court finds that by only blocking in response to user complaints or system volume triggers, U.T.’s filtering is sufficiently tailored to minimize collateral damage. Then the court throws in a parenthetical bombshell, reminding that the filtering must be content- and viewpoint-neutral…
The content-neutral gorilla – don’t bring it out of the cage if you aren’t willing to feed it.
This last half-sentence is easily the 800-pound gorilla in this case.
Granted, this case was not about filtering based on content, but about filtering based on mail volume. It is perhaps right that the 5th Circuit shouldn’t spend much time on an issue not technically before the court, but content-based spam filtering is pervasive, and in all likelihood there are many state agencies that employ content-based email filtering, in part because it blocks a significant amount of spam. (Think about all those spam filters set up to screen out words like ‘viagra,’ ‘porn,’ ‘free offer,’ ‘multi-level marketing,’ or the like.) Should the 5th circuit have brought up such a potentially large problem so tangentially without addressing it squarely?