The Fourth Circuit issued a decision last week that for lack of a better description beat back an anti-spam activist.  The decision contains the motherlode of arguments for anti-anti-spam lawyers.  (Omega World Travel, Inc. v. Mummagraphics, Inc., 2006 U.S. App. LEXIS 28517, No. 05-2080 (4th Cir. Nov. 17, 2006)) [access a pdf version here].  The case decides several typically contested legal issues.  It also contains helpful general language to the effect that while spam is a serious problem, the law should not be stretched to address situations not intended by Congress.  This is a constant refrain for lawyers who defend spam cases.  Now they finally have some powerful judicial backup.

In Mummagraphics, the Fourth Circuit affirmed the decision of the district court (Judge Brinkema, who sits in AOL’s home jurisdiction and is no stranger to these types of cases).  The court found CAN-SPAM preempted the state cause of action – based on immaterial errors in the emails – alleged by Mummagraphics, Inc. (Mummagraphics).  The court also rejected the underlying CAN-SPAM claims.  The decision is important for one simple reason:  anti-spam lawyers (and plaintiffs) often advance the exact arguments advanced by the plaintiff in this case.  Lawyers on the other side know these arguments lack merit, but do not have any court decisions to back them up.  As a result, a vicious settlement cycle results.  This case probably represents the start of the tide turning in the other direction.  

Preemption

CAN-SPAM preempts state law dealing with commercial email except to the extent state law is aimed at preventing falsity or deception.  The few published federal decisions addressing preemption reach the opposite conclusion as Mummagraphics.  For example, Gordon v. Impulse Marketing Group, Inc., 375 F. Supp. 2d 1040, 1042 (E.D. Wash. 2005), concluded that CAN-SPAM does not preempt claims under Washington’s anti-spam statute.  (See also Beyond Sys. v. Keynetics, Inc., 422 F. Supp. 2d 523, 528 (D. Md. 2006); but see White Buffalo Ventures, LLC v. Univ. of Texas, 420 F.3d 366, 371 (5th Cir. 2005), cert. denied, 2006 U.S. LEXIS 61 (Jan. 9, 2006).)  One would have expected a similar result in Mummagraphics, given that the Oklahoma spam statute roughly tracks the Washington statute.  The Washington statute prohibits the misrepresentation of “any information in identifying the point of origin or the transmission path” and is not limited to material misrepresentations.  See RCW § 19.190.020.  Gordon found claims under this statute to survive CAN-SPAM.

Mummagraphics notes that the Oklahoma statute “seems to reach beyond common law fraud and deceit . . . [and] is not limited to inaccuracies in transmission information that were material.”  The court acknowledges there are no Oklahoma cases interpreting the Oklahoma spam statute.  According to the court, the percise scope of the Oklahoma statute is irrelevant:  Mummagraphics’ state law action is premised on immaterial errors (see the discussion below), and for this reason cannot survive CAN-SPAM’s premption clause. 

In analyzing the scope of preemption, the court focused on the words “falsity” and “deception.”  First, because falsity appears with the more serious word deception, this should color the interpretation of falsity.  The exceptions should cover deception and "other torts involving misrepresentations, rather than to sweep up errors that do not sound in tort."  Second, CAN-SPAM uses the terms (albeit in headings) “materially false,” and “materially misleading.”  Congress intended CAN-SPAM to cover materially false or misleading emails.  Finally, and most importantly, the court reasons that

Congress did not intend ‘falsity’ to encompass bare error because such a reading would upset the Act’s careful balance between preserving a potentially useful commercial tool and preventing its abuse.

The narrower interpretation given to the preemption clause allows it to avoid commerce clause issues as well as First Amendment concerns (!).  (I had written previously that both of these are essentially dead issues in the context of CAN-SPAM analysis, but this decision illustrates how they could have an indirect influence on the court’s interpretation of CAN-SPAM and preemption.) 

CAN-SPAM Violations 

With respect to the CAN-SPAM claims, Mummagraphics first asserted the emails in question contained “header information that is materially false or materially misleading.” Mummagraphics claimed the emails violated the statute because they originated from one server (FL-Broadcast.net) but contained a different "from" line ([email protected]).  Established, legitimate, advertisers do this all the time, particularly where a third party service provider or service is used to send the email – the “sender” is technically someone other than the company whose products and services are being sold, but the advertiser wants the recipient to know its identity and relationship to the email.  The court smacks down Mummagraphics’ often-used argument, noting that the emails are “chock full of methods to ‘identify, locate, or respond to’ the sender or to ‘investigate [an] alleged violation’ of the CAN-SPAM Act.”  In other words, everyone knew where the email came from so there could be no violation of the header information prong.  (The fact that the plaintiff was able to locate and call the sender’s general counsel is also a testament to this.)  The court also notes that in the context of an email which legitimately proposes a commercial transaction (as opposed to a phising scam or something which injects malware into your system) it would come “as little surprise” that the email in question identifies the sender or the company offering the product or service.      

Mummagraphics also asserted a claim based on the sender’s failure to honor an opt-out request.  The court easily disposes of this claim, noting that there was no allegation of a “pattern or practice” of failure to honor an opt-out. 

At the end of the day, this will be widely cited by lawyers who defend CAN-SPAM cases.  You can’t really get better than when a court says the statute “does not impose liability at the drop of a hat.”  Obviously, time will tell as to the precise effects of this decision. 

A few other minor points:

1. The court enclosed the words “spam” and “spammer” in quotes.  This could be a signal that these are legally loaded terms that carry little meaning when used by the parties in pleadings.
   
   
2. The court’s brief First Amendment discussion (fn. 2) will no doubt get some mileage (maybe in the Jaynes case?).
   
   
3. The result on the state claim feels right, but the approach is slightly awkaward.  The court have easily interpreted the state law to avoid preemption and the issue noted in fn 2 -- i.e., concluded that Mummagraphics failed to state a claim under OK law.
   
   
4. The decision goes out of its way to mention Mummagraphics’ refusal to use the opt-out mechanism.
   
   
5. It's disappointing that the court does not mention/address the other preemption decisions.
   

 [link to the decision originally via How Appealing] [corrected to include references to the White Buffalo and Beyond Sys. cases; and again correct to clarify that White Buffalo does not reach a contrary conclusion to Mummagraphics.]