Kleffman v. Vonage Holdings Corp.

Opinion

Business and Professions Code section 17529.5, subdivision (a)(2)^[1] (section 17529.5(a)(2)) provides that it is unlawful to advertise in a commercial electronic mail (e-mail) advertisement—commonly known as “spam”—if the advertisement “contains or is accompanied by falsified, misrepresented, or forged header information.” The issue this case presents is whether, under this section, it is unlawful to send commercial e-mail advertisements from multiple domain names for the purpose of bypassing spam filters. We hold that, on the undisputed facts of this case, the answer is “no.”

Factual Background

“The Internet is an international network of interconnected computers” that enables millions of people “to communicate with one another and to access vast amounts of information from around the world.” (Reno v. American Civil Liberties Union (1997) 521 U.S. 844 , 849-850 [ 138 L.Ed.2d 874 , 117 S.Ct. 2329 ].) “In order for the Internet to function, each entity connected to it (e.g., computer, router, network, etc.) must have a unique numeric ‘address.’ A unique identifier is required to enable one connected computer or network to identify and send information to another connected computer or network. Those unique addresses are known as Internet Protocol Addresses or ‘IP addresses.’ [Citation.]” (National A-l Advertising, Inc. v. Network Solutions, Inc. (D.N.H. 2000) 121 F.Supp.2d 156 , 159 (National A-l Advertising).) An IP (Internet Protocol) address consists of “four sets of numbers separated by periods” (Kremen v. Cohen (9th Cir. 2003) 325 F.3d 1035 , 1038), such as “12.34.56.78.” “IP addresses function much like Social Security numbers or telephone numbers: each IP address is unique and corresponds to a specific entity connected to the Internet.” (National A-l Advertising, supra, at p. 159.)

Because the number strings that make up IP addresses can be difficult to remember, the Internet community developed the domain name system, which enables users to link a numeric IP address to a unique and easier to remember domain name, “thereby making it more convenient for users to access particular addresses on the Internet.” (National A-l Advertising, supra, 121 F.Supp.2d at p. 159.) “Domain names—e.g., bettyandnicks.com—consist of at least two groups of alphanumeric characters, each known as a string, separated by a period or dot. The last string—the farthest to the right— denotes the top-level domain. The second-to-last string is the second-level

domain name and identifies the person’s or organization’s Internet computer site.”^[2] (Thomas v. Network Solutions, Inc. (D.C. Cir. 1999) 336 U.S. App.D.C. 74 [ 176 F.3d 500 , 503].)

In March 2007, plaintiff Craig E. Kleffman filed this class action in state court against defendants Vonage Holdings Corp., Vonage America, Inc., and Vonage Marketing, Inc. (Vonage), asserting a claim under section 17529.5(a)(2). As noted above, that section makes it unlawful to advertise in a commercial e-mail advertisement that “contains or is accompanied by falsified, misrepresented, or forged header information.” In relevant part, Kleffman alleged the following: Vonage, by and through its marketing agents, sent him 11 unsolicited e-mail advertisements for its broadband telephone services using “11 different domain names: superhugeterm.com; formycompanysite.com; ursunrchcntr.com; urgrtquirkz.com; countryfolkgospel.com; lowdirectsme.com; yeamfrmore.com; openwrldkidz.com; ourgossipfrom.com; specialdlvrguide.com; and stmggletailssite.com.” These “11 different domain names can [all] be traced to a single physical address” in Nevada where Vonage’s marketing agent “is located.” “None of these domain names provides any indication to the recipient (or its spam filter) that the advertisement is from Vonage.” Vonage’s “use of these multiple domain names . . . reduces the likelihood that an internet service provider will identify these . . . advertisements as spam and block them before they reach the email inboxes of [Kleffman] and class members.” An ISP (Internet service provider) “ ‘may block a message because ... [a] domain name is associated with the sending of high volumes of spam,’ ” so recipients “could easily block all of’ Vonage’s e-mail advertisements “[i]f Vonage and its marketing agents were to use a single domain name to send [those] advertisements.” Vonage “could have easily (and less expensively)” sent all of its e-mail advertisements “using a single domain name,” and “the only reason” it used “multiple domain names is to mislead email service providers and recipients, and their spam filters.” “In other words, Vonage essentially creates multiple identities, as represented by the multiple domain names, in order to ‘spread out’ the total volume of [its e-mail advertisements] and reduce the volume sent via each domain name, a strategy deliberately calculated to trick the ISPs into believing there are multiple senders, when in actuality the emails are sent for

the ultimate single beneficiary: Vonage.” “The multitude of ‘from’ identities falsifies and misrepresents the true sender’s identity and allows unwanted commercial email messages to infiltrate consumers’ inboxes.” Vonage’s use of “multiple domain names to bypass spam filters,” its “failure” to use “a single domain name” in sending its advertisements, and its “failure to identify Vonage in the domain name from which the . . . advertisements were sent, i.e., through the use of a generic subdomain name such as adfor.vonage.com, constitute[] falsified and misrepresented header information prohibited by” section 17529.5(a)(2).

After Vonage removed the case to federal court, it moved to dismiss the complaint, arguing the complaint failed to state a claim under section 17529.5(a)(2). The district court agreed and dismissed the action with prejudice and without leave to amend, stating: “Kleffman does not actually allege that the content of Vonage’s email was false, misrepresented or forged, and indeed points to nothing misleading about any single given email.” “The headers are allegedly falsified because, though they literally and truthfully identify the sender, they are part of a mechanism to avoid anti-spam legislation and therefore imply that they originate from different sources. However, under the plain language of the statiite, which requires that an email message contain a falsified, misrepresented or forged header, the claim fails. The failure to send mail from a single domain name that includes the word ‘Vonage’ is simply not a misrepresentation in any ordinary sense of the word.” “Moreover, while [Kleffman] might characterize an email as containing the implicit misrepresentation T am not from the same sources as the others,’ . . . this is more than the plain language of the statute would bear.”^[3]

Kleffman appealed to the United States Court of Appeals for the Ninth Circuit. Pursuant to rule 8.548 of the California Rules of Court, the Ninth Circuit asked us to decide the following question: “Does sending unsolicited commercial e-mail advertisements from multiple domain names for the purpose of bypassing spam filters constitute falsified, misrepresented, or forged header information under [section] 17529.5(a)(2)?” We granted the Ninth Circuit’s request.^[4]

Discussion

At issue here is the scope of section 17529.5(a)(2), which makes it “unlawful ... to advertise in a commercial e-mail advertisement” that

“contains or is accompanied by falsified, misrepresented, or forged header information.” “ ‘As in any case involving statutory interpretation, our fundamental task [in considering this issue] is to determine the Legislature’s intent so as to effectuate the law’s purpose.’ [Citation.]” (People v. Cole (2006) 38 Cal.4th 964 , 974-975 [ 44 Cal.Rptr.3d 261 , 135 P.3d 669 ].)

In resolving the parties’ disagreement over the meaning of section 17529.5(a)(2), it is useful to begin by noting the matters on which they agree. There is no dispute here that the domain names in question were part of the e-mails’ “header information” within the meaning of section 17529.5(a)(2).^[5] There also is no dispute that the domain names used to send Vonage’s e-mail advertisements, and reflected in the header information of these e-mail advertisements, actually exist and are technically accurate, literally correct, and fully traceable to Vonage’s marketing agents. Finally, there is no dispute that, in light of this conceded fact, the e-mails neither contained nor were accompanied by “falsified ... or forged header information” within the meaning of section 17529.5(a)(2). Thus, the parties agree that the question here is whether the e-mails contained or were accompanied by “misrepresented . . . header information” within the meaning of that section.

Vonage’s answer to this question is relatively straightforward. It asserts that header information is not “misrepresented” within the meaning of section 17529.5(a)(2) unless it contains “a false representation of fact.” Vonage reasons that, when the Legislature drafted the statute, this was “the established legal definition” of the term “misrepresent” for purposes of the tort of misrepresentation, and nothing indicates the Legislature intended to use the term in section 17529.5(a)(2) to convey some other meaning. Applying this definition, Vonage argues that e-mail advertisements from multiple domain names with fully accurate and traceable header information do not violate the statute because they contain no false representation.

Kleffman, quoting Cooley v. Superior Court (2002) 29 Cal.4th 228 , 249 [ 127 Cal.Rptr.2d 177 , 57 P.3d 654 ], asserts that Vonage’s construction violates the rule of statutory construction that “ ‘ “[c]ourts should give meaning to every word of a statute if possible, and should avoid a construction malting any word surplusage.” ’ ” He reasons that because section 17529.5(a)(2) also expressly prohibits “falsified” header information, “misrepresented” header information must encompass something in addition to a false statement of fact; otherwise, the term “misrepresented” adds nothing to the statute and has no meaning.

Instead, Kleffman asserts, in defining the term “misrepresented” for purposes of section 17529.5(a)(2), we should look to “other statutory claims in the false advertising sections of the Business and Professions Code, such as section 17200, which prohibits fraudulent business practices, and section 17500, which prohibits false or misleading advertising.” These statutes, Kleffman argues, apply where advertising “ ‘is not actually false, but thought likely to mislead or deceive, or is in fact false. By their breadth, [they] encompass not only those advertisements which have deceived or misled because they are untrue, but also those which may be accurate on some level, but will nonetheless tend to mislead or deceive.’ ” In Kleffman’s view, this established legal definition is what we should presume the Legislature had in mind when it drafted section 17529.5(a)(2) to prohibit “misrepresented” header information. Alternatively, Kleffman asserts, we should, construe the term “misrepresent” in accordance with its “ordinary” meaning as set forth in several “lay” dictionaries, i.e., to give a “misleading” representation or idea of something.^[6]

Applying these tests, Kleffman argues the e-mails at issue here contained or were accompanied by “misrepresented” header information within the meaning of section 17529.5(a)(2). In Kleffman’s view, the problem is not that the e-mails were sent from multiple domain names, or that the domain names

did not include the word “Vonage.” Indeed, Kleffman expressly disavows any requirement that an advertiser use a single domain name or identify in the domain name the e-mail’s contents, the name of the sending party, or the advertiser’s identity. The problem, Kleffman asserts, is the “random,” “variéd,” “garbled,” and “nonsensical nature” of the multiple domain names, which created the “misleading” or “deceptive” impression—the “misrepresentation”—that they were from different entities when in fact they were all from Vonage “via its solo marketing agent.”^[7] According to Kleffman, the “use of multiple garbled and nonsensical domain names” was “unnecessary” and “serve[d] no purpose but to conceal the single source of these . . . e-mail advertisements” in order “to bypass spam filters,” which “block e-mail when a domain name is associated with the sending of high volumes of spam.” Therefore, “the garbled and nonsensical [nature of the] domain names evidence[d] an intent to bypass spam filters that does not exist with multiple domain names in and of themselves.” In short, Kleffman argues, section 17529.5(a)(2) “permits e-mail advertisements sent from multiple domain names,” but only if “they are not deceptive by virtue of their utterly random and nonsensical nature.”

There are several problems with Kleffman’s analysis. First, his view that we should look to sections 17500 and 17200 fails to account for the significant linguistic differences between those statutes and section 17529.5(a)(2). Unlike section 17529.5(a)(2), neither section 17500 nor section 17200 uses the word “misrepresented” or any form of that word. Rather, as relevant here, section 17500 applies to statements that are “untrue or mislead ing” (italics added), and section 17200 applies to advertising that is “unfair, deceptive, untrue or misleading''’ (italics added). Thus, these statutes expressly use the term—“misleading”—that Kleffman asserts we should read into section 17529.5(a)(2) in order to define the term “misrepresented.” This approach contravenes the principle that “[w]hen the Legislature uses materially different language in statutory provisions addressing the same subject or related subjects, the normal inference is that the Legislature intended a difference in meaning. [Citation.]” (People v. Trevino (2001) 26 Cal.4th 237 , 242 [ 109 Cal.Rptr.2d 567 , 27 P.3d 283 ].)

Second, Kleffman’s approach, including his reliance on dictionaries that use the word “mislead” to define the word “misrepresent,” overlooks the language of the provision that immediately follows section 17529.5(a)(2). Section 17529.5, subdivision (a)(3), prohibits the sending of an e-mail

advertisement with a subject line that is “likely to mislead a recipient.” (Italics added.) Thus, in the very next provision of the same statute, the Legislature expressly used the “likely to mislead” language Kleffman would use to define the word “misrepresented” in section 17529.5(a)(2). This approach contravenes the principle that “when different words are used in contemporaneously enacted, adjoining subdivisions of a statute, the inference is compelling that a difference in meaning was intended.” (People v. Jones (1988) 46 Cal.3d 585 , 596 [ 250 Cal.Rptr. 635 , 758 P.2d 1165 ], italics omitted; see also Briggs v. Eden Council for Hope & Opportunity (1999) 19 Cal.4th 1106 , 1117 [ 81 Cal.Rptr.2d 471 , 969 P.2d 564 ] [“[w]here different words or phrases are used in the same connection in different parts of a statute, it is presumed the Legislature intended a different meaning”].) It also ignores the principle that “when a statute contains a list or catalogue of items, a court should determine the meaning of each by reference to the others, giving preference to an interpretation that uniformly treats items similar in nature and scope.” (Moore v. California State Bd. of Accountancy (1992) 2 Cal.4th 999 , 1011-1012 [ 9 Cal.Rptr.2d 358 , 831 P.2d 798 ].) Under this principle, the meaning of the word “misrepresented” in section 17529.5(a)(2) takes color from the other words listed in the same provision—“falsified” and “forged”—not from the distinctly different “likely to mislead” language found in the next provision, section 17529.5, subdivision (a)(3).

Moreover, it is significant that the language in section 17529.5, subdivision (a)(3), fully articulating the standard applicable to e-mail subject lines— “likely to mislead a recipient, acting reasonably under the circumstances”—is virtually identical to the language that, only months before section 17529.5’s passage, a California appellate court announced for applying sections 17500 and 17200. (See Lavie v. Procter & Gamble Co. (2003) 105 Cal.App.4th 496 , 512 [ 129 Cal.Rptr.2d 486 ] (Lavie) [standard for claims under §§ 17200 and 17500 is whether “the ordinary consumer acting reasonably under the circumstances” is likely to be deceived or misled].)^[8] Thus, the Legislature clearly knew how to draft language invoking the “likely to mislead” standard of sections 17500 and 17200. That it did so in drafting section 17529.5, subdivision (a)(3), but not in drafting the immediately preceding subdivision—section 17529.5(a)(2)—significantly undermines Kleffman’s argument.

Nevertheless, Kleffman argues that the relevant legislative history supports his construction of section 17529.5(a)(2). He acknowledges that, other than repeat section 17529.5(a)(2)’s language verbatim, the legislative analyses of

certain falsified, misrepresented, obscured, or misleading information.” (Sen. Bill No. 1457 (2003-2004 Reg. Sess.) as introduced Feb. 19, 2004.) In the next version of the bill, this statement was revised to explain that existing law prohibited the sending of an e-mail advertisement that “contains or is accompanied by . . . falsified, misrepresented, obscured, or forged header information, or if the e-mail has a misleading subject line.” (Assem. Amend, to Sen. Bill No. 1457 (2003-2004 Reg. Sess.) June 9, 2004, italics omitted.) This revised language expressly recognized the linguistic differences between subdivision (a)(2) and (3) of section 17529.5. Thus, the legislative history of the 2004 amendment does not support Kleffman’s view that “misrepresented” in section 17529.5(a)(2) means “misleading” or “likely to mislead.”^[9]

More broadly, for several reasons, we cannot reasonably interpret the statute as making it unlawful to use the multiple domain names at issue in this case. First, it seems evident the Legislature did not intend section 17529.5(a)(2) generally to prohibit the use of multiple domain names. At the same time it enacted that section, the Legislature addressed the subject of multiple domain names by passing another section—section 17529.4, subdivision (c)—that prohibits the “use [of] scripts or other automated means to register for multiple electronic mail accounts from which to” send, or enable another to send, an unsolicited commercial e-mail advertisement. If mere use of multiple domain names, which requires registration of multiple electronic accounts, constituted “misrepresented . .. header information” for purposes of section 17529.5(a)(2), then section 17529.4, subdivision (c), would be essentially useless. Of course, in construing section 17529.5(a)(2), we must avoid interpretations that would render related provisions unnecessary or redundant. (People v. Fierro (1991) 1 Cal.4th 173 , 262 [3 CalH.ptr.2d 426, 821 P.2d 1302 ].) Notably, as explained above, Kleffman concedes that mere use of multiple domain names does not “in and of itself’ violate section 17529.5(a)(2).

Second, it also seems evident the Legislature did not intend section 17529.5(a)(2) to make it unlawful to use in a single e-mail a domain name that does not make clear the identity of either the sender or the merchant-advertiser on whose behalf the e-mail advertisement is sent. To begin with, a domain name in a single e-mail that does not identify the sender, the merchant-advertiser, or any other person or entity simply does not make any

“representation” regarding the e-mail’s source, either express or implied, within the common understanding of that term, so it cannot be said to constitute “misrepresented” information within the meaning of section 17529.5(a)(2).^[10] Moreover, a contrary conclusion would raise significant preemption problems. Federal law provides that the CAN-SPAM Act “supersedes any statute, regulation, or rule of a State or political subdivision of a State that expressly regulates the use of electronic mail to send commercial messages, except to the extent that any such statute, regulation, or rule prohibits falsity or deception in any portion of a commercial electronic mail message or information attached thereto.” ( 15 U.S.C. § 7707 (b)(1).) Regarding the scope of this provision, a congressional committee report stated that “a State law requiring some or all commercial e-mail to carry specific types of labels, or to follow a certain format or contain specified content, would be preempted.” (Sen.Rep. No. 108-102, 1st Sess., p. 21 (2003), reprinted in 2004 U.S. Code Cong. & Admin. News, p. 2365.) Relying on this statement, the Ninth Circuit has held that a state law requiring an e-mail’s “from” field to include the name of the person or entity who actually sent the e-mail or who hired the sender constitutes “a content or labeling requirement” that “is clearly subject to preemption. [Citations.]” (Gordon v. Virtumundo, Inc. (9th Cir. 2009) 575 F.3d 1040 , 1064 (Gordon).) Thus, construing section 17529.5(a)(2) as requiring this kind of information would contravene the rule that courts should, if reasonably possible, construe a statute “in a manner that avoids any doubt about its [constitutional] validity. [Citations.]” (Association for Retarded Citizens v. Department of Developmental Services (1985) 38 Cal.3d 384 , 394 [ 211 Cal.Rptr. 758 , 696 P.2d 150 ]; see also County of Los Angeles v. Superior Court (1999) 21 Cal.4th 292 , 298 [ 87 Cal.Rptr.2d 441 , 981 P.2d 68 ] [“supremacy clause of the federal Constitution . . . prohibits a state court from applying state law that is inconsistent with federal law”].) Reinforcing application of this general principle in this case is the legislative history of the 2004 amendment to section 17529.5(a)(2), which explains that a purpose of the amendment was to “conform” the statute to the CAN-SPAM Act. (Assem. Com. on Judiciary, Analysis of Sen. Bill No. 1457 (2003-2004 Reg. Sess.) as amended Aug. 5, 2004, p. 1.) Again, as explained above, Kleffman agrees that section 17529.5(a)(2) does not require a domain name in a single e-mail to include such identifying information.

Given these conclusions, we find that a single e-mail with an accurate and traceable domain name neither contains nor is accompanied by “misrepresented . . . header information” within the meaning of section

17529.5(a)(2) merely because its domain name is, according to Kleffman, “random,” “varied,” “garbled,” and “nonsensical” when viewed in conjunction with domain names used in other e-mails.^[11] An e-mail with an accurate and traceable domain name makes no affirmative representation or statement of fact that is false. It is true that the term “misrepresent” may encompass situations where someone, having undertaken to provide information regarding a matter, fails to disclose all facts that “ ‘materially qualify’ the limited facts disclosed. [Citations.]” (Randi W. v. Muroc Joint Unified School Dist. (1997) 14 Cal.4th 1066 , 1082 [ 60 Cal.Rptr.2d 263 , 929 P.2d 582 ].) However, this principle does not apply here. Contrary to Kleffman’s assertion, as a matter of law, the use of an accurate and traceable domain name in an e-mail cannot reasonably be understood to be an implied assertion that the source of that e-mail is different from the source of another e-mail containing a different domain name. This is especially true in this case, given that the header information in each of the 11 e-mails Kleffman allegedly received contained the term “GreatCallRates” in the part of the sender’s e-mail address that preceded the domain name.^[12] Therefore, the sender’s failure to include additional information did not render the header information “misrepresented.” And, absent a misrepresentation, use of a given domain name cannot constitute “misrepresented . . . header information” within the meaning of section 17529.5(a)(2), even if the sender chose the domain name for the purpose of bypassing spam filters.^[13]

Moreover, as a practical matter, the rule Kleffman would have us adopt— that using multiple domain names violates section 17529.5(a)(2) if those

names are “random,” “varied,” “garbled,” and “nonsensical”—is unworkable. Kleffman offers no definition of these terms and no standard for applying them.^[14] As amici curiae supporting Vonage assert,^[15] “[determining whether a domain name is sensible, nonsensical, random, or non-random is so subjective as to make the inquiry meaningless.” “Advertisers, distributors, and even casual e-mail senders (of ‘commercial e-mail advertisements,’ as broadly defined by Section 17529.1, subdivision (c)) would face tremendous uncertainty about whether their actions run afoul of this undefined standard.”^[16] We would add that the uncertainty inherent in Kleffman’s construction is especially problematic given that a violation of section 17529.5(a)(2) constitutes a misdemeanor that is punishable by imprisonment in county jail for up to six months. (§ 17529.5, subd. (c); see People v. Avery (2002) 27 Cal.4th 49 , 58 [ 115 Cal.Rptr.2d 403 , 38 R3d 1 ] [where reasonable constructions of statute prescribing criminal penalties “ ‘stand in relative equipoise,’ ” courts generally adopt construction more favorable to offender].)

Kleffman insists his construction is consistent with the relevant legislative history. Although acknowledging that the legislative analyses of section 17529.5(a)(2)’s enacting bill did not mention the types of header information that would violate the statute, he cites a letter written by the legislative author of both the enacting bill and its 2004 amendment, which stated: “Examples of violations of [section 17529.5] could include,” among other things, “[t]he use of multiple email addresses and/or domain names created for the sole purpose of bypassing spam-filters and blacklists.” However, this statement is entitled to no weight, because we do not consider statements of a bill’s author (or any other legislator) unless they reiterate legislative discussion and events leading up to the bill’s passage (Martin v. Szeto (2004) 32 Cal.4th 445 , 450-451 [ 9 Cal.Rptr.3d 687 , 84 P.3d 374 ]), and

Kleffman concedes there is no evidence the statement he cites meets this requirement.^[17]

Kleffman also relies on references in various sources to the use and effectiveness of spam filters. He first notes that the Legislature’s statutory list of “problems” that made section 17529.5(a)(2) and other antispam legislation “necessary” (§ 17529, subd. (m)) included the following: “[s]pam filters have not proven effective” (id., subd. (f)) and “[m]any spammers have become so adept at masking their tracks that they are rarely found, and are so technologically sophisticated that they can adjust their systems to counter special filters and other barriers against spam and can even electronically commandeer unprotected computers, turning them into spam-launching weapons of mass production” (id., subd. (i)). He next notes a similar statement in an enrolled bill report to the Governor,^[18] and cites arguments against the bill that “[i]t would be better to rely on technology to solve the problem of spam.” (Sen. Republican Floor Commentaries, Sen. Bill No. 186 (2003-2004 Reg. Sess.) Sept. 10, 2003, p. 5; see also Dept. Consumer Affairs, Enrolled Bill Rep. on Sen. Bill No. 186 (2003-2004 Reg. Sess.) Sept. 22, 2003, p. 9 [“it could be argued” this bill is “unnecessary” because “[tjhere are software programs available to consumers that filter unsolicited e-mail”].) Based on these references, Kleffman asserts we should construe section 17529.5(a)(2) to prohibit e-mail advertisements “from multiple random and nonsensical domain names intended to bypass spam filters.”

Kleffman’s argument is unpersuasive. It is true that, in passing section 17529.5(a)(2), the Legislature generally noted the limitations of spam filters. However, this circumstance does not justify contorting the meaning of “misrepresented . . . header information” in the statute to prohibit every practice that might decrease the effectiveness of spam filters. Because, as explained above, Kleffman’s proposed construction is inconsistent with the statutory language read in context and would be unworkable in practice, we decline to adopt it.