by Paul Alan Levy
We have previously blogged about the crucial role played by section 230 of the Communications Decency Act in protecting the ability of consumers to speak effectively about corporate or political wrongdoing, by protecting the hosts of web sites and email services from being sued over the contents of communications that are made using their services. Absent such protection, Internet hosts would have to worry about being held liable each time the target of speech complained about posted content, and many hosts would take the communications down rather than run the risk of having to defend a lawsuit or even the risk of being held liable. Indeed, just before the new libel tourism bill was passed with specific protection against enforcement of judgments that defy section 230, the bill's sponsor expressly acknowledged on the House floor this fundamental aspect of section 230's purpose. Instead of subjecting Internet hosts to liability, Congress decided to put them in a litigation-free zone that enables them to make judgments of what speech to leave alone and what to take down. Instead, the law says, if someone is disgruntled about speech, the remedy has to be against the speaker and not the medium.
But what happens after the plaintiff wins that case against the speaker? Must the plaintiff accept the fact that the content remains online? This week, two courts addressed that question, and although the cases are distinguishable on their factual and procedural contexts, the results point in different directions. Not surprisingly, both cases involve postings on Ripoff Report.
This is a longer discussion post than I usually post here, but the issue is important. Bear with me.
Seventh Circuit Ruling in Blockowicz
In the Seventh Circuit's decision in Blockowicz v. Williams, 2010 WL 5262726, section 230 provides the background rather than figuring in the decision itself. The case involved some nasty, personal, sex-related accusations directed against some members of an Illinois family, posted anonymously on Ripoff Report. The plaintiffs sued two individuals — whom plaintiffs identified as the former husband of one member of the family (who was not, however, a plaintiff) and his current female companion — and presented some evidence that could certainly tend to link those individuals with the posts. However, after plaintiffs failed to serve either of these individuals, the court ruled that its service efforts had been sufficient, and eventually issued a default judgment against them, finding (based on plaintiff’s undisputed affidavits) that the statements about plaintiffs were false, and compelling plaintiffs to secure the removal of the derogatory material. In addition to enjoining specific statements that the Court found were false, the court issued a general injunction against posting “any false and/or defamatory statements or ‘reports’” about the plaintiffs.
That ruling set up the conflict that produced this week’s Seventh Circuit decision. Having failed to bring the individual posters before the court, plaintiffs decided that the only way they could get effective relief would be to enforce the injunction against Ripoff Report (more precisely, against XCentric Ventures, the company that owns Ripoff Report). But, apparently recognizing that they could not sue Ripoff Report directly, because Section 230 provided it with immunity, plaintiffs argued that Ripoff Report could be subjected to the injunction against the defendants on the theory that it was a third party that was acting in concert with the third party’s violation of the injunction. The district court, and then the Seventh Circuit, rejected this argument. Both courts held that neither the act of hosting expression that is later enjoined, nor the refusal to take any further action to facilitate removal of that content after the posters lose in court, were acts "in concert with” the injunction target, or constituted “aiding and abetting” the injunction target’s violation of the injunction. Indeed, paradoxically, because the injunction target cannot remove the content, there was no violation to aid and abet. Although the significance of section 230 was briefed on appeal, the Seventh Circuit did not even mention the issue; the district court noted that Section 230 prevented the Blockowiczes from suing Ripoff Report in the first place in explaining why the Blockowiczes had sued only the posters themselves. Both courts’ Rule 65 analysis, and district court's passing reference to 230 as barring suit against the host of tortious expression, seem to me plainly correct.
Florida Trial Court Ruling in Giordano
The second decision issued this week is Giordano v. Romeo, No. 09-68539-CA-25 (Fla. Cir. Miami-Dade Cy. Dec. 28 2010). That case arose out of a single anonymous posting on Ripoff Report NOTE UPDATE BELOW complaining about sanitary conditions, bad staff, poor treatment, and the alleged criminal past of one of the owners of the “G&G Holistics Addiction Treatment Center.” G&G and its president, Giordano, sued both Donna Romeo, the poster, and XCentric, the host; XCentric moved to dismiss, raising Section 230 among other grounds, and the Court agreed with that ground for the dismissal. After the litigation proceeded against Romeo, she eventually agreed to allow an order styled as a TRO to be entered against her, requiring her to make good faith efforts to have the post removed; this consent order suggested that, if Ripoff Report adhered to its announced policy of not taking down reports, even at the request of the poster, the court would “will consider an appropriate motion by Plaintiffs at that time.” Romeo duly asked Ripoff Report to remove the report, and it refused to do so.
Plaintiffs then moved for a TRO against Ripoff Report and, one day after XCentric was served with the motion papers, the trial judge held a hearing on that motion; XCentric’s Arizona lawyer was able to participate only by telephone given the lack of notice. Plaintiffs argued that an injunction was not barred by section 230 because they were not seeking damages and not seeking to hold XCentric liable for the defamation, but only seeking an injunction to implement the consent order with Romeo. The judge then signed plaintiffs’ proposed TRO, which included this pre-printed reasoning: “Xcentric [sic] refusal to comply with this Court’s Order and the demand of the publisher to remove the statements makes XCentric the publisher of the statements.” The Court crossed out the pre-printed language “and therefore liable for damages” and hand-wrote the following: “This is different from determining that they are the publisher solely because of the posting.”
Part of the problem with the reasoning is that it is circular – Xcentric is being ordered to take down the posting because it failed “to comply with [the previous] order,” but that order did not compel XCentric to do anything; thus there was nothing for it not to comply with. And in other respects, the reasoning is just wrong. The order expressly treats XCentric as the “publisher” of the statements in defiance of the statute’s express command that a provider of interactive computer services shall not be "treated" as the publisher of content provided by another; and the courts addressing section 230 have consistently agreed that the decision whether to remove posts is part of the “publisher” function that Congress intended to leave to web hosts’ discretion. Nor does the court explain how it can issue the equitable remedy of an injunction -- especially an injunction forbidding speech -- against a party that has not been held liable for the speech.
Moreover, allowing motions for injunctive relief is inconsistent with the statutory language providing an immunity from suit as well as from liability, and with one of the basic reasons why host immunity is such an essential element of the system of online freedom of speech. Even if the host is protected against damages claims, the very need to respond to motions for injunctive relief imposes the expense of litigation and thus threatens to enable the heckler’s veto that animated Congress to adopt Section 230 in the first place, and to make it a part of American law so fundamental that a foreign defamation judgment cannot be enforced in the United States unless it meets section 230's requirements. And hosts are understandably chary of being enjoined, because even if it is s very specific statement that has already been held to be false that is enjoined, the host has to worry about being held in contempt if somebody later posts that same statement online.
Also curiously not addressed by the Florida court was why it was entitled to issue a prior restraint on less than a full evidentiary hearing on the merits. I discuss below the fact that, in many jurisdictions, the old rule that “equity will not enjoin a libel” remains good law to this day. But even the jurisdictions that allow the entry of a permanent injunction against speech after the defendant has had his day in court and a jury has decided that the enjoined speech is false, consistently hold that a preliminary injunction is a prior restraint. And as the Supreme Court said in Organization for a Better Austin v. Keefe, 402 U.S. 415, 419-20 (1971), in overturning a preliminary injunction against leafleting on First Amendment grounds, “No prior decisions support the claim that the interest of an individual in being free from public criticism of his business practices . . . warrants use of the injunctive power of a court.”
How Internet Hosts Generally Respond to Court Orders Directed at the Original Poster
In our experience, there is wide variation among Internet hosts with respect to the circumstances in which they will do voluntary takedowns.
On some web sites (for example, the owner of a blog on the Google or Typepad blogging platforms), someone who puts material online has physical control of their own content and can remove it at will. But when access for removal remains with the host, the poster can only send a request for removal. Some hosts will just take material down on request from posters, but many hosts will not. They may demand a good reason for the removal, and they may be concerned about the possibility that posters may want to have a complaint removed for some reason other than recognition that what they said was false. Many times, the target of the criticism will threaten suit, and posters may remove the criticism just because they know they can’t afford to defend against such a suit. In those circumstances, maintaining a policy of not removing posts on request can actually be a way of protecting such posters against intimidation, as well as protecting the marketplace of ideas against having valuable contributions removed from the public debate for spurious reasons.
Hosts also vary enormously in their responses to the issuance of a court order to an alleged defamer. For some hosts, an order finding that a specific post is defamatory and ordering its removal will be a sufficient basis to take down the content no matter the circumstances attending the order. But other hosts worry about the fact that it can be too easy for a wealthy individual or company that is the target of critical speech to secure a court order even if the report is true (or is a nonactionable opinion). For example, the defendant may simply consent to a judgment, or fail to oppose its entry, because it is too expensive to defend herself, recognizing that she is essentially judgment proof in any event and hoping that the plaintiff will not pursue further litigation at that point. Thus, just as some hosts will not take down content just because the original poster asks the host to do so, they will not take down the content just because the original poster has consented to a court order requiring its removal.
Hosts also worry that sometimes a plaintiff may deliberately do very little to try to serve the defendants, and then seek a default judgment, using home court advantage to secure a ruling from a judge that want to help a local lawyer or local company against a disgruntled individual, especially if the alleged defamer is a far-away defendant. Or the plaintiff may deliberately file suit in a court that lacks personal jurisdiction or that is, in any event, too far away for the defendant to effectively mount a defense.
Thus, some hosts choose not to remove content that is the subject of a court order against an information content provider unless they are confident that the order was obtained under circumstances that reflect a considered decision that shows that the content really did violate the plaintiff’s rights. They may insist on satisfying themselves that the trial court had jurisdiction, or they may decline to act pursuant to orders obtained by default or by consent; only a case that is litigated on an adversary basis may satisfy such hosts. Or they may apply a variety of other tests of their own devising to determine whether to go along with a court order that is not, after all, binding on them.
Ripoff Report’s Stance on Voluntary Removal of Content – and Its Explanation About Its Refusal to Back Down in Blockowicz
Ripoff Report's "non-removal policy" is well-known -- it takes the position that it will not remove content once it has been posted. Instead, Ripoff Report’s view is that someone who wants to express disagreement with a report should express that disagreement through a comment on the post. As explained on Ripoff Report's home page, "ALL complaints remain public and unedited in order to create a working history on the company or individual in question." And that policy extends to the person who originally posted the report – if she later regrets the posting, and wants to clarify or correct or even retract it, Ripoff Report’s view is that the correction or retraction should be posted as a comment. Even a customer that participates in the very expensive “Corporate Advocacy Program” does not get criticisms removed, although, as I discuss here, the criticisms are buried in praise.
As Blockowicz and Giordano show, Ripoff Report's stance of not taking down posts is maintained even when the target of the criticism has obtained a court order.
Interestingly, Ripoff Report general counsel David Gingras has appended to the posts at issue in the Blockowicz case a lengthy discussion defending the company’s decision to resist removal of the posts instead of "do[ing] the right thing and remov[ing] the statements.” Interestingly, his statement indicates that, although Ripoff Reports has a general policy against removing posts: “Truth be told, on occasion we may remove reports when it’s clearly appropriate to do so.” However, his posting explains at considerable length why he was skeptical of the propriety of removal in this case.
With the kind assistance of the lawyers on both sides of the Blockowicz case, who supplied me with various briefs and other court filings, I have compared Gingras’ explanations with the record in the case. I am not posting all of these papers but am prepared to provide them to readers who want to evaluate my assessment of them.
A number of Gingras' reasons seem dubious to me. For example, Gingras cites the fact that the complaint does not mention the fact that the original postings at issue in the case had been posted in 2003. Gingras’ point is that a suit over a 2003 post would be barred by the statute of limitations. But the only reason omission why from the complaint would matter would be if there were some reason to believe that plaintiffs were trying to hide the dates from the court. And that would be a serious concern in an ex parte proceeding. But the dates are all over the affidavits submitted in support of injunctive relief and of the default judgment. So Gingras’ lengthy discussion of the omission of dates from the complaint just makes it seem that there is a problem that doesn’t really exist.
Gingras makes much of the point that the original posts from 2003 were largely about Megan Blockowicz, who is not a party to the case (and he correctly notes that only Megan Blockowicz can sue over statements that defame her — the “of and concerning” requirement has constitutional dimension in fact). However, review of the record shows that the suit was brought over statements about the three family members who are plaintiffs in the case. He also claims that the plaintiffs never showed that the defendants actually got notice of the lawsuit against them, and never showed why there was reason to believe that Michelle Ramey was a defendant in the case. But my own review of the record shows that there was a fair amount of evidence tending to show these facts. To be sure, at the time plaintiffs sought leave to proceed without service of the summons, there were some indications that plaintiffs had not taken all the steps that I would have required, had I been the judge, to ascertain wether plaintiffs had received actual notice. But this circle may have closed later on, when there another anonymous post, apparently from the unserved defendants, saying that they knew of the lawsuit and proclaiming their defiance.
Gingras also asserts that he examined the server logs connected with the posts in question, and reviewed the email addresses that the posters had given when registering to post, and that none of this information was consistent with the information provided by the plaintiffs as their basis for accusing Williams and Ramey of being the posters. Although this contention is a tantalizing one, it is not specific enough for me to be able to assess the soundness of Gingras’ judgment in that regard. Indeed, given the extent to which some of his arguments seem to grasp at straws to justify skepticism of plaintiffs' claims, I find the generality of this assertion troubling.
Gingras also says that many of the statements in the posts at issue are hyperbole or rhetoric, which the courts generally treat as non-actionable opinion. Reviewing the posts I would have to agree with him; and the fact that the injunction includes a general prohibition against posting any “false and/or defamatory statements” in the future suggests that the order was entered without consideration of the First Amendment. But there are some highly factual statements in the posts, and the injunction does not appear to be based on the hyperbolic statements that could fairly be defended as mere opinions. Moreover, there is evidence in plaintiffs’ affidavits showing that those factual statements are false.
There is another reason set forth in Gingras’ account that resonated with me much more. Gingras was plainly offended that the original complaint accuses Ripoff Report of charging individuals and companies for the privilege of responding to reports, and that the motion to enforce the injunction against Ripoff Report accused it of “aiding and abetting” a violation of the injunction based on “simple greed.” His post concludes:
"Here, because the plaintiffs never bothered to contact us prior to filing their lawsuit, we were never given any opportunity to decide whether to work with them to delete or redact parts of these reports. In addition, by making false statements against us in court which cost tens of thousands of dollars to defend, the plaintiffs’ lawyer made it extremely difficult for us to consider voluntarily agreeing to remove these reports."
It has certainly been my observation over the years that when plaintiffs bring suit against web hosts, forcing them to hire lawyers to litigate the cases, and when plaintiffs accuse the hosts of misconduct, that tends to stiffen the host’s spines and leads them to defend strongly when they might otherwise be inclined to be gracious. And because many of his other arguments seem weak in light of what I saw in the record, I am inclined to think that this last argument may be the best explanation of Ripoff Report's decision not to remove the posts.
Does Ripoff Report’s Standing on Its Rights Threaten Section 230?
Some of the blog coverage of the Blockowicz case has suggested that rulings upholding Ripoff Report’s immunity in such cases pose a long-term threat to section 230, because they take the statute beyond where Congress would have wanted to apply it.
For example, blogging about the district court ruling, Ben Sheffner argued that although the ruling follows from the statutory language, I can't imagine Congress would have enacted Section 230 back in 1996 if it knew this would be the result. On Ars Technica, Nate Anderson wondered whether the district court ruling is “‘safe harbor’ run amok, or just an unfortunate and rare side effect of an otherwise well-crafted statute?” And Eric Goldman, our leading academic authority on Section 230, whose analyses of section 230 have sometimes taken its protections further than even I would go, expressed “normative” concerns with the district court’s ruling; after the Seventh Circuit ruled, Goldman said:
"If UGC websites refuse to voluntarily honor injunctions against user postings and the initial posters can't be found, there could be circumstances where pernicious content simply can't be forced off the web. I would be more troubled by this if I expected this outcome to arise frequently, but it won't [because it happens so seldom]"
I am much less worried about the normative consequences of holding that the web host can’t be enjoined, because there is another significant problem with extending either injunction to Ripoff Report. There is a hoary doctrine holding that “equity will not enjoin a libel” — that no matter how false and hurtful a statement may be, the only proper remedy for defamation is an award of damages, not an injunction. Several states still follow his maxim, and both Illinois and Florida are among them. See Montgomery Ward & Co. v. United Retail, Wholesale & Department Store Employees, 400 Ill. 38, 51, 79 N.E.2d 46, 53 (1948); Moore v. City Dry Cleaners & Laundry, 41 So. 2d 865, 873 (Fla. 1949). My research indicates that the Illinois Supreme Court ruling has never been overruled, although it has been narrowed, so that it does not apply to cases of commercial disparagement among competitors. And there are quite a number of state appellate decisions in Florida reiterating the rule. E.g., Weiss v. Weiss, 5 So. 3d 758, 760 (Fla. Dist. Ct. App. 2009). And regardless of state law on the question, it remains an open question whether the First Amendment treats permanent injunctions against statements finally determined to be false as impermissible prior restraints. A few years ago, the Supreme Court of the United States granted certiorari in Tory v. Cochran to decide that issue, but the case was later decided on narrower ground when Cochran died during the pendency of Supreme Court review.
Gingras does not mention this problem in his posted defense; neither party discussed the issue before the Seventh Circuit; nor was the issue briefed in connection with the TRO proceeding in Florida against Ripoff Report (given the short notice, I doubt this will be treated as a waiver; Ripoff Report had briefed the issue earlier in the proceedings). But given the fact that many states would simply not allow an injunction even after a final determination of falsity in an adversary hearing, and given the fact that the constitutionality of such an injunction remains an open question, it is hard to see why construing Section 230 to leave removal to the web host’s discretion is normatively wrong. It may just be that damages have to be the only remedy, unless the web host exercises its discretion.
Securing Removal Without An Injunction
In that regard, it is worth noting that, as noted above, Gingras’s posting indicates that even Ripoff Report has removed posted material in circumstances it deems appropriate. Moreover, once it won in the Seventh Circuit, Ripoff Report placed this statement at the beginning of each of the posts over which suit was filed:
"although the existing reports remain visible in their original form, we have made minor redactions to the titles of the affected reports to remove language that was needlessly offensive and profane. Furthermore, despite our decision not to remove this text, anyone reading these reports should keep in mind that a court order has been entered which finds the statements below are not true."
A reader has to scroll down through a long preface about the case reaching the statement just quoted, before reaching the accusations over which suit was filed. In effect, the accusations have now been buried on the page.
It would surely be interesting to know in what other circumstances Ripoff Report has deemed it “appropriate” to remove comments, although at the same time I can also appreciate that Ripoff might not want to publicize those instances for fear of encouraging takedown requests that it is unlikely to honor. But the generosity of this action once Ripoff won on appeal may suggest that it would have considered taken action had the Blockowiczes simply come to Ripoff in the first place, presenting the order against Williams and Ramey as well as evidence of falsity.
And it is certainly true that filing a lawsuit or motion is likely to be quite counterproductive with respect to any host's exercise of its section 230 discretion. I have had any number of conversations with web hosts or their lawyers who told me that their willingness to consider removing content was reduced or eliminated because they were being threatened (or sued) instead of just being asked. They felt, understandably, that if their reputation was to give up in response to threats or suits, they would just be inviting more.
Indeed, as a practice tip, I would generally advise lawyers for plaintiffs like Giordano or the Blockowiczes to recognize the very strong position in which Section 230 puts a web host. Instead of initiating legal action against the host, it is sometimes more effective to go hat in hand, simply laying out a factual case to show how false and how harmful posted accusations are, and asking the hosts to exercise the discretion to take the posts down that Congress gave them through Section 230. I know of a number of cases where that more modest approach has been successful. Although I myself don’t represent plaintiffs in such cases, I have been able to guide plaintiff counsel who approached me for advice down such a successful path.
The Blockowiczes, by contrast, have litigated the case at length and, in the final analysis, they have probably done more to publicize the accusations made against them than to inhibit their future publication. Giordano’s counsel is likely to accomplish the same for his client once that case goes up on appeal. If we assume that the statements about them are both false and malicious — and perhaps there is good reason to think so — that is a pity.
Since this analysis was posted, XCentric has complied with the state court TRO by replacing the entire report (as the injunction required it to do, even though it was only the statement that Giordano is a "convicted fellon" that the court decided was false) with a report on the case, including links to a number of the underlying documents. The link to the post in the discussion above now goes to that discussion.
A new trial judge, who reportedly was assigned to the case after the original judge failed to win re-election, has dissolved the temporary restraining order and dismissed the action against XCentric.