CCR Symposium: What Does the “Civil Rights” Label Add?

Having written on why I’m not sure the problems of this symposium are civil rights problems, let me now move on to what exactly the label adds. Let’s say, for the sake of argument, that everyone agrees that the issues here are civil rights issues. What follows from that?

I ask the question because I don’t see how altering Section 230 liability in response to the problem of cyberbullying is a “civil rights” response. For example, two years ago, I wrote a blog post floating the idea that we should restrict Section 230 liability in response to the problem of Internet cyberbullying, and in particular the facts of the AutoAdmit case. I wrote:

Would it help to somehow link up provider immunity with search robot exclusion? Under current law, site owners are immune from liability for the speech of others under 47 U.S.C. 230. This means that a site owner can allow anonymous comments, announce that anything goes, and then sit back and watch as the trolls engage in all sorts of foul play. Search engine robots then pick up the foul play, resulting in harm weeks or months later when a third party googles that person or event. A lot of people may be harmed, but the law can’t stop it: the provider is immune and the commenters are anonymous.

If I’m not mistaken, though, the same provider who is immune under Section 230 also controls the scope of the resulting harm. Why? Because, at least as I understand it, the same provider controls whether search engine robots are permitted to come to the site and collect the information in the first place. I believe that blocking search engine robots is pretty easy, or at least could be configured to be easy; it just requires a line of htmlcoding.

Where does that take us? Well, it suggests to me that we might consider conditioning legal immunity on disabling search robots. Providers would be immune for liabililty relating to particular content only if they had taken technical measures to block search engine robots from collecting that content. So if you wanted to host a free-for-all for others and be immune from liability, you coulld do that: you would just have to keep the resulting content from being fed into Google. On the other hand, if you wanted Google to pick up the content, for whatever reason, you would need to assume the risk of liability for that content you’re letting Google collect.

I emphasized at the time that the idea was tentative and amateurish, and that I didn’t know if it would work. Indeed, maybe the idea is nutty. But I didn’t conceive of the proposal as a civil rights proposal, or of myself as a civil rights crusader. Rather, I was just trying to figure out if you could tinker with the law in a way that restored incentives lacking in the law that seemed to be causing significant social harm.

So I return to the question, what does the label “civil rights” add? I wonder if it’s mostly a rhetorical move. On one hand, the label appears to genuinely excite some people. On the other hand, folks who might oppose the proposals might be reluctant to be seen as opposing something labeled “civil rights.” Either way, I’m not sure what the label adds in substance. Social harms are social harms, and I’m not sure the label does much to change how we assess them.

You may also like...

26 Responses

  1. Dave says:

    I may be missing something, but the answer to this question seems pretty easy. I read Citron as arguing that categorizing cyber sex harassment as a civil rights issue will have two positive impacts: substantive and expressive.

    One way the civil rights label adds value is by allowing victims of online sex harassment access to a greater range of federal causes of action. I don’t know much about the content of federal civil rights law, but I think it has to be right that this expands the kinds of legal theories of liability, penalties, and resources available to combat this kind of online conduct. This is in the BULR piece at 89-96.

    So I don’t think it’s merely a rhetorical move, but I think it is certainly the latter as well. This is the point Citron elaborates in the Michigan L Rev essay that categorizing cyber gender harassment as a civil rights issue sends a message that the law takes that kind of conduct seriously, and counteracts the perception that it’s trivial (a point I question in my post from yesterday).

    One might disagree that the civil rights categorization will be a positive move, but I think in both of her pieces Citron makes pretty clear what practical implications the move will have.

  2. It’s because you’re not a liberal – the approach was designed to appeal to liberals and civil-libertarians who are strong supporters of free speech as a civil right. So it’s operating within that social framework, to re-assure them that this is compatible with their values of supporting civil rights.

    This was all pretty well analyzed in the 80’s with the Mackinnon-Dworkin proposed laws, and it’s strange to see that approach being rerun with nobody seeming to remember how much we’ve been around it before (by which I mean not that the issue hasn’t been even mentioned, but it’s not novel at all, and much explored, sometimes quite negatively).

  3. I wrote a short essay for the Yale Pocket Part arguing that this robots-exclusion proposal is a bad idea.

    More directly, my second post in this symposium responds to the question of what the “civil rights” label adds. You talk about how reducing the scope of section 230 isn’t a particularly “civil rights” response, but I think the “civil rights” label is helpful in directing us to look at remedies beyond just imposing liability on intermediaries. What would be the online equivalent of an escort from the 101st Airborne, for instance?

  4. Orin Kerr says:


    I did very much appreciate your explanation of why you see this as a civil rights issue. I didn’t find it persuasive, though: Many crimes target people who are weak and vulnerable, and yet the fact that they target the weak and vulnerable are not normally thought to make them civil rights issues. For example, fraud schemes often target the elderly, and yet we do not say fraud is a “civil rights” offense. (Do you think we should?)

  5. Paul Gowder says:

    Orin, there’s a difference between “weak and vulnerable” and “regularly treated badly,” and on that difference hinges the label “civil rights.” — see my five-minutes-ago comment in the other thread for more (and an attempt to unpack the stuff you objected to in my working definition).

    As for the work it does, it identifies the bad behavior at issue as part of a broader pattern of bad behavior, and helps us understand the way that behavior has a particularly harmful and morally objectionable quality.

  6. Orin Kerr says:


    It seems to me that the elderly in our society are “regularly treated badly”: They are ignored, ostracized, shuttled off into “retirement homes,” and made the targets of crimes ranging from frauds and robberies. Why do these grievous injustices not merit the label of civil rights issues?

  7. Paul Gowder says:


    Alas, we seem to have forked this terribly, because my answer to this is my same as my answer to the last point in the discussion in the other thread, so, well, see that.

  8. A.W. says:

    I’ll tell you what it does. it makes us canada, where one of the most basic human rights, freedom of speech, is sacrificed on the altar of “human rights.” Ask Mark Steyn or Ezra Levant.

  9. It adds a lot. Because without that label and the premise that this is a “civil rights” issue, it removes a large swath of potential criticism of the work. Scott Greenfield says it far more eloquently than I can:

    “The gist of Danielle Citron’s theory begins with the proposition that women and minorities are disproportionately victimized by vicious anonymous online attacks and threats of real world harm, which both extends beyond mere bad speech into a metaphysical punch in the face and the inhibition of free expression by the victims due to the excess of free expression by the attackers. This is supported by anecdotal evidence and studies of dubious value, though most of the symposium participants have accepted this part of the proposition without any critical thought, since it’s consistent with their sensibilities and academic orthodoxy. It’s a value judgment, and few in the Academy are inclined to be naysayers whenever women’s victimization is at stake.”

  10. AF says:

    What does the civil rights label add? Concretely, new causes of action.

    For example, Citron argues that cyber-harassers should be held liable under 42 USC 1981 or 1985, which outlaw discriminatory interference with contracts and conspiracy to violate civil rights, respectively. See pp. 89-96 of her paper.

    Whether these statutes (and others mentioned by Citron) apply to cyber-harassers is open to debate, but it is hard to deny that if “everyone agrees that the issues here are civil rights issues” that may open new causes of action under the civil-rights laws.

  11. Howard Wasserman says:

    But it’s more than causes of action, because we do not need the Civil Rights label to get more causes of action. We can get there by saying “Cyber Torts” or “Cyber Federal Torts” or “Cyber Effects on Commerce”–all of those would open the door to new causes of action. No, calling it “civil rights” works to establish a tone about *types* of cause of action and to appeal to visceral sense of “justice.”

  12. Orin Kerr says:


    I appreciate the response, but that’s just a doctrinal claim, right? If it’s a doctrinal claim, then the label doesn’t matter: What matters are the facts and the law. To the extent the debate here is really over whether certain specific doctrines apply, then I think we should be citing cases and text. I would certainly welcome that, and learn from it, but that doesn’t seem to be what either the participants or the commenters are doing.

  13. Orin Kerr says:


    Yes, I think that’s right: My point is that “establishing a tone” and “appealing to visceral senses” seem more like rhetorical questions than anything else.

  14. Orin: It seems to me that the elderly in our society are “regularly treated badly”: They are ignored, ostracized, shuttled off into “retirement homes,” and made the targets of crimes ranging from frauds and robberies. Why do these grievous injustices not merit the label of civil rights issues?

    They do. When the elderly are targeted because of their age, that’s often a civil rights problem. See the Age Discrimination in Employment Act, elder abuse acts, Social Security, etc.

    Next question?

  15. JP says:


    Another commenter, DevinB (who I don’t think has shown up in this symposium), raised a similar question about a month ago in response to a post by Danielle:

    DevinB’s comment is worth reading in full, but I’ll quote my own short follow-up to his remark:

    “I would venture that, all else being equal, framing the issue as a women’s issue, as opposed to a social order issue, will result in greater visibility within the academy (marginally higher likelihood of publication, higher readership, etc…). But as Devin notes, the same framing has the opposite effect when the issue is raised in the policy-making context.”

  16. And, as for fraud, no, it’s not generally a civil rights problem. But it can easily become one. The mortgage scams directed at blacks moving into integrated redlined neighborhoods were a civil rights issue. Recognizing it as one helped give us the Fair Housing Act, which took on the fraud, in part, by targeting discriminatory lending practices elsewhere in the system. “Civil rights” helps direct our attention to patterns of victimization, to targeting on the basis of membership in vulnerable classes, and to systemic responses.

  17. Orin Kerr says:


    Your “next question?” retort suggests your patience with me is running thin, and for that I apologize. But my question to Paul was this: “fraud schemes often target the elderly, and yet we do not say fraud is a civil rights offense. Do you think we should?” I would be very interested in your answer.

    Also, I suppose it’s a very interesting result if you and Paul think these questions are trivially easy and yet disagree with each other on the answers. Indeed, as I read the answers in this thread and the earlier one, all three people who have responded have quite different definitions of what makes a case a civil rights issue.

  18. No, my apologies; the “next question” was meant more along the lines of “Well, that has a crisp answer, but I bet Orin’s next question will be more of a head-scratcher” than as a dismissal. I like to use the phrase in class as a comic moment; everyone takes a quick breath at getting an answer they can rely on, instead of one hard imponderable after another. I shouldn’t have tossed it in here; it came across as far too flippant. I’m enjoying this symposium, even if I can’t give it all the attention it deserves. And yes, I agree with you that I think there are some very different definitions of “civil rights” floating around here; I’ve been trying to point out and emphasize one that I think has real value.

  19. Orin Kerr says:

    James, it seems we cross-posted, and you did indeed offer an answer to the question. Thanks for that. At the same time, your answer seems unclear to me: Your answer seems to be, “”not generally” but “maybe”. Well, why? What’s the test? Is it an empirical one? A doctrinal one? An ideological one?

    You also comment: “civil rights helps direct our attention to patterns of victimization, to targeting on the basis of membership in vulnerable classes, and to systemic responses.” Who is the “us”? What do we do once our attention is directed? And why are we having problems with our attention without this? Or maybe that’s the rhetorical point again — that the label is a powerful rhetorical device that could be used to get people to favor a more severe remedy than otherwise?

  20. Orin Kerr says:

    Oh, and thanks James, re the misunderstood comment.

    About the case of mortgage scams and redlining, my recollection (and I don’t claim to have followed it closely) was that it was a civil rights issue because there was intent to discriminate on the basis of race by specific wrongdoers, thus triggering the elements of existing law. It seems here that we’re dealing with something else — more of a disparate impact concept looking at a class of conduct as a whole.

  21. The easiest answer is that “we” is “me” — Danielle’s first draft of the paper, as I saw it last summer, struck me as wrong on its 230 prescriptions but really opened my eyes to looking beyond intermediary liability. It also gave me a more concise way of thinking about the gendered aspect of online harassment. My understanding about trolls targeting the vulnerable grew directly out of those musings. And that’s helped me realize that there are really two online harassment problems: one involving specific dedicated abusers and the other involving the fury of (often randomly targeted) mobs. Civil rights categories, such as group-based animus, have helped me sort through the facts. That’s something that’s severely undervalued in legal-academic discourse, in comparison with making bold legal proposals, but it’s incredibly important. Even if CCR does no more than help us with the factual description, that’s a real use — it makes clear why some legal responses miss the problem, and why, say, naming-and-shaming (to the extent it’s possible) might be helpful.

    And now, I really have to prep my afternoon class, but ping me again on your first question, which deserves reply, too.

  22. Disparate impact? Looking at a lot of the sort of online harassment that CCR discusses, the invidious discriminatory intent seems obvious on the face of it. We’re talking about the sort of comments that would lead to substantial liability in, say, an employment discrimination lawsuit.

  23. Orin Kerr says:


    I am not well-informed about employment law, but I thought the theory of the workplace discimination claims was that comments were made on the basic of race, color, religion, or sex can create a hostile work environment. As I understand it, the employer doesn’t need intent to discriminate to be liable; rather, the employer needs to have hosted a work environment that was hostile because of comments made on the basis of race, color, religion or sex.

    I suppose part of my confusion is that I’m not sure what level of generality you’re looking at to determine whether there is intent to discriminate: Is that intent of specific commenters, intent of specific hosters of online environments, or a more generalized global intent of those who engage in cyberharassment?

  24. JP says:


    Most employment discrimination laws do require intent. The question of employer liability in hostile environment cases is (somewhat loosely) based on agency principles, so a fair amount depends on context (e.g., whether the individual[s] intentionally creating a hostile work environment is a supervisor, what complaint mechanisms are in place, etc.).

    Disparate impact cases do not require intent, but the comments in CCR are not analogous; they would be considered under the law prohibiting intentional disparate treatment.

    An employer that was aware that its managers or employees made comments similar to those at issue in CCR, but did not respond by disciplining the offenders, would be presumed to have accepted them, and would be liable.

    Mapping employment discrimination law directly to the question of the liability of online hosts would effectively shut down blog comments and discussion boards. Much to her credit, Danielle does not seem to advocate this. However, it isn’t clear how different her proposed “standard of care” would end up, since–aside from requiring “traceable anonymity”–she mostly leaves it to be defined later.

  25. Orin Kerr says:

    Thanks for the info, JP.

  26. AF says:

    I take Howard and Orin’s point that this discussion is about more than doctrine.

    But I would disagree that doctrine is just about “facts and law” and that a offering a new conceptual framework has nothing to do with doctrine. Doctrine sometimes changes when people decide that new types of harm fall under preexisting statutes or causes of action. This is often the result of “reconceptualizing” either the nature of harm, or the nature of the existing law, or both.

    A pertinent example is sexual harassment, which was determined to be a kind of sex discrimination after legal scholarship argued that it should be so conceptualized.

    I had thought Professor Citron had similar goals in her paper.