The Volokh Conspiracy

Over the last few days, I’ve posted several posts about liberal vs. conservative attitudes on a range of free speech questions, including “hate speech,” anti-religious speech, and more, both based on questions from the General Social Survey and based on an analysis of Supreme Court Justices’ decisions on these questions (whether from the 1970s to now or the 1990s to now). All the results show a similar pattern: (1) liberals tend to be somewhat more sympathetic to free speech protection, including for racist speech, than are conservatives or moderates, (2) the difference is fairly slight, and (3) all three groups are divided on the subject, often sharply divided.

As readers might have gathered, these posts have been triggered by the common argument that I hear from conservatives — for instance, in the comments to these very posts — about how liberals are supposedly eager to support restrictions on “hate speech” and on anti-Muslim blasphemy. Being someone who generally leans more conservative than liberal (and definitely more Republican than Democrat), I would have liked to endorse this argument, and see liberals as culpable in the occasional moves to impose such restrictions. But I think the argument is misguided, for four closely related reasons.

1. It’s factually unsound. Liberals are generally no more likely than conservatives to support restrictions on “hate speech,” and are probably somewhat less likely to do so (though the apparent distance between the groups generally isn’t that great, and is small enough that it’s possible that true distance is virtually nil).

It’s true that some “hate speech” restrictions, such as campus speech codes and even more aggressive restrictions, have been mostly proposed by liberals (in this instance meaning generally people who are left of center). But some of the leading opponents of such restrictions, such as Nadine Strossen and Harvey Silverglate, have likewise been liberals. And while I’ve seen no surveys that break down liberal vs. conservative public opinion on speech codes as such, the general support among many conservatives for restrictions on, say, speech that argues that blacks are genetically inferior suggests that many conservatives among the public would support speech codes as well.

2. It’s unfair. Liberals and conservatives, according to the GSS, are both pretty sharply split on such questions; even if a “hate speech” restriction is proposed by some liberals, it seems very likely to be opposed by a substantial share of liberals, perhaps even a majority.

I understand that our mental images of either group may be a lot more homogeneous — but I suspect that this is a reflection of our own limited experience and limited memories. I doubt that many of us have discussed these questions in details with large groups of liberals and conservatives, much less representative groups. We might remember a few prominent speakers on one or another side who have taken one or another view. But that doesn’t mean much about what “liberals” generally think or what “conservatives” generally think.

3. It’s ungrateful. We owe a great deal of our free speech protections — including protections against university speech codes, protections for alleged “hate speech,” protections for anti-Muslim speech, and so on — to many people, including many liberals, both past (such as Justices Brennan, Marshall, and Douglas, and many liberal lawyers and activists) and present or only very recently past (Justices Ginsburg and Souter, and many current liberal lawyers and activists). Even if we disagree with these people on many things, we should appreciate and respect their contributions on this issue, on which we are allied. And condemning the whole liberal movement, of which they have been integral parts, as supposedly being on the wrong side of this issue strikes me as a poor way of showing such appreciation.

4. It’s politically unwise. Any continued protection against campus speech code, “hate speech” bans, and bans on speech that offends religious groups generally and Muslims in particular can’t come from conservatives and libertarians alone, especially given the number of conservatives who don’t support protecting such speech. It can only come from a continuation of the coalition that has protected such speech in the past — a coalition that has contained many liberals, moderates, conservatives, and libertarians. Unfairly condemning liberals on this point may well alienate some of them from this coalition. And suggesting that supporting such speech restrictions is indeed the liberal way (even when that suggestion comes in the form of condemnation) might in some measure become a self-fulfilling prophecy.

I want the law to keep protecting speech, including what some label as “hate speech”; I oppose campus speech codes; I oppose restrictions on religiously offensive speech. But I think the best way of accomplishing this — both practically best and most reflective of the truth — is by viewing liberals as potential allies (and, in many instances, actual allies, past and present) and not as adversaries.

As I mentioned before, the General Social Survey has long asked respondents the following questions (among many others):

76. There are always some people whose ideas are considered bad or dangerous by other people. For instance, somebody who is against churches and religion... a. If such a person wanted to make a speech in your (city/town/community) against churches and religion, should he be allowed to speak, or not? ...

78. Or consider a person who believes that Blacks are genetically inferior. a. If such a person wanted to make a speech in your community claiming that Blacks are inferior, should he be allowed to speak, or not?

Two quite different questions, it seems to me, with the first referring to an attitude with which people on the left are generally somewhat more likely to sympathize than are people on the right (though in America, both people on the left and on the right are likely to be religious), and the second referring to an attitude that is generally associated with extremists on the right. Yet looking at the results in the 2000-2010 surveys, 76% of all respondents gave the same answer to both questions: Of those who said that the anti-religious person should be allowed to speak, 75% said also that the person who claims Blacks are inferior should be allowed to speak; of those who said that the anti-religious person shouldn’t be allowed to speak, 83% said also that the person who claims Blacks are inferior shouldn’t be allowed to speak.

Likewise, if one looked at people’s answers to (1) whether the person who believes blacks are inferior should be allowed to speak, and (2) whether an “admitted Communist” should be allowed to speak, 77% of all respondents gave the same answer to both questions. So there does seem to be something of a pro-protection vs. anti-protection orientation out there, and not just a whose-ox-is-being-gored orientation.

And one mechanism through which this orientation is spread, it seems to me, is education. For instance, here is the data mapping attitudes on whether the blacks-are-inferior speaker should be protected to the respondent’s highest degree; I have limited this to the years between 2000 and 2010, and only to whites, to minimize confounding effects:

The junior college and graduate numbers are small enough that the margin of error gets a bit larger than I like, but excluding the year limitation yields similar results (48.9% / 62.7% / 66.8% / 76.5% / 80.3%) in the “Allowed” category, with the differences in each category being statistically significant at the 95% confidence level.

I blogged yesterday about liberal/conservative divides among General Social Survey respondents on free speech questions: For anti-religious speech, speech arguing that blacks are genetically inferior, and speech by anti-American Muslim clergy, liberals were somewhat more likely than conservatives to conclude that the speech should be protected, though the gulf wasn’t great, and both liberals and conservatives were divided within themselves.

I’d like to finish by reporting on the answers to three other questions that were asked between 2000 and 2010, and that yielded 8000 responses, split roughly evenly among liberals, moderates, and conservatives:

1. “Now, I should like to ask you some questions about a man who admits he is a Communist: a. Suppose this admitted Communist wanted to make a speech in your community. Should he be allowed to speak, or not?” Liberals said yes by 77-23%, moderates by 64-36%, conservatives by 67-33%.

2. “And what about a man who admits that he is a homosexual? a. Suppose this admitted homosexual wanted to make a speech in your community. Should he be allowed to speak, or not?” Liberals said yes by 90-10%, moderates by 85-15%, conservatives by 81-19%.

Now this might be explicable by attitudes about the particular topic, and not about free speech generally. As to Communists, people might be more open to protection for the extremists on their own side (even if they don’t agree with those extremists) than for extremists on the other side. And of course liberals are less likely to disapprove of homosexuality — and thus disagree with the hypothetical speech by the homosexual speaker — than are conservatives.

But here’s one more question:

3. “Consider a person who advocates doing away with elections and letting the military run the country. a. If such a person wanted to make a speech in your community, should he be allowed to speak, or not?” Liberals said yes by 74-26%, moderates by 65-35%, conservatives by 66-34%.

The pattern is the same: Liberals are somewhat more likely than conservatives to support protection for the speech, though the gulf isn’t very wide, and there are divides within each group. And this is so even though there’s no reason to think the liberals are likely to sympathize more with the pro-military-dictatorship message, or even view as “my extremists, who should be free to speak even if I don’t agree with them.” I don’t think that either conservatives or liberals would view these people as their sort of extremists; though conservatives seem to have a higher opinion of the military than liberals do (also judging by a GSS question), I don’t sense any affinity either on the Left or on the Right for military dictatorship.

Some comments on some of the earlier posts suggest that the difference in responses might stem from conservatives’ being more likely to answer “no” to “should he be allowed to speak” when they would contemplate only social or economic pressure (i.e., private property owners should refuse to rent him space to speak). But I’m skeptical that this is the likely answer. Many speakers, after all, would speak in traditional public forums such as parks and sidewalks; no private economic pressure can stop that. More broadly, I suspect that “allowed to speak” in such a question — which doesn’t specifically refer to any particular private venue, such as a private church or private university or even privately owned meeting hall — would likely refer to “allowed by the government” (as opposed to prohibited, whether by a specific statute or by discretionary government action).

I blogged yesterday about whether liberal or conservative Americans are more likely to support protections for various kinds of speech, including speech arguing that blacks are genetically inferior — the General Social Survey question that’s most relevant to the debate about protection for supposed “hate speech.” It turns out that liberals are somewhat more likely than conservatives to support protection for such speech, though the gulf isn’t wide, and there’s a substantial split of opinion on both sides.

What about Supreme Court Justices? Since 1970, there have been several cases in which the Court has considered restrictions on what might be said to be “hate speech,” usually racist speech but in one instance misogynistic pornography. As I’ll note below, there are limits to how much this dataset tells us, but I pass it along for whatever it’s worth. So here is the data, with the caveats later. Each vote is classified as “+” if it supported protection for racist or misogynistic speech (or hinted substantially in that direction), “-” if it opposed such protection (or hinted substantially in that direction, and blank if the Justice didn’t express an opinion on the subject.

The bottom line result seems to be much the same as what I described for the public at large — liberals are somewhat more likely than conservatives to support protection for what is sometimes labeled “hate speech,” though the gulf isn’t vast, and there’s a substantial split of opinion on both sides. The same also remains even if we exclude Hudnut (which had to do with sex, and which might be hard to categorize for reasons I mention below) and even if we exclude both Hudnut and Collin (perhaps in trying to limit ourselves to post-1990 cases, to reflect the possibility that both conservative and liberal views on free speech have shifted in recent decades). Indeed, even excluding Hudnut and Collin, the result in terms of total votes shows 75% pro-hate-speech-protection for the liberals, 50% for the moderates, and 67% for the conservatives. If you rearrange some of the categories, the results might change slightly; but on balance, I think this probably captures the situation pretty well.

Now on to the details. First, a few general caveats that might limit the relevance of these cases to “hate speech” debates more broadly: The cases don’t generally deal with any solid attempts to outlaw bigoted advocacy generally, and often focus on very narrow restrictions, such as bans on racist “fighting words” (a subcategory of “fighting words,” which is already a recognized First Amendment exception). The cases may have had other factors in play that influence the Justices, such as broader debates about the death penalty (Dawson v. Delaware) or pornography (Hudnut v. American Booksellers Ass’n). Some of the decisions involved denials of discretionary review, so most of the Justices expressed no opinion on the merits (since the refusal to hear a case is often based on factors other than agreement or disagreement with the decision below). Some of the decisions, especially the denials of discretionary review, are somewhat ambiguous in their positions. No Justice was on the Court for all these cases.

Second, an explanation of my ideological labels for the Justices: As with my earlier posts, I combine liberals and moderate liberals into one group, and conservatives and moderate conservatives into one group. I label White and Powell as moderate because that’s how they have generally been understood as Court-watchers, and I think correctly so, if one looks at their views across a wide range of issues. I label Blackmun a moderate because he was generally seen as a conservative in the 1970s but then a liberal in the late 1980s and 1990s; given the makeup of the cases, that averages out to moderate. I label Stevens a liberal because he was generally seen as a moderate in the late 1970s and early 1980s but then a liberal in the 1990s and 2000s; given the makeup of the cases, that averages out to liberal. (I think Stevens genuinely changed his views on some topics, though not necessarily on free speech; I don’t think his labeling as liberal later in his career can be explained solely by the Court’s becoming more conservative.) I label Souter a liberal because I think he was a moderate liberal throughout his tenure on the Court; though he was seen as a conservative when he was first appointed by President Bush, I don’t think he really changed in any appreciable way, and he was quickly recognized to be a liberal. I label O’Connor and Kennedy as conservative because I think that they are on balance moderate conservatives, even though they’ve at times disappointed conservatives with their votes.

Third, a brief summary of the cases:

1. Smith v. Collin (1978), denying discretionary review, and also the earlier decision in the case denying a stay. These came from the famous litigation over the Nazi parade in Skokie, Illinois, in which the lower court upheld the Nazis’ right to march. Justices White and Rehnquist voted to deny the stay, and Justices White and Blackmun voted to rehear the case; both opinions seemed to express some openness to the continued survival of Beauharnais v. Illinois (1952), a decision that upheld bans on defamation of racial and religious groups (an early form of “hate speech” ban). Beauharnais is widely considered to no longer be good law, given New York Times Co. v. Sullivan and later cases, but I read Justices White’s, Rehnquist’s, and Blackmun’s votes and accompanying opinions as at least suggesting the view that Beauharnais should be still seen as valid.

2. Hudnut v. American Booksellers Ass’n (1986), affirming without opinion the Seventh Circuit decision striking down the Indianapolis anti-pornography ordinance (drafted in large part by Catharine MacKinnon), which banned even pornography that fell outside the existing “obscenity” exception, so long as that pornography presented women “as sexual objects who enjoy pain or humiliation,” “as sexual objects for domination,” “in scenarios of degradation,” “as sexual objects who experience sexual pleasure in being raped,” and so on. The Court’s affirmance was a decision on the merits, but one that set a very narrow precedent. (Such “summary affirmances” were not uncommon until the Court’s jurisdictional statute was changed in the late 1980s.) Chief Justice Burger and Justices Rehnquist and O’Connor voted to hear oral argument in the case, which I read as suggesting that the ordinance might well be constitutional.

3. Dawson v. Delaware (1992) reversed a death penalty determination that was based partly on the defendant’s membership in the Aryan Brotherhood prison gang; the Court held that the membership by itself only showed the “abstract beliefs” of the gang, and wasn’t sufficiently linked at trial to the culpability or future dangerousness of the defendant. Justice Thomas dissented, arguing that such a link was present.

4. R.A.V. v. City of St. Paul (1992) struck down an ordinance that specially punished certain kinds of insulting words that were based on race, religion, and similar categories. Chief Justice Rehnquist and Justices Scalia, Kennedy, Thomas, and Souter held that the ordinance was unconstitutional even if it was limited to so-called “fighting words,” which could generally be banned; selectively banning bigoted fighting words, the majority held, violated the First Amendment. The concurring Justices, White, Blackmun, Stevens, and O’Connor, concluded that selectively banning bigoted fighting words would be constitutional, though it voted to strike down the ordinance because the concurring Justices didn’t think it was limited to fighting words. Probably the best way of viewing the split was five Justices being more open to protecting racist speech, and four Justices being more open to some quite modest restrictions on such speech.

5. Avis Rent a Car System, Inc. v. Aguilar (2000) declined to review a lower court case that upheld an injunction against racist epithets in the workplace. As with Smith v. Collin, this decision didn’t express an opinion on the merits. Still, Justice Thomas did dissent from the denial of review, and strongly suggested that the lower court decision was wrong.

6. In Virginia v. Black (2003), the Supreme Court considered three convictions under a Virginia statute banning cross-burning; it struck down one (which I label “Black A”) by an 8-1 vote, on the grounds that that cross-burning was just a statement of racist political ideas, and upheld two others (which I label “Black B”) by a 6-3 vote, on the grounds that the single cross-burning involved in those two cases was an individually targeted threat that therefore fell within the “true threats” exception. The exact details are complicated, but I think it’s fair to say that Justice Thomas was the sole dissenter (on the more speech-restrictive side) in Black A, and Justices Kennedy, Souter, and Ginsberg were the dissenters (on the more speech-protective side) in Black B.

I do not include in this list Forsyth County v. Nationalist Movement (1992), in which the speakers happened to be racist, but in which the law did not target racist speakers as such, but rather imposed extra fees on speakers that were likely to draw a hostile audience.

Also from the General Social Survey, though this time with a question that was only asked in 2008 and 2010, yielding a total of 2500 respondents:

“Now consider a Muslim clergyman who preaches hatred of the United States. If such a person wanted to make a speech in your community preaching hatred of the United States, should he be allowed to speak, or not?”

• Liberal — 52% yes, 48% no.
• Moderate — 38% yes, 62% no.
• Conservative — 39% yes, 61% no.

Here, the difference is considerably greater than as to anti-religious speech and speech claiming blacks are genetically inferior — discussed here — though it’s still in the same direction of liberals being more in favor of speech protection.

UPDATE: I originally accidentally typed “52% yes, 38% no” for liberals; I meant to say, “52% yes, 48% no.”

Here’s data from the General Social Survey, limiting the years to 2000, 2002, 2004, 2006, 2008, and 2010 (with nearly 8000 total responses for each question), with respondents self-identifying as liberals, moderates, and conservatives. (I have flattened out the gradations within each category — extremely liberal/conservative, plain liberal/conservative, slightly liberal/conservative — partly because otherwise some of the cell sizes become small enough that the margin of error gets quite large.)

“There are always some people whose ideas are considered bad or dangerous by other people. For instance, somebody who is against churches and religion... a. If such a person wanted to make a speech in your (city/town/community) against churches and religion, should he be allowed to speak, or not?”

• Liberals — 83-17% yes.
• Moderates — 76-24% yes.
• Conservatives — 75-25% yes.

“Or consider a person who believes that Blacks are genetically inferior. a. If such a person wanted to make a speech in your community claiming that Blacks are inferior, should he be allowed to speak, or not?”

• Liberals — 67-33% yes.
• Moderates — 58-42% yes.
• Conservatives — 62-38% yes.

So looking at the public at large, liberals support protection both for racist speech and anti-religious speech more than conservatives do, though the gulf is not wide. Similarly, when the question is whether “such a person [should] be allowed to teach in a college or university,” liberals are likewise somewhat more likely to say yes, both as to the anti-religious person (72-38% liberal, 60-40% moderate, 57-43% conservative) and the person who believes blacks are genetically inferior (53-47% liberal, 47-53% moderate, 47-53% conservative).