This post will address the concern that the methodology of the Cato study is biased. And an explanation to the irregularities in Murphy's data (provided by Guitar), and how to view the similarities and differences in all three surveys regarding free speech (Cato, Murphy using GSS, Economist).
Guitar said
Cato study is "bought and paid for by the ultra-conservative Koch brothers is obviously going to have some bias issues" and that "There are easy to see flaws designed into the study. I already addressed the Koch brothers issue many times here, each study should be judged on its own merit, not by who paid for it or did it. You are more than welcome to attack the data, methodology, and the analysis. But if you have to resort to refuting studies using because it has an "ideology tilt". It's really pointless to go any further. To be fair, you raised a methodology problem, lets take a look.
The "flaws designed into the study", such as questions targeting liberals, was indeed
designed in such a way to illustrate "82% Say It’s Hard to Ban Hate Speech
Because People Can’t Agree On What Speech Is Hateful or Offensive". In fact a set of questions targeting conservatives can be found in Appendix B, including "A person who says that all white people are racist", and "A person who says the police are racist" just to drive the point home. And looking at the survey results in Appendix B, the answers match the party line division. This is a
feature of the study to showcase it's hard for people to agree on what is hateful or offensive, not a bug or inherent bias as you said. The claim you made regarding bias is therefore erroneous.
A side musing when I first looked at
Murphy's analysis: I was almost immediately drawn to the similar shapes of "extremely liberal" and "extremely conservative", which reminded me of the
Political Horseshoe.
The Murphy analysis is problematic, not only to me, but also to Kurt, Guitar's source whom commented on the Murphy analysis.
"Well, I can understand the conservatives, because many of them are racists, but the extreme liberals?"
"Again, I’m a bit confused. The people who are the most censorious of racist speech are those who are slightly liberal, yet those are people who, compared to other liberals, are also most censorious of militarist speech (those at about 2.5 on the scale)."
"One note: deplatforming of college speakers, as judging by the FIRE “disinvitation database”, is being
done far more these days by the Left than the Right. That doesn’t comport with Murphy’s finding that the left is the least censorious wing of politics."
It is fortunate then, that Murphy himself provided us with the tools to examine his work: his own
robustness check. Namely, we can sample the 2014 data points from Fig.2 and map it on the same X-scale from fig.3. If Murphy's analysis is robust and what he had observed truly present, both the
trend and
shape of both lines should be at least comparable. I have attached a basic xlsx file, and you can all see that the two lines are nothing alike, especially for the "racist speakers". Worse, when we actually go on to
GSS Explorer and plot the graph ourselves, separated by
political affiliation (under break down), we see the Dems are decidedly less open to let a racist speaker speak, which is not necessarily always a bad thing. But it directly
contradicts Murphy's finding and analysis. Why? Murphy himself gave us an answer.
"One might object that my finding about the “Slight Leftist Speech Suppressor,” which surprisingly exonerates the radical leftists from the speech suppression tendency,
might be a false artificat of the ambiguous meaning of the term “liberal.” Perhaps one could say that survey questions measuring “liberalism” are not accurate measures of leftism because they may be tapping Classical Liberalism (more libertarian than leftist). I used the “liberalism” measure of ideology because it had the most data over time, but we can check our findings against survey questions asking about “leftism,” although they have much less data. "
Murphy used survey questions measuring "liberalism" to make his own analysis, but as we just saw, it indeed suffered from a false artifact of the ambiguous meaning of the "liberal", which resulted in it failing the robustness test and showing the opposite of what the raw data shows.
Despite all this, Murphy acknowledges that there is "a puzzling minority of vaguely leftist activists, who happen to have gained media attention, wish to suppress free speech", this coincides with the Cato finding that "Strong liberals (52%), racial minorities (54%), stand out with slim majorities who believe it’s more important for colleges to prohibit offensive and biased speech on campus".
The difference being Murphy did not know whom these leftists were, and described them simply as a
puzzling minority of vaguely leftist activists. Knowing Murphy's analysis suffered from a false artificat of the ambiguous meaning of the term “liberal", we need to look elsewhere to solve this mystery, after all, if a phenomenon is real, it's bound to be observed again and again no matter who does the survey, as long as the data is factual and analysis artefact free.
Enter the
Knight Foundation survey which the
Economist piece alluded to.
"More students now (61%) than in 2016 (54%) agree that the climate on their campus prevents some students from expressing their views. Although a majority of college students, 69%, believe political conservatives on campus are able to freely and openly express their views, many more believe political liberals (92%) and other campus groups are able to share their opinions freely. (pg 16)"
Seeing how Rep leaning students accounts for less than 1/4 of all students surveyed (weighted 27%), this means a sizeable number (if not all) of Rep leaning students feel the pressure to at least self-censor.
Compared with the 2016 survey, students now perceive the five freedoms as significantly less secure. This includes a 21-percentage-point decline in perceptions that freedom of the press is secure and nine-point declines for free speech. Shockingly, the biggest drop came from the Dems, while the Reps numbers stayed flat (pg 4). Why did Reps' numbers stay flat? The study says "Democrats were more likely to participate in all types (diversity and free speech) of protests than Republicans were." They simply refuse to engage and keep to themselves.
Recall what Kurt said "One note: deplatforming of college speakers, as judging by the FIRE “disinvitation database”, is being done far more these days by the Left than the Right." The drop in Dems' "security" numbers reinforces the notion that the "deplatformers" efforts are negatively impacting moderates and liberals.
The study concludes "Further, college students acknowledge that campus norms can act to deter speech — a larger majority than a year ago perceive that their campus climate prevents some people from expressing their ideas for fear of offending others. Also, students perceive that some groups on campus have a lesser ability to voice their opinions than other groups do."
Look. At this point, I can show you a thousand studies and it will likely do nothing. I have responded to the two studies you quoted. The Cornell study is of very limited use, if any, in the context of what we are discussing, and the Murphy analysis likely suffers from a false artefact of the ambiguous meaning of the term “liberal".
All the studies examined point to a group of leftists (perhaps extreme) actively seek to push their agenda to "supress free speech" as Murphy put it. I will leave it at that, you can decide for yourself if this is a problem.
----------------------------------------------------------------------------------------------------------------------------
A few words being I finish. My original lengthy post was the result of being invited by the OP, who had since gone silent. I was very reluctant to share my view at first because I knew it would be unpopular and I would be denounced by multiple posters.
At first I was attacked because I didn't show any data and most of it were "anecdotes". I then supported my points with studies but were accused of "bias" and worse. I showed in many follow-up posts that the studies I used are not biased, and the counter studies presented actually either reinforces my points, of limited use in the context of this discussion, or suffers from serious analytical artefact.
To the people who say studies/data are biased, do you even know what are the steps of designing a study/experiment? 1. Investigators have a hypothesis 2. Investigators design an experiment/study that might best capture and observe the hypothesized effect. And somehow that becomes a bias? IT IS A FEATURE, NOT A BUG.
True experimental bias comes in the form of non-factual data, or introducing artefacts into the analysis. It has nothing to do with whom paid or worked on it.
Do you, like, not remember anything you learned from high school science classes?
So far, I have been the only person (other than perhaps Guitar and occasionally golden1) that supported my view with studies and examined counter studies based on their own merits and robustness (instead of screaming bias with no proof whatsoever). Plenty of people have accused whites simply being resentful because "white people are giving up a clear advantage that racism brings them" yet have so far show no widespread data that supports such assertion.
This attitude that whites don't like this because they are racist (so far no one has provided any data that supports this. It is amazing to see such serious accusations levelled, casually, without adducing a shred of supporting evidence), is
very different from my points that whites don't like this because they feel excluded (which I have supported with studies).
I will end here because I see there is no point to engage further. It's been fun, I hope you all enjoyed the show.
ps. golden1, the clear contrast in terms of public opinion in 1963 and 1964 (40% favorable to 60% favorable) is clearly correlated with MLK's speech delivered in mid 1963. Kennedy's death prob also helped. It's not like you have personally come up with a reason that might explain the drastic change in public opinion between 1963 and 64. I am open to learning more and I will read the links you provided in my spare time.