Whose Speech, Which Religion?

May 8, 2014
Posted by Jay Livingston

Does a justice’s view of the First Amendment’s “establishment clause” depend on which religion is being established? 

The First Amendment doesn’t specify any religions as more or less establishable. It just says no establishment.

This week, five conservative justices on the Supreme Court voted to allow a town council in Greece, NY to open their meetings with Christian prayers. These referred to “our Christian faith,” Jesus Christ, and the Resurrection. The justices ruled that these Christian prayers were in perfect accord with the First Amendment.  Needless to say, the five justice majority was all Christain (Catholic in fact). The two Jews and two other Catholics dissented. (The Court has no Protestants.)

 I wonder what the decision would have been if the town had had a Muslim majority that opened each meeting with “There is no god but Allah, and Mohammed is his prophet,” and the local Christians filed the First Amendment suit.

We don’t know, but on Wednesday, the Times (here) published some data from a study by Lee Epstein of decisions on another First Amendment issue – freedom of speech.  In general, the liberal justices were more likely to hold for freedom of speech, conservative justices were for regulation. But the content of the speech mattered. When the speech was conservative, the conservative justices suddenly saw the virtues of civil liberties arguments. The effect also held for the liberal justices but was much weaker.

* p < .05
(Click on the chart for larger view)

Look at Scalia and Thomas (Nino and Silent Clarence). When the speech in question is liberal, they come down on the side of free speech in maybe one case in five.  But when the speech is conservative, they take the speakers’ side in nearly two-thirds of the cases.  They are three times more likely to rule for a conservative speaker than for a liberal speaker. 

The effect for Alito and Roberts is even stronger; they are four or five times more likely to support free speech for conservatives than for liberals. The number of cases is small – 27 and 24, respectively, compared with 104 for Thomas and 161 for Scalia – but the differences are still statistically significant.

As for a comparison with the hypothetical Muslim town council and its prayers, a similar study would be impossible because there is no variation. The religion being established is always Christianity.  Still, given the evidence on speech cases, those five justices might become downright Jeffersonian in their views of church and state. Scalia would don his originalist cape and claim that “no establishment” means just what it says and that this prayer, that just happens to be Islamic, is in violation of the First Amendment. Pay no attention to that Christian prayer we green-lighted in that other case last year.

How might Scalia respond to the evidence of the justice’s bias in speech cases?  Probably the way he responded to evidence of racial bias in death penalty cases – to dismiss it as irrelevant. In McCleskey v. Kemp (1987), he agreed with the argument that social science evidence about Black-White differences in these cases was irrelevant. McCleskey would have to prove bias in his particular case – an impossible hurdle to clear.  Of course, if social science data supported a position that Scalia agreed with, he just might make a belated, if temporary, discovery of the virtues of sociological findings.


brandsinger said...

You summarize the study here:
"Look at Scalia and Thomas (Nino and Silent Clarence). When the speech in question is liberal, they come down on the side of free speech in maybe one case in five. But when the speech is a conservative, they take the speakers’ side in nearly two-thirds of the cases."

Surely the validity of such a study depends on what the researchers define as "liberal speech" and "conservative speech." I guess you believe that this distinction is obvious and objectively made ("liberal speech" vs. "conservative speech"), as if with the difference between fire and ice.

I would say that categorizations of "liberal speech" and "conservative speech" are subjective — and that this study — like so many others conducted by and for the liberal establishment to prove conservative opponents dunces and hypocrites — is bogus.

In my lifetime, it has been liberals who have wanted speech restricted (e.g. political correctness on campus, shouting down people they object to, campaign finance restrictions, tossing Juan Williams out of NPR, impact of the "fairness doctrine," etc. etc.). These activities should be seen as obnoxious to all believers in the First Amendment — and would be opposed by jurists who defend freedom of speech… (of which Justice Scalia would be among the foremost).

Jay Livingston said...

Designating a speaker as liberal or conservative is subjective. But in most cases, most of us share that same subjectivity. Look at your own statement about campus speech codes, campaign finance restrictions, Juan Williams, etc. If I took your “it’s all subjective” perspective, I would argue that your characterizing these as liberal positions is purely subjective, and therefore you’re conclusions are bogus.

C’mon Claude (as a certain brand consultant I know might say), in most cases, we all know what’s liberal and what’s conservative. A kid who unfurls a “Bong hits 4 Jesus” banner – how many people would say he’s a conservative speaker? Pro-lifers who want to take their protests closer than 36 feet to an abortion clinic – would you classify them as liberals? (And take a wild guess as to which side Nino and Silent Clarence came down on in those cases. Hint: they were for expression on one and against it on the other.)

In fact, the intercoder reliability on this was close to 100%. Also, the fuzzier the measurement of a variable, the less likely it is to produce clear differences. Imagine that we’re trying to see if height is correlated to weight. We have a yardstick and a scale, and we find a strong correlation. We conclude that people who are taller tend to weigh more. But suppose the scale is broken and gives us completely random readings on weight. The measurement of that variable is unreliable. The result is that our correlation will disappear; we will not be able to make any conclusions about weight differences between tall people and short people.

You can read the study here Actually reading a study before judging it biased and bogus – what a concept!