John McGinnis on Christmas and the Constitution
At Liberty Law Blog, John McGinnis: The Christmas Holiday of 1870 and the Establishment Clause. The main argument:
But instead of thinking what the current Establishment Clause doctrine means for the Christmas holiday, we can turn the question around and ask what the Christmas holiday means for the meaning of the Establishment Clause. Christmas was first declared a federal holiday in 1870. At that time Christmas was a predominantly religious holiday without the overwhelming commercial aspect of our contemporary world.
As far I can tell, there was no substantial objection to the holiday on the basis of the Establishment Clause. The lack of objection to giving official status to a predominantly religious holiday suggests that the meaning then attributed to the Establishment Clause does not comport with the so-called “endorsement test” employed today to test for violations of the Clause. The government proclaimed a Christmas holiday, although that proclamation, like the forbidden religious Christmas display today, may give the appearance of endorsing a religion.
Recognizing the constitutionality of religious holidays, however, is consistent with the so-called “coercion test” for the Establishment Clause that holds that only coercive support of religion, like forcing citizens to pay taxes to a church, triggers its prohibition [Ed.: As Justice Scalia argued]. A holiday, even one which is predominantly religious, is not coercive in this respect. Even if it suspends commercial obligations till the next day, as does the bank holiday created by the 1870 law, citizens will not be made substantially worse off by a brief delay that is applied equally to all.
[S]ome constitutional theorists might object to this type of evidence entirely, because it concerns an “an expected application” of a constitutional principle rather a parsing of the language of the principle itself. Mike Rappaport and I have argued that such a dichotomy is false. The meaning of moral and legal principles can sometimes be best be pinned down by looking at their applications.
I especially agree with this last point. The Constitution's commands are formed by its text, not by its expected applications. But expected applications can be important evidence of meaning. In this example, "establishment of religion" doesn't have an meaning in the abstract; for an originalist, its meaning is whatever the phrase meant at the time it was enacted. How people expected the phrase to be applied to concrete circumstances (e.g., did they think declaring a Christmas holiday was an "establishment"?) is good evidence of what they thought an "establishment" was. (And there is no relevant meaning of "establishment" apart from what they thought it meant).
True, they could have been wrong (or, as likely, they could have failed to live up to their own aspirations). But demonstrating that is something of an uphill battle, requiring evidence from the relevant period of a contrary meaning. I would say that the expected application, if directly on point, puts the burden of proof on the side that would find a contrary meaning.