I take the title of this post from Alexander Bickel. The post is a reflection on the articles in the Symposium on "Testing the Constitution" recently published in the NYU Law Review. The symposium's editors frame the articles by observing that scholarship in constitutional law is rarely empirical (according to their definition). The articles (purport to) describe Supreme Court decisions that rest in part on empirical judgments, and offer studies aimed at illuminating the accuracy of the Court's empirical judgments.

I had difficulty early on, when the editors explain their own study of the incidence of empirical studies of constitutional law. Their coding "counted articles as empirical is they made some use of data (facts about the world)." There's no indication in the article that they included law review articles dealing with constitutional history, defined here to include originalist inquiries. Yet, to me, history is an empirical social science, seeking data in the sense of facts about the world. (Or, if you raise fancy epistemological questions about the status of historical studies, you surely ought to raise the same fancy questions about the studies the editors clearly do describe as empirical -- including the articles in the Symposium.) I'll return to the omission of historical studies from the category "empirical" at the conclusion of this post.

Many of the articles in the Symposium have the air of "lamppost" scholarship (you know the joke about looking for car keys under the lamppost). Authors tug and haul Supreme Court opinions to extract some proposition plausibly described as empirical, then look around for "rough" proxies for information directly bearing on the empirical proposition ("rough" and similar terms crop up reasonably frequently in the studies), and then test the proposition using the proxies.

Sometimes this is reasonably effective, but more often it is quite strained. So, for example, one can manipulate the "marketplace of ideas" metaphor (a) to make it empirical (maybe not that difficult), and then (b) to locate empirical data that shed some light on your version of the empirical claim. But, as executed, the study makes Holmes and Brandeis look like idiots because they thought that they were making as vapid an empirical claim as the one the article examines. If the metaphor makes an empirical claim, it a far more complex one, and almost certainly has to do with the effects of dissemination of ideas over reasonably extended periods. The study "begets" an answer, but it's an answer to a question that Holmes and Brandeis weren't asking.

Another example comes in an article examining the purportedly empirical claim made by Justice Kennedy in Citizens United that "the appearance of influence or access ... will not cause the electorate to lose faith in our democracy." I have some questions about the construct-validity of the study used to test that proposition. But, more important, though one can read Justice Kennedy as making an empirical claim, anyone who did so is, to quote John Hart Ely, a booby. (I think he made that observation in connection with some arguments around Cohen v. California, but don't want to take the time to track them down -- and I don't endorse his implicit critical comments about boobies, which are quite attractive [in their own way] birds.) Justice Kennedy was, it seems pretty clear, making a statement about his (our?) faith in the good sense of the people of the United States.

Had the editors of the Symposium treated history as an empirical discipline, they might have framed the discussion differently. For, there's a well-developed literature on the difference between history-as-historians-do-it (call it "History sub-one"), and history-as-lawyers-do-it (call it "History sub-two"). The two ways of doing history are simply different, and it's a category mistake, though one committed quite often, to assess History-sub-two according to the criteria of History-sub-one. So too, I think, with respect to seemingly empirical claims in Supreme Court opinions. The practice of social-science-as-justices-do-it is a different social practice from social-science-as-social-scientists do it. So, though some of the articles in the Symposium illuminate something, it's not clear to me that they illuminate constitutional law. "No answer is what the wrong question begets."