Much of the time, Pildes argues, our doctrine is framed in formalist terms—that is the black box version. For instance, when federal courts yield to state courts under a range of abstention doctrines, they employ a formal rule that presumes that state courts are just as adequate as federal ones. And they do this without actually worrying about whether that presumption is empirically true in the specific case, either for this state or this moment in time. On the other hand, sometimes doctrine takes a realist turn, for instance when the Supreme Court dealt with certain southern state courts in the 1950s and 1960s.
There’s much more specificity to the article, of course, across a range of institutions, but I think the basic point is insightful and profound. And it also helpfully frames the dilemma that courts are facing in reviewing executive actions of the Trump presidency. If one take an “institutional formalist” approach to the separation of powers, then one might defer to the President’s national security decisions even if one doubts that he has made them on the basis of national security. If one takes an “institutional realist” approach things might look different—as reflected in Ben Witters & Quinta Jurecic’s posts on “what happens when the judiciary doesn’t trust the President’s oath.”
Now it probably will not surprise readers to hear that in general my sympathies lie with formalism. But I think it’s helpful to realize that this is actually a recurring problem in public law, not a completely unprecedented situation created by the current administration. Unfortunately, it is not a problem that public law has solved, and it doesn’t look to me like that will change in the near future.