To follow up on Paul’s post below, since political science models have played a role in advancing the strongest versions of legal realism, it’s worth explaining what a more realistic version of legal realism looks like.
One way I like to explain it is borrowed from Michael Klarman’s superb book From Jim Crow to Civil Rights. In short, legal texts range on a continuum from determinate to indeterminate, and a judge’s ex ante political and/or legal policy preferences range on a continuum from weak to strong. If the law is clear, judges will generally follow the law unless they have extremely strong reasons for not doing so. As the law becomes less determinate, legal questions will (and, indeed, must) be decided by factors other than strictly “legal” ones.
The reason the attitudinal model works fairly well for Supreme Court decisions on the merits is that — at least under the current rules which give the Court has near-complete control over its docket — cases generally cluster well towards the “indeterminate” end of the spectrum and are also disproportionately likely to involve issues justices have strong ex ante preferences about. But it would be wrong to therefore conclude that the strongest versions of legal realism were correct. The law is still doing a lot of work to determine which issues end up in court and which don’t, how the questions that do come to the court are framed, etc. And, of course, the votes the attitudinal model can’t easily explain are often the most important.
The other text I like to point to is Mark Graber’s essay about judicial decision-making. Not only do many important decisions have demonstrable legal, political, and strategic components, judicial decisions cannot be fit into neat scientific categories in part because what is “legal” and what is “political” is always being contested at the margin. But this cuts both ways — “law is nothing but politics” can be as misleading as “judges are neutral umpires mechanically calling balls and strikes.”
Still, legal constraints are ultimately norms. Whether American democracy will avoid serious further decline will depend on whether institutions like the Supreme Court are bound by crucial norms or decide to start doing anything they have the power to do. The Supreme Court had the formal authority to issue a (lawless) ruling that may well have allowed the party of the majority of the Court to retain control of the House in 2018. That they didn’t is actually pretty important.