This study examines determinants of cohesion in the Supreme Court's reliance on precedent in the network of U.S. Supreme Court precedents. We use a database of all Supreme Court Citations and test several measures of annualized Court action as independent variables in a regression on several measures of network cohesion. We find that the magnitude of ideological decisionmaking is consistently associated with a reduction in network cohesion, suggesting that there is indeed a conflict between legal application of precedent and ideological decisions. This effect was roughly the same, though, regardless of whether the Court was conservative or liberal. We also find significant independent effects for Chief Justice eras, with the Burger Court associated with greater cohesion and the Rehnquist Court associated with lessened cohesion.
(Here's the link to the paper.)
In other words, we purport to find (objective) empirical evidence that when Supreme Court justices take more ideological positions, they depart from precedents more, and this happens with both liberal and conservative opinions. I think this is a very interesting result -- it suggests an essential compatibility between the rule of law and relatively apolitical, non-ideological, maybe even "objective" judging. Of course, one may ask how we measure how ideological a Court is, but I think I will get into that only when somebody actually asks. We use what are now widely used, perhaps becoming standard, measures in political science, but that is not to say they should be completely uncontroversial.
I think Robert is trying to make a serious philosophical argument, and my response is not on the same level of seriousness. But that said, I will just hazard the view that I think mere contestability is too high a standard to which to hold political science or most social science. This may mean, as Robert may be suggesting, that some contestable concepts may have to be agreed upon up front before any further enquiry can take place, and that this agreement should not be taken for granted. And, it happens all the time that an assumed, but never granted, agreement on such concepts can be leveraged so that some views get effectively excluded from the conversation. Any libertarian or conservative who has realized he is the only one at a conference with his views has had that experience, or try being the only liberal or theist at a conference where everyone else is a Marxist. But, even given all that, I still think in law for example, that much useful work can be done based on agreements about terms that may be contestable, but need not be contested for present purposes. So we might agree that Miranda and Roe v. Wade are "liberal" decisions, or that Justice Brennan or Marshall were liberal justices, and Scalia conservative. These are not objective calls, but having settled on them, you can go from there. And you may want to push back definitions further, for example by saying that Democrats are liberal compared to Republicans. I may be missing Robert's point, and I apologize if I am. Those of us who want to do empirical legal scholarship, however, or at least I, think that you can discover interesting things about our legal system if you are willing to agree on (what seem like) relatively uncontroversial starting points.