Jeff Rosen has a very good piece on John Paul Stevens in the Times Magazine. The central point is that Stevens isn’t so much a liberal as someone who plays one on the Rehnquist and Burger Courts:
Stevens, however, is an improbable liberal icon. “I don’t think of myself as a liberal at all,” he told me during a recent interview in his chambers, laughing and shaking his head. “I think as part of my general politics, I’m pretty darn conservative.” Stevens said that his views haven’t changed since 1975, when as a moderate Republican he was appointed by President Gerald Ford to the Supreme Court. Stevens’s judicial hero is Potter Stewart, the Republican centrist, whom Stevens has said he admires more than all of the other justices with whom he has served. He considers himself a “judicial conservative,” he said, and only appears liberal today because he has been surrounded by increasingly conservative colleagues. “Including myself,” he said, “every judge who’s been appointed to the court since Lewis Powell” — nominated by
Richard Nixonin 1971 — “has been more conservative than his or her predecessor. Except maybe Justice Ginsburg. That’s bound to have an effect on the court.”
It is a measure of how not only how much the Court has changed but how much the Republican Party has changed that Rockefeller Republicans now seem like liberals on the Supreme Court. There’s no Brennan, Marshall or Douglas on the modern Court. There have been some liberal advances, but that have been mostly modest expansions of existing doctrines agreeable to moderate northern Republicans: overturning a widely derided 5-4 decision that the swing vote repudiated almost immediately to strike down laws that were sporadically and arbitrarily enforced, and striking down a couple unusual applications of the death penalty that represent a small fraction of the total number of cases. And as Souter demonstrated, a Harlan-like incrementalist is obviously going to look more liberal after the Warren Court than when on the Warren Court.
The other thing to mention here is that the Burger and (especially) Rehnquist Court shifted doctrine to the right in subtle ways that makes it seem as if it changed less than it did. It’s true that the Court has generally avoided overturning major Warren Court landmarks — but it has often substantially alerted their content. The Court, of course, has never considered overturning Brown, but it has defined it to require only formalistic non-segregation as opposed to actual substantive desegregation (and has also made it very difficult for school districts to voluntarily desegregate.) Miranda has never been overturned, but any number of exceptions to it have been carved out. Casey is remembered primarily for re-affirming Roe, but also allowed the states (and the federal government) substantially more leeway to regulate abortion. And so on. Especially when dealing with the Roberts Court, it’s important to look at the substantive outcomes of the cases, not at how they characterize the precedents.