Ten years ago today, Judge John Roberts took the oath of office to become the 17^th Chief Justice of the United States. Although we speak of “the Roberts Court”—its 10^th term now behind it, its 5^th under its current composition—it’s somewhat misleading to do so since it seems to imply that the chief justice has more power than in fact he has. To be sure, he leads the Court in a number of administrative respects, including the not inconsiderable power of assigning opinion writing when he’s in the majority in a given case. But at the end of the day, his vote counts for no more than that of any other justice.

Nevertheless, that’s the custom, so with those milestones before us, it’s worth asking how the Roberts Court is doing from a classical liberal perspective—liberty through limited constitutional government—the perspective we at Cato’s Center for Constitutional Studies have advanced since our inception over a quarter of a century ago. Given Roberts’ 2012 and 2015 opinions upholding Obamacare and his ringing dissent last June in the same-sex marriage case, one is tempted to answer “not well.” Those opinions speak volumes, about which I’ll say a bit more shortly. But on balance, it’s been a fairly good record. There are exceptions, for sure, and many cases are decided on technical grounds having little to do with substantive issues. But the Roberts Court has generally been supportive, for example, of property rights, religious liberty, free speech—especially political speech in the campaign finance context—and the Second Amendment, and it has mostly stood against affirmative action, executive branch overreach, and a number of other governmental intrusions.

What then of the exceptions—in particular, of Roberts’ own opinions in the two most important cases to come before the Court since he took over its helm, 2012’s NFIB v. Sebelius, upholding Obamacare’s individual mandate and thus the Act itself, and last June’s King v. Burwell, the statutory ruling that saved the Act from collapsing from its own unartful design? And what of the chief’s impassioned dissent a day later in Obergefell v. Hodges, the landmark same-sex marriage decision? It’s worth asking what those opinions say about Roberts’ very approach to the law and to the role of the Court in securing it.

I address those questions in some detail in the Foreword to the new Cato Supreme Court Review, the title of which, “Roberts’ Rules: Deference Trumps Law,” captures what I take to be the heart of the problem. Let me simply sketch the point here.

Recall the metaphor Roberts used during his Senate confirmation hearings: “Judges are like umpires. Umpires don’t make the rules; they apply them.” All well and good as a critique of the liberal judicial activism conservatives often complain about, often rightly. But it was followed immediately by talk of judicial “modesty” and “humility”—suggesting that Roberts might err in the opposite direction, that he might be too deferential to the political branches. More important, it doesn’t tell us anything about his conception of the Constitution.

When we read those three unfortunate opinions, however, we see how it is that Roberts’ extraordinary concern to avoid overriding the political branches trumps an accurate reading of the law. In NFIB, for example, he upheld the Affordable Care Act on the ground that the individual mandate could be read simply as a tax on those who chose not to buy insurance—even though he never said what kind of a tax it was, much less whether the tax was constitutionally cognizable.

In King, he dismissed as “ambiguous” what he himself said was “the most natural reading” of the ACA’s text, which authorized tax credits only for those who purchased insurance through “an Exchange established by the State,” ruling instead that that text could also be read as authorizing subsidies for those who purchased insurance through an exchange established by the federal government, thus saving the Act a second time by allowing subsidies in the 34 states that had declined to establish exchanges.

And in his lengthy dissent in Obergefell, Roberts offered perhaps his clearest statement to date of his understanding of the Constitution and the Court’s role in checking state actions that might violate rights or deny equal protection of the laws. In a nutshell, as with so many conservatives, it was clear from the start that he hasn’t yet come to grips with the full scope of the changes the Fourteenth Amendment brought to America’s federalism, which infused in the Constitution at last the classical liberal vision of the Founders that slavery precluded their accomplishing originally. Instead, like his fellow conservatives in this case, each of whom wrote an opinion, Roberts struggled within the deeply flawed post-New Deal methodology Progressivism set in motion, offering up an analysis that was far removed from the nation’s First Principles. His vision, like Robert Bork’s before him, is of a nation in which majorities are entitled to rule in wide areas simply because they are majorities, while in some areas minorities are entitled to be free from such rule. That gets Madison’s Constitution exactly backwards. Madison stood for liberty first, as our natural right, limited majoritarian democracy second, as a means toward securing that liberty.

Here too, then, from a mistaken inversion of our moral and political order, coupled with an understandable but inordinate fear of judicial activism, Roberts was led to defer to the states that had denied same-sex couples the equal protection of the laws. Judicial “modesty” became a cover for judicial abdication. That will not change until a more accurate understanding of the Constitution itself comes into view. Again, see that Foreword for details.