Skip to main content
ABC News
Is Chief Justice Roberts A Secret Liberal?

In 2009, The New Yorker’s Jeffrey Toobin wrote that Chief Justice John Roberts, more than any of his colleagues, had “served the interests, and reflected the values, of the contemporary Republican Party.”

My, how things change.

More than a decade after he was first appointed, Roberts does not appear to be the justice he was first made out to be. That could prove pivotal in the next few years, as any of several justices may retire. We could soon see a Roberts Court in more ways than one: with him as both chief justice and swing justice, sitting at the ideological center of the bench. But what kind of court would that be? It’s relatively early in Roberts’s Supreme Court career, but he is beginning to fit the historical pattern, at least quantitatively, of an ideological defector.

The case for a liberal John Roberts starts with his 2012 decision that determined the fate of the Affordable Care Act. The chief justice was widely expected to vote to kill the law — the crown jewel of the Democratic president whom he’d sworn into office three years earlier. Instead, he performed some contorted judicial yoga, declaring that the law’s individual mandate was a constitutionally allowed tax, siding with the liberal bloc and saving Obamacare.

Ever since that decision, the right has been concerned. Just this summer, the conservative news site The Daily Caller noted that Roberts “sided with the Supreme Court’s liberal bloc in two racially-tinged cases this term, breaking with his conservative colleagues in favor of housing-rights activists and a black death row inmate.” In 2015, the conservative National Review described the concern colorfully, and geographically: “John Roberts and Anthony Kennedy will, if the goblins in their heads are sufficiently insistent, ratify whatever Starbucks-customer consensus exists for 80 miles on either side of Interstate 95.”

The concern has even manifested in an adjective that’s been repeatedly applied to Roberts: “wobbly.”

We don’t know for sure what caused Roberts’s leftward shift. Those who talk about the inner workings of the court don’t know, and those who might know don’t talk. None of the many former Roberts clerks I reached out to for insight would talk to me. But we do know there’s been a shift.

To quantify ideology, I’ll use Martin-Quinn scores, a prominent measure created by two political scientists that use justices’ actual votes to place them quantitatively on a left-right spectrum, like DW-NOMINATE scores do for legislators.developed a probability model that turns a justice’s votes on cases, combined with the votes of her colleagues, into a single number, which can vary over time.

">1 Justices’ shifts over time are usually mild — a matter of degree rather than of kind. But sometimes these shifts have been enough to unravel and reweave the fabric of the court, remaking the law of the land for decades after.

Four modern justices were appointed by Republican presidents and began their high-court careers as conservatives, but shifted left, crossed the center line, and ended up as liberals.

A brief history of the men above:

  • Justice Harry Blackmun was nominated by Richard Nixon in 1970 and confirmed unanimously by the Senate. In 1973, he authored the court’s opinion in Roe v. Wade, which protected the right to abortion.
  • Justice John Paul Stevens was nominated by Gerald Ford in 1975. A 2007 Times magazine profile of Stevens called him “The Dissenter” and mentioned the word “liberal” 44 times.
  • Justice David Souter was nominated by George H.W. Bush in 1990; Bush’s chief of staff said that he would “be a home run for conservatives.” He later voted to reaffirm Roe and dissented in the court’s decision in Bush v. Gore, which handed the presidency to the younger Bush.
  • And Justice Anthony Kennedy was nominated by Ronald Reagan in 1987. (Kennedy is not a solid liberal but is known as the court’s swing vote.) In 2015, he wrote the opinion that legalized gay marriage nationwide.

Roberts may yet join that list.

It’s not just the Martin-Quinn scores that provide evidence of a leftward shift. We can also see the change in the scores’ ingredients — the justices’ votes. In 2010, for example, Roberts agreed with the positions Justice Clarence Thomas, the most conservative justice according to the Martin-Quinn scores, 89 percent of the time and with Justice Samuel Alito, another conservative, 96 percent of the time, per the SCOTUSblog Stat Pack. He agreed with liberal Justices Elena Kagan and Ruth Bader Ginsburg 69 and 65 percent of the time, respectively. By 2015, with that same court intact, his agreement with those two conservatives dropped to 75 and 84 percent. His agreement with those two liberals spiked to 87 and 78 percent.

How often Roberts agreed with …
2010 2011 2012 2013 2014 2015
Samuel Alito Jr. 96% 91% 90% 85% 81% 84%
Antonin Scalia 90 86 85 90 84 88
Anthony Kennedy 90 84 85 92 70 88
Clarence Thomas 89 88 86 88 70 75
Stephen Breyer 72 70 74 85 72 84
Sonia Sotomayor 71 71 65 79 69 77
Elena Kagan 69 73 67 83 65 87
Ruth Bader Ginsburg 65 64 65 77 69 78

Source: SCOTUSblog

Despite the trends in the metrics, many of Roberts’s past Supreme Court positions have, without question, been archly conservative. His majority opinion in Shelby County v. Holder invalidated key provisions of the Voting Rights Act. He was in the majority in Citizens United, the liberal bête noire that protected the injection of corporate money into politics. The Hobby Lobby case was a victory for corporations and a blow to contraception access. He was highly skeptical of social science arguments against partisan gerrymandering in oral arguments this term. (That case has not yet been decided.)

Perhaps because of those decisions, legal experts I spoke with cautioned against reading too much into Roberts’s leftward slide, and they doubted that he is in the early stages of some bloodless liberal judicial coup.

Josh Blackman, a law professor and author of “Unprecedented: The Constitutional Challenge to Obamacare,” attributes the historical “defections” not so much to shifting ideologies but rather to poor initial vetting or presidential indifference to the true political beliefs of their nominees. Souter: merely a New Hampshire Republican. Stevens: never conservative in the first place. Kennedy: ditto. Blackmun: a moderate picked as a compromise after the Senate rejected Nixon’s two previous nominees. But Roberts: a real-deal conservative? “Roberts is no shrinking violet,” Blackman said. Roberts becoming a reliable liberal vote “is the liberal fantasy. But no matter how much Linda Greenhouse wishes it, it’s not going to come true.”2

But it’s not as though there’s nothing to the trend. This type of empirical shifting is common, but not universal — it means something that Roberts is on the move.found that, “of the twenty-six Justices who served on the Court for ten or more terms since 1937, all but four exhibit ideological drift over the course of their tenures. Twelve moved to the left, seven to the right, and three in more exotic ways.”

">3 For context, let’s take a look at one of Roberts’s colleagues.

Alito was confirmed to the court four months after Roberts. Both were nominated by George W. Bush and confirmed by a Republican-controlled Senate. Both were federal circuit judges before joining the Supreme Court. Roberts was an editor for the Harvard Law Review and Alito was an editor for the Yale Law Journal. And in their first two terms on the bench, both had nearly identical ideologies.

But beginning around 2007, Roberts and Alito diverged. Roberts now has a Martin-Quinn score more liberal than Kennedy’s score was five years ago, and more liberal too than that of Souter — a “defector” — at certain points in his career. Alito’s score, meanwhile, is about the same as Justice Antonin Scalia’s was when Scalia died last year.

And so we return to the question of why Roberts is drifting to the left. Kevin Quinn, a political scientist at the University of Michigan and co-developer of the Martin-Quinn scores, thinks Roberts’s shift may have something to do with his status as chief justice. “I think it is probably reasonable to assume that at least some of that movement was more a reflection of the institutional realities of being chief than a genuine change in his view of the law.”

Quinn said that former Chief Justice William Rehnquist, rather than Blackmun or Souter, might be the best point of comparison. Roberts clerked for Rehnquist in the early 1980s and took over his seat on the bench after Rehnquist died in 2005. Visually, Roberts’s ideological path is a remarkably fitting continuation of Rehnquist’s.

The chief justice, as opposed to the eight associate justices, is afforded certain special powers. One of these is critical. After the justices hear arguments in a case, they sit in a closed-door meeting called a conference where they cast their initial votes. If the chief is in the majority, he assigns the justice who will write the court’s opinion for that case. The theory goes that this protocol might give Roberts incentive to side with the liberals and vote against his truly held ideology so that he can assign himself the opinion, or assign it to another conservative justice in the majority. That way, the court’s opinion could be crafted in a way that’s more palatable to the chief justice. Strategic moves like these would shift Roberts’s Martin-Quinn score to the left but have no major impact on the actual decisions of the court.

There is evidence, however, that this isn’t Roberts’s game.found no evidence that the former chief behaved this way either.

">4 Roberts has never not been chief, so if this strategy is what’s causing him to appear more liberal, he would have had to have started using it only in the last six or seven years, since he was steadily conservative for his first few years on the court. And Rehnquist, who underwent a similar shift, began moving left well before he became chief, so in at least one case we know that the chief’s powers weren’t the sole cause of a political realignment.

If Roberts really is liberalizing, how might that help define the law of the land? One place to look for clues is how Roberts votes in close, 5-4 cases. Since he took his seat, he has presided over 169 such cases, according to the Supreme Court Database. It’s hard to tease out much of a trend because each year only has so many cases to work with. In the 2006 term, Roberts voted in a liberal direction in just 8 percent of close cases.5 The next term, that number was 8 percent again, then 17, then 31, but then 6. By the 2013 term, the number was 30, and then 28 in 2014.6 If you squint, you can see a liberalizing trend, but it’s highly erratic.

This term alone, the court will decide the fate of partisan gerrymandering, shaping the future of American democracy. It will decide a case about a business owner refusing to serve a same-sex couple, shaping the future of gay rights. It will decide a case about cell phones and the privacy of our data. These votes will likely be close. In terms to come, the court may tackle the death penalty, abortion rights, gun rights and issues we can now only imagine.

Chief Justice John Roberts may wobble. No one but he knows where he will fall.


  1. Scholars Andrew Martin and Kevin Quinn developed a probability model that turns a justice’s votes on cases, combined with the votes of her colleagues, into a single number, which can vary over time.

  2. Greenhouse was the New York Times’ Supreme Court correspondent and was known for criticizing the decisions of conservative justices. The theory goes that these justices, not wanting to incur her wrath in print, would shift their positions to the left — the so-called Greenhouse Effect.

  3. In 2007, Martin, Quinn and two other coauthors found that, “of the twenty-six Justices who served on the Court for ten or more terms since 1937, all but four exhibit ideological drift over the course of their tenures. Twelve moved to the left, seven to the right, and three in more exotic ways.”

  4. And at least one study that tested the “strategist chief justice” hypothesis on Warren Burger’s court found no evidence that the former chief behaved this way either.

  5. Two out of 24.

  6. Because Scalia’s death led to a yearlong vacancy on the court, there were no 5-4 decisions in the 2015-16 term, and there were only three such decisions in the 2016-17 term.

Oliver Roeder was a senior writer for FiveThirtyEight. He holds a Ph.D. in economics from the University of Texas at Austin, where he studied game theory and political competition.