No one denies that Christopher Monfort, who is black, shot and killed Seattle police officer Timothy Brenton, who is white, in 2009. But was the act aggravated murder or insanity? That's up to a jury of Monfort's peers.

In order to select that jury, King County Superior Court spent months narrowing down a pool of 3,000 potential jurors to 12 final jurors and four alternates. In a death-penalty case, jurors have to be willing not only to consider guilt, but willing to consider giving someone a death sentence. If a potential juror would never sentence anyone to death for any reason, he or she is disqualified, just as a juror who would always sentence a person to death would be disqualified.

Lawyers call this part of the process of jury selection "death qualification." Capital-defense lawyers have long argued that death-qualified juries tend to exclude more black jurors than white jurors. And research has shown that death qualification does indeed exclude a greater proportion of black jurors than whites. But it's been only in the last few years that social scientists have found new tools to look at how death qualification primes juries to sentence black defendants to death.

"It is the essential unfairness of the death penalty," says Ben Cohen, a lawyer at the Promise of Justice Initiative in Louisiana. "[Death qualification] transforms Seattle into a suburb of Dallas, Texas, because everyone who opposes the death penalty is gone."

The way Cohen explains it, the history of American racism, the process of death qualification, and the death penalty are inseparable. One of the first recorded instances of a jury made to pass a death-qualification test was in 1859, when the point was to make sure those opposed to slavery and Quakers—whose religious beliefs led them to object to the death penalty no matter what—didn't get in the way of the execution of rebel abolitionist John Brown. In the end, Brown was found guilty of murder, treason, and conspiracy, and he was hanged.

In the Monfort trial, which is under way right now, defense attorneys tried to use new research on death qualification to get the death-penalty option thrown out. In early October, Monfort's defense team arranged for University of Hawaii law professor and death-penalty social scientist Justin Levinson to fly to Seattle. In King County Superior Court, Levinson explained why jurors who wouldn't even be aware of their biases might make decisions based on the automatic assumption that black lives are less valuable than white ones.

Levinson has been a pioneer in this arena. Six years ago, Levinson sat 67 students at the University of Hawaii in front of computer screens. He primed the students by asking them to sort digital images of light-skinned male faces and dark-skinned male faces into categories of "Black" or "White." Then Levinson instructed the students to hold their fingers over the "D" and "K" buttons on their keyboards, and the students sorted words like "fault," "convict," "innocent," "wrongfully accused," and "perpetrator" into the categories of "Black," "White," "Guilty," and "Not Guilty" as fast as they could.

The computers measured the students' reactions in milliseconds. The student test subjects were undergoing something called the implicit-association test (IAT), a tool social scientists have been using since the late 1990s to try to illuminate pre-conscious biases in people's brains. Other IATs had measured how people linked race to "good" and "bad" qualities, but Levinson was the first to try to see how implicit bias—judgments that people may not even be aware that they make—might affect the presumption of innocence in a courtroom.

Not only did Levinson find that the students implicitly linked "Black" and "Guilty," he also discovered that those biases predicted how students would judge evidence as mock jurors in cases against dark-skinned defendants. But that wasn't even the strangest or most disturbing part of the University of Hawaii experiments. When Levinson asked the same students to rank the "warmth" of their feelings toward black people, the implicitly racist students actually rated friendlier feelings toward black people than the students who held less bias.

"People have come to understand that expressing racist and racially biased attitudes and stereotypes is not socially appropriate," Levinson says. "So this trend has emerged at the same time that social-science research that has figured out ways to measure automatic biases, and that research established again and again that implicit biases are strong and pervasive."

In 2013, Levinson published a study in the NYU Law Review, this one looking at how covert racism might affect sentencing in death-penalty trials in particular. He and his colleagues asked 445 jury-eligible citizens to imagine themselves as jurors studying evidence inspired by an actual case. This IAT, conducted online, measured how the race of a defendant might affect a person's judgment on the value of that person's life. This time, the test found that mock jurors associated "White" with worth and "Black" with worthless.

Death-qualified mock jurors, the ones who had answered questions about their ability to consider the death penalty and weren't excluded, displayed even stronger implicit and explicit racial biases than the general study sample. Stronger racial bias, Levinson found, also predicted more death sentences for black defendants when the victim was white. "The higher the level of bias on the Value of Life IAT, the more likely they were to vote for death when the defendant was black," Levinson says.

After Levinson testified about all of this, the King County court had to decide: Would jury selection in Monfort's trial taint the judicial process itself? King County prosecutors rebutted Levinson's testimony by filing a response poking holes in his studies. Implicit bias doesn't necessarily affect group decision making, they argued, and Levinson had never tried out his theories with a real jury. Prosecutors also tried to discredit the legitimacy of the IAT, despite the fact that the IAT is widely regarded as a credible scientific tool.

The prosecution went one step further, too. They argued that Levinson had a wider agenda of discrediting criminal justice altogether by highlighting implicit racism throughout the system. And they flipped the racism allegation: "The assertion that members of a particular minority group cannot and should not be held criminally responsible for their actions in the way that other citizens are is demeaning and insulting to them, and serves to increase racial suspicion, resentment, and bias throughout our nation's diverse population," the prosecutors claimed.

"I would never suggest that anyone shouldn't be held responsible for their criminal actions," Levinson argues back. "The whole point of this research is to move toward fairness in the criminal justice system."

King County prosecutors declined to be interviewed for this article, as did the judge in Monfort's trial, Judge Ronald Kessler. But court documents speak for them. Judge Kessler ended up dismissing the defense's motion to get rid of the death-penalty option based on Levinson's testimony. (Judge Kessler accepted another of the defense's motions to get rid of the death penalty for the case, but it was subsequently reversed by the state supreme court.)

Monfort's attorneys, however, still argue that death qualification impacts death-penalty juries, including the jury that is now hearing Monfort's case. (The trial began January 20.) The whole process of jury selection is wrong from the get-go, argues public defender Stacey MacDonald, "because you're focused on, before he's found guilty or not, can you kill him or not."

But even if courts do select for implicitly racist juries, is that enough to call the practice of death qualification unconstitutional? "No court has decided that," says Jim Lobsenz, an adjunct professor at the Seattle University School of Law and vice president of Washington Coalition to Abolish the Death Penalty. "It's a pretty cutting-edge kind of question. You have to be more like a philosopher to wrestle with this question."

The judge makes the ultimate decision as to whether jurors can be disqualified because of their feelings about the death penalty. In that way, the death-qualification process could be read as one of the few ways judges are able to make sure a jury is fair. The quality of the judge has an impact on the makeup of the jury.

But sussing out a juror's life-and-death philosophy can be a pretty circular task. When asked about the death penalty, one of the candidates who didn't end up as a Monfort juror wrote that he or she had taken a Buddhist oath not to kill. "But on the other hand," the juror wrote, "my spiritual beliefs tell me that death is not that big of a deal."

Even with a good judge, a jury might not be made up of a defendant's peers—depending on how you define "peers." If a juror of a different race than the defendant lacks insight into the experience of racism that the defendant has lived, is that juror a peer or not?

Levinson's work strongly suggests that death qualification "transforms the jury," says Cohen, of the Promise of Justice Initiative. And Cohen backs up this finding anecdotally: "You can see that just by watching death qualification, and seeing African Americans looking over at an African American defendant and understanding a history of lynching in the United States, that they're uncomfortable with the use of the death penalty, and white people feeling the exact opposite, and watching how race plays out. Justin's work proves what you would understand just by watching."

King County prosecutors would probably disagree—but they already got what they wanted: a death-qualified jury.