A North Carolina judge commuted the death sentence of Marcus Robinson, a black man convicted of killing a white teenage boy after abducting him in a car and robbing him at gunpoint, based on a new state law (the idea came up on the federal level in the 90s, but never came to fruition) that prohibits race from being a factor in the application of the death sentence. The ACLU explains what happened in the Robinson case:
[P]owerful statistical evidence of racial bias in jury selection was introduced [by defense attorneys], including a study from Michigan State University finding that North Carolina prosecutors were twice as likely to remove qualified Black jurors from jury service as other jurors, even after the researchers controlled for alternative explanations such as criminal background or reservations about imposing a death sentence.
The state offered no meaningful rebuttal to the statistical evidence. No statistical expert testified for the state that race did not play a role in jury selection. Rather, the State lodged a frontal attack on the concept of statistical evidence itself. In its closing argument, the prosecution argued that the problem with the Robinson's statistical evidence is that it tries "to get people to lose sight of the trees and focus on the forest." At the end of the argument, the prosecution was even more direct: it pleaded with the judge not to make a decision "with respect to the Racial Justice Act based upon numbers."
In a 167-page ruling, Judge Gregory Weeks disagreed with the prosecution's argument, ruling that "race was a materially, practically and statistically significant factor in the decision to exercise peremptory challenges during jury selection by prosecutors."
Of course none of this would matter if the death penalty became a thing of the past.