First Ruling Under the Racial Justice Act

Last Friday, Cumberland County Senior Resident Superior Court Judge Greg Weeks issued the first merits ruling on a claim under the Racial Justice Act (“RJA”), G.S. 15A-2010 et seq. Judge Weeks’ order, which vacates the death sentence previously imposed on Marcus Robinson, is here. It’s 168 pages long, so I suspect most people haven’t read the whole thing. Maybe you’ve read a newspaper article about it, like this one from the News and Observer, or this one from the New York Times. But maybe you’ve been waiting to get the skinny from this blog. Here’s what happened.

In 1991, Robinson killed Erik Tornblom. Robinson was tried capitally in 1994, and was convicted and sentenced to death. His direct appeal and collateral attacks were generally denied. When the RJA became law, Robinson filed for relief under the Act. Earlier this year, Judge Weeks heard evidence on one aspect of Robinson’s RJA claim: his contention that “[r]ace was a significant factor in decisions to exercise peremptory challenges during jury selection.” (The Act also allows defendants to claim that the race of the defendant or the race of the victim influenced the decision to seek or to impose the death penalty. I believe that Robinson also made claims under those provisions of the Act, but those claims weren’t heard by Judge Weeks.)

The order begins by summarizing the testimony presented by the parties at the hearing. The key evidence for Robinson was a study conducted in part by, and the testimony of, Dr. Barbara O’Brien, a law professor at Michigan State University. In a nutshell, the study examined jury selection in 173 North Carolina capital trials resulting in death verdicts. O’Brien and her colleagues looked at each juror the state had an opportunity to strike; identified the juror’s race; noted other factors that might be relevant to the prosecutor’s decision regarding the juror; and recorded whether the state struck the juror. The study concluded that racial discrimination pervaded jury selection in capital cases in virtually every county, prosecutorial district, and judicial division in the state, as well as in the state as a whole. The state’s presented testimony from Dr. Joseph Katz, a retired business professor from Georgia State University, who was critical of Dr. O’Brien’s methodology but who did not himself conduct a competing study. The order also discusses a number of other witnesses called by the parties, and explains that Judge Weeks generally disallowed testimony from several judges who had presided over capital trials in Cumberland County.

The order then proceeds to interpreting the RJA. Judge Weeks’ interpretation of the relevant portions of the Act is as follows:

• The Act provides for relief if the defendant can show that race was a “significant factor” during jury selection or in the decision to seek or impose the death penalty. It doesn’t define “significant factor.” The order states that significance will be interpreted to mean (1) statistically significant, in the sense that the results are less than 5% likely to be the result of chance (statisticians everywhere are probably cringing at that way of expressing statistical significance, but hey, I’m a lawyer, not a statistician), and (2) practically significant under the four-fifths rule used by the EEOC in employment discrimination cases. (“Under this basic rule-of-thumb, disparate impact will be presumed if the minority’s success rate under a challenged employment policy is equal to or less than four-fifths (80%) of the majority’s success rate.”)
• The order states that the Act provides for a burden shifting approach: “[T]o establish a prima facie case, a defendant may introduce statistical proof of unadjusted data demonstrating significant racial disparities in prosecutors’ peremptory strikes. If a defendant establishes a prima facie case that race was a significant factor, it becomes the State’s burden of production to actually rebut the defendant’s case, or to dispel the inference of discrimination, not merely advance a non-discriminatory explanation.”
• The order concludes that the RJA does not require a showing of intentional discrimination. Rather, relief is required even if prosecutors discriminated subconsciously.
• The RJA does not require that a defendant show prejudice or an impact on his particular case: “The Court likewise holds that the plain words of the RJA demonstrate the absence of any requirement to prove race was the basis of the decision to seek or impose a death sentence in a defendant’s particular case. In clear and unambiguous terms, the RJA permits showings of patterns of discrimination by county, district, division, and state.”

Next, the order makes findings of fact concerning the evidence presented. Principally, it states that the court “finds the [O’Brien study] to be a valid, highly reliable, statistical study of jury selection practices in North Carolina capital cases between 1990 and 2010. The results of the unadjusted study, with remarkable consistency across time and jurisdictions, show that race is highly correlated with strike decisions in North Carolina. The adjusted, regression results show that none of the explanations for strikes frequently proffered by prosecutors or cited in published opinions, such as death penalty views, criminal backgrounds, or employment, diminish the robust and highly consistent finding that race is significantly correlated with strike decisions in North Carolina.”

As to the “unadjusted” results, the study concluded that “prosecutors statewide struck 52.6% of eligible black venire members, compared to only 25.7% of all other eligible venire members,” and that similar disparities existed in the various smaller political subdivisions listed in the Act. The court ruled that “[b]ased solely upon the unadjusted analysis of the decisions to peremptorily challenge black venire members, the Court finds that race was a materially, practically and statistically significant factor in decisions to exercise peremptory challenges during jury selection by prosecutors” statewide and in smaller political subdivisions. The court further found that the unadjusted disparities supported an inference of intentional discrimination.

The order then discusses the efforts made in the study to control for the influence of non-racial variables that may be correlated with race, such as attitudes about the death penalty. The order undertakes a rather technical discussion of cross-tabulation and regression analysis, but the bottom line is that the study concludes that non-racial factors cannot explain the disparity in jury selection procedures, and Judge Weeks agreed that the study properly controlled for such factors. He generally rejected the state’s argument that the study did not include all the relevant variables, and that the regression analysis could be run differently.

The order also addresses several alleged errors that the state identified in coding the data included in the study. Judge Weeks agreed with the defense that most of the alleged errors were actually not errors, and that correcting the genuine errors did not change the results.

The court also rejected the state’s suggestion that any discrimination by prosecutors against black prospective jurors was counterbalanced by discrimination by defense attorneys against white prospective jurors. The court ruled that even if the juries actually seated were representative of the jury pools from which they came, discrimination by the state violates the RJA. In an interesting aside, the court suggested that any racial discrimination against white prospective jurors by defense attorneys might actually provide another basis for a defendant to seek relief under the Act.

The court also considered nonstatistical evidence of discrimination presented by the defense, including evidence that racial discrimination has a long historical legacy in the state; that people generally are susceptible to unconscious biases; that, in certain specific instances, prosecutors struck jurors without a clear non-racial reason; and that, in the court’s view, prosecutors had not conducted appropriate trainings regarding how to conduct jury selection in a non-discriminatory manner.

Summing up, the court found evidence of racial discrimination in jury selection statewide and at the level of the judicial division and the county. (Cumberland County is its own prosecutorial district.) Further, although the court stated that such findings were not required under the Act, it found that such discrimination was intentional and that it occurred in Robinson’s case. Based on those findings, the court ordered that Robinson be re-sentenced to life without parole.

The state has announced that it plans to appeal. Let me end by noting that I don’t think that this decision has any binding force with regard to the 150 or so other inmates who have filed claims under the Act. Even though the vast majority of the court’s conclusions aren’t specific to Robinson’s case, neither res judicata nor collateral estoppel allow those conclusions to control the outcome in other cases, because each case has a different defendant and so the element of identity of parties is not present. (Update: a helpful reader pointed me to Thomas M. McInnis & Assoc. v. Hall, 318 N.C. 421 (1986), and Rymer v. Estate of Sorrells by and Through Sorrells, 127 N.C.App. 266 (1997), which indicate that the requirement of identity of parties has been removed as to defensive and perhaps at least some offensive uses of collateral estoppel. So there may be room to argue that the ruling in the Robinson case has a broader impact. It should be obvious by now that I’m not the world’s leading expert on collateral estoppel, so I won’t offer an opinion on the merits of such an argument.) Of course, depending on how it is written, an opinion by the appellate courts could have a statewide effect. Stay tuned.