News Release

U study provides new insight toward reducing racial bias in courtroom

Peer-Reviewed Publication

University of Minnesota

The American criminal justice system relies on jurors to regularly decode the mental states of criminal defendants. When those defendants are people of color, decoding minority mental states is a centerpiece of the justice process.

A new study released by University of Minnesota Law School Professor Francis Shen finds that it may be possible to make these "minority mens rea" determinations (mens rea is a legal term referring to criminal intent) without significant racial bias.

The study, titled "Minority Mens Rea" and published in the Hastings Law Journal, offers positive news for a criminal justice system that has become keenly aware of the need for improved responses to race biases. Ultimately, Shen hopes further research will help the legal system better understand how implicit racial biases lead, or don't lead, to unjust outcomes.

"The rush to tackle implicit bias is based on a presumption that these implicit biases are contributing, perhaps dramatically, to racial injustice in criminal outcomes," Shen said. "Do implicit racial biases exist? Yes. But must they affect every decision in the justice system? My results suggest that the answer may be no."

Shen's research presents the first empirical investigation of how jury-eligible subjects engage in assessing the mens rea, or criminal intent, of minority defendants, and builds on the 2003 economics field study, "Are Emily and Greg More Employable Than Lakisha and Jamal?" In that study, researchers from Harvard and the University of Chicago sent out resumes to potential employers. While keeping the substance of the resumes the same, the researchers randomly changed the name (e.g. replacing John with Jamal). Their findings were striking: the fictional Lakisha and Jamal were less likely to receive interviews than their white counterparts.

Shen's study, conducted through 16 separate online experiments with more than 1,200 subjects and evaluation of 2,400 unique vignettes, explored whether subjects assessed the criminal mental states of "John," "Jamal," "Emily," and "Lakisha" differently. Specifically, subjects read a vignette and determined whether its protagonist acted Purposely, Knowingly, Recklessly, or Negligently. These four mental states, defined by the Model Penal Code, can lead to very different penal outcomes. Thus, if black defendants are systematically deemed more culpable than white counterparts, it would be of great concern.

However, the study instead found that assessments of minority mens rea were not biased by race. On the Model Penal Code's hierarchy of mental states, higher levels of culpability were no more likely to be misattributed to "Lakisha" and "Jamal" than to their white counterparts. In other words, implicit biases did not affect mens rea determinations.

To explain these results, Shen draws on psychology and cognitive neuroscience. He argues the most compelling explanation is the nature of the experimental task subjects were asked to perform. It might have served as a distraction task, he says, focusing subjects on the cognitively intense work of determining mental states and thereby diverting their attention from the perceived race of the protagonist.

"The distracted brain may sometimes be a less racially biased brain," Shen said.

Despite his findings, Shen cautions that it is premature to draw policy implications from a single study, which relied only on the manipulation of names.

"Until we complete more ecologically valid mock jury studies, we can only speculate," Shen said. "Still, the research suggests that further research in this domain could lead to new insights about how to reduce racial bias in criminal justice outcomes."

###

The research was supported by the University of Minnesota and by the MacArthur Foundation Research Network on Law and Neuroscience at Vanderbilt Law School.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.