Subtle Bias in Legal Decision Making: How Attitudes and Social Norms Affect Primary and Peripheral Targets

Before the 1990s controlled research using mock jurors consistently found black defendants guilty more often than white. However, more recently, research has generally failed to find this effect. One explanation is that prejudice has been reduced so much that there is no longer an effect. While there does seem to have been a reduction in overt prejudice, it is unlikely that it has decreased to the point that it does not affect verdicts. A more likely explanation is that strong social norms exist concerning prejudice which result in efforts to avoid being (or appearing) biased. Thus, when motivation to reduce prejudice is salient, mock jurors and perhaps real jurors will display little or no prejudice; but when motivation to reduce prejudice is not salient, decision-making becomes spontaneous and whatever prejudice does exist will affect decisions. In a series of 6 studies, race of defendant, race of witness, and the salience of the importance of being unbiased were varied. Results revealed a complex situation with many factors playing a part. Race of key alibi witness played a key role, with the white witness favoured, and the black witness mistrusted. Outcomes may be partially predicted based on Social Dominance Orientation (SDO). Those high in SDO treat incongruent defendant/witness race pairs more harshly than congruent race pairs. Modern apparently egalitarian outcomes are perhaps due to low prejudice participants' bias in favour of black defendants while high prejudice participants were biased against black defendants -- effectively cancelling out each others' verdicts. Those low in SDO treat incongruent defendant/witness race pair too leniently as compared to congruent race pairs. When race is made salient, bias is reduced, and though the average results are still essentially egalitarian, these verdicts are more truly egalitarian -- much fairer and less biased when considered at an individual level. These results also reinforce theories of dual process models of attitudes. Individuals may have common stereotype knowledge, but separate activation based on prejudice levels. Both high and low prejudiced individuals can control bias with the proper motivation.