Dan Simon, Doug Stenstrom and Stephen J. Read (USC Gould School of Law, USC Department of Psychology , University of Southern California and University of Southern California – Department of Psychology) have posted Partisanship and Prosecutorial Decision Making: An Experiment on SSRN. Here is the abstract:
One of the distinct features of the American criminal justice system is the central and powerful role played by prosecutors. Many scholars agree today that the criminal justice system has become more administrative than adjudicative in nature, in that more than ninety percent of felonies and an even greater number of misdemeanors are settled by means of plea bargaining. That means that prosecutors, rather than judges and juries, dispense the lion share of criminal convictions. Naturally, the terms of the plea are strongly determined by the prosecutor’s assessment of the evidence in the case.
This study tests simulated prosecutors’ assessments of the facts in a quasi-criminal case. The study is part of a larger research project that examines the effects of the adversarial system on the outcomes of the criminal justice process. A prior study (presented at CELS, 2008) demonstrated that the mere assignment of participants to adversarial roles distorts and polarizes their views of the case (as compared to one another, and vis-a-vis a neutral assignment). The current study tests whether one’s view of the case is affected also by situational variables which can be characterized partisanship.
The study assigned participants to the role of a prosecutor in an academic disciplinary procedure, and compared groups that were given information that evoked either low or high degrees of partisanship. The findings indicate that the manipulation had a strong effect on the perceptions of the evidence as well as on a host of other variables that could affect a prosecutor’s judgments of the case.
Second, the study was designed to test the specific role of anger in the making of a prosecutorial decision. Normatively, judgments of culpability in such cases ought to be based entirely on a factual determination: did the suspect commit the crime, or not. Emotions, particularly anger, ought to play no role. Prior research has shown that the arousal of anger can have strong implications for determinations of guilt and punishment, in that anger increases attributions of blame, lowers thresholds of evidence, and increases vengefulness and punitiveness. We found that the assignment did have an effect on participants’ anger towards the suspect. Mediational analysis and Structural Equation Modeling revealed that while anger mediated the effect of the assignment on the verdict and the interpretation of the facts, it was also mediated by them in return. This bi-directionality is consistent with coherence based reasoning and the underlying mechanisms of parallel constraint satisfaction.
The study was designed also with basic-psychological objectives in mind. First it seeks to replicate previous work on coherence based reasoning. More importantly, it was designed to explore a novel question concerning the underlying cognitive mechanism by which 'hot' and 'cold' cognitions interact. Though a considerable body of past research has demonstrated that the two types of cognition affect one another, there is scant evidence pointing to the mechanism by which this occurs. Previous work on coherence based reasoning was limited to cold cognitions: primarily decision making based on judgments of facts. This study provides experimental support for the conclusion that hot cognitions too are incorporated in coherence based reasoning. Both types of cognitions are driven by Gestaltian forces (specifically, parallel constraint satisfaction processing) towards a point of equilibrium, at which all cognitions involved—the assessments of the facts, overall judgments, motivation, affect, and emotions – come together at a point of maximum coherence.