Neuroscience and institutional choice in federal sentencing law.

• Author: Krauss, Rebecca

Advances in functional Magnetic Resonance Imaging (fMRI) have shed light on how the human brain operates in different decisionmaking contexts, including legal ones. (1) In roughly the past five years, legal scholars have begun incorporating fMRI evidence into scholarship on criminal responsibility, evidence, health law, employment law, and various other fields. (2) In this Comment, I consider the implications of a recent fMRI study, The Neural Correlates of Third-Party Punishment, (3) for sentencing law and institutional choice. One neuroscience study will never resolve the extended debate on federal sentencing, but this Comment uses The Neural Correlates of Third-Party Punishment as a jumping-off point to demonstrate that evidence from neuroscience is relevant--to criminal sentencing specifically and to institutional choice analysis generally.

In The Neural Correlates of Third-Parry Punishment, Joshua Buckholtz et al. ("Buckholtz") describe the results of a recent neuroimaging study in which subjects were scanned while evaluating culpability and determining appropriate punishments in hypothetical criminal cases. Buckholtz found that different brain regions were involved in these two tasks: subjects used their right dorsolateral prefrontal cortices to determine culpability and their socio-affective brain networks (particularly the right amygdala) to assign appropriate sentences. (4) In plainer English, this finding means that subjects used brain regions associated with impersonal judgment and classically cognitive tasks to make culpability determinations, but they used brain regions associated with social-emotional processing and rapid stimulus response to make punishment decisions. (5) In more legal terms, then, the Buckholtz study has found something akin to a neural correlate for the bifurcation of the criminal trial into culpability and sentencing phases. (6)

Part I of this Comment explains the significance of these results and shows how the Buckholtz study relates to other scholarship on cognitive function and legal decisionmaking. Part II points to real-world correlates of this study in the history and characteristics of federal sentencing. Finally, Part III suggests how this evidence might bear on federal sentencing law, and particularly on institutional choice within sentencing. Specifically, the Buckholtz study, viewed in conjunction with other research, raises the possibility that criminal sentencing is a quintessentially legal realist domain in which "hunches" and intuition determine outcomes. (7) If this is the case, then judges may not have specialized sentencing expertise, and we should reconsider what each institutional actor--judges, juries, Congress, and the Sentencing Commission--can contribute to sentencing decisions. The Sentencing Commission and its Guidelines, for example, can be understood as a means of confining intuitive and nonrational sentencing decisions to one decisionmaking body and then applying those decisions equally to all criminal defendants. Section III.B considers the merits of the current advisory Guidelines system when viewed in this light and suggests that the advisory Guidelines may represent a compromise between arbitrary rule-based sentencing and purely intuition-driven sentencing. Section III.B also argues that the Guidelines could execute this compromise more effectively by incorporating empirical evidence about the effects of criminal punishment, rather than merely replacing judicial intuitions with the Commission's intuitions.

1. THE BUCKHOLTZ STUDY IN CONTEXT

   The idea that cognitive processes can be divided into two basic categories--"intuition" and "reason"--is an ancient one. Today, cognitive scientists widely embrace this "dual-process theory" of brain functions, distinguishing quick and associative cognitive processes from slower, more reflective ones. (8) Keith Stanovich and Richard West coined the phrases "System 1" and "System 2" to describe these collections of processes, a convention adopted in this Comment. (9) System 1 processes tend to be automatic, effortless, associative, rapid, and opaque; System 2 processes are generally controlled, effortful, deductive, slow, self-aware, and rule-following. (10) Daniel Kahneman and Shane Frederick have developed a decisionmaking model that describes how the two systems interact: "System 1 quickly proposes intuitive answers to judgment problems as they arise, and System 2 monitors the quality of these proposals, which it may endorse, correct, or override. The judgments that are eventually expressed are called intuitive if they retain the hypothesized initial proposal without much modification." (11) Chris Guthrie, Jeffrey Rachlinksi, and Andrew Wistrich have applied this model to judicial decisionmaking, calling it an "intuitive-override" model of judging. (12)

   The Buckholtz study is best understood in the context of this dichotomy. Using fMRI, Buckholtz scanned sixteen participants while they evaluated culpability and determined appropriate punishments in a series of fifty written scenarios. (13) The scenarios varied in the gravity of harm committed by the would-be defendant and in the would-be defendant's level of culpability; in some cases, participants were aware that mitigating circumstances excused or justified the would-be defendant's behavior. (14) Subjects rated each scenario on a scale from zero to nine, with zero indicating no punishment and nine indicating extreme punishment. (15) Buckholtz found that the right dorsolateral prefrontal cortex (rDLPFC) was differentially activated based on the would-be defendant's culpability, with greater activation levels corresponding to more culpability. (16) This result was expected, since the rDLPFC has previously been associated with impersonal moral reasoning and cognition. (17) The rDLPFC activity did not, however, track the magnitude of punishment assigned. (18) Instead, there was a linear relationship between punishment magnitude and activation in brain regions that have been "extensively linked to social and affective processing," particularly the amygdala. (19) The amygdala is an evolutionarily older part of the brain that is implicated in emotional processing and immediate stimulus response. (20) In the terms of dual-process theory, the Buckholtz study suggests that assigning criminal punishment is an intuitive, System 1-based decision, unchecked by System 2 processes.

   Cass Sunstein and his coauthors have described a behavioral phenomenon that corroborates this idea. Based on "mock-juror" experiments, Sunstein et al. ("Sunstein") argue that punishment decisions are complicated by a "translation problem"--a "distinctive problem involved in translating a moral judgment of some kind into the terms made relevant by the legal system, such as monetary penalties, civil fines, or criminal punishment."...