Confronting cognitive "anchoring effect" and "blind spot" biases in federal sentencing: a modest solution for reforming a fundamental flaw.

• Author: Bennett, Mark W.

TABLE OF CONTENTS INTRODUCTION I. THE POWER AND ROBUSTNESS OF THE "ANCHORING EFFECT" A. Background B. The Cognitive "Anchoring Effect" Studies II. JUDGES AND THE ANCHORING EFFECT A. The German Judges Studies B. The American Judges Studies 1. The U.S. Magistrate Judges Study 2. The U.S. Bankruptcy Judges Study 3. One Final Anchoring Study--Information Obtained in Settlement Conferences 4. Summary of Cognitive "Anchoring Effect" Studies III. THE FEDERAL SENTENCING REVOLUTION A. Brief Overview of Federal Sentencing B. The Booker Revolution C. The Post-Booker Sentencing Regime D. The Importance of Gall v. United States E. The Pre-sentencing and Sentencing Process IV. POST-BOOKER SENTENCING AND THE GRAVITATIONAL PULL OF THE GUIDELINES RANGE V. ANCHORING AND FEDERAL SENTENCING VI. A MODEST PROPOSAL CONCLUSION INTRODUCTION

"God not only plays dice. He also sometimes throws the dice where they cannot be seen."

--Stephen William Hawking (1)

Trial judges, too, roll dice in sentencing. They just do it unwittingly. Like God's dice roll in Hawking's quote above judges' dice rolls are never seen--except in one startling series of studies establishing that the actual number rolled on the dice, when disclosed to the judges, affected the length of sentences they gave! For state and federal judges who sentence pursuant to advisory guidelines, there are potent psychological heuristics at play. "Psychologists have learned that human beings rely on mental shortcuts .. . 'heuristics,' to make complex decisions. Reliance on these heuristics ... can also produce systematic errors in judgment. ... [C]ertain fact patterns can fool people's judgment, leading them to believe things that are not really true." (2) These heuristics have a strong potential to affect the length of sentences. Whether judges consider their sentencing philosophy to be tough, lenient, or in between, to be the best judges they can be, they need to recognize and understand how these cognitive and implicit forces tend to increase judges' sentences without their conscious knowledge.

This Article explores how judges' hidden cognitive biases, specifically the "anchoring effect" and, to a lesser extent, the "bias blind spot, impact the length of sentences they impose by subconsciously influencing judges to give greater weight to the now-advisory Federal Sentencing Guidelines than to other important sentencing factors. Biologically, every mammalian eye has a scotoma in its field of vision--colloquially known as a blind spot. Everyone, including sentencing judges, has blind spots. This Article is not concerned with our scotomas, the physical blind spots of our eyes, but with their psychological corollary: the cognitive bias known as the bias blind spot." This psychological blind spot prevents us from seeing our own cognitive biases, yet allows us to see them in others. (4) This "tendency to see bias in others, while being blind to it in ourselves," means that judges impacted by the anchoring effect in sentencing are unlikely to recognize it. (5) This creates a double bind for judges. First, a lack of awareness prevents perception of the powerful and robust impact of the anchoring effect in sentencing. Second, once one becomes aware of the anchoring effect, an inability to see how the anchoring effect impacts one's own sentencing persists because of the "bias blind spot." "Moreover, to the extent that judges might consider themselves experts in the law, they are probably more confident of their abilities to disregard biases than they should be." (6)

Even more troubling, research indicates that sentencing judges are influenced by anchors, even irrelevant anchors, to the same extent as lay people and that the effects of the anchors are not reduced by the judges' actual experience. (7) Compounding this conundrum is that while more experienced judges are equally susceptible to the effects of anchoring as novices, they "feel more certain about their judgments." (8) That is why it is critically important for sentencing judges, probation officers who prepare presentence reports, and practicing lawyers to understand the potential robust and powerful anchoring effect of advisory Guidelines and the effect of the "bias blind spot" in determining just sentences.

In the last quarter century, federal sentencing has undergone enormous upheaval: from unbridled discretion to sentence as low as probation up to the statutory maximum, to the mandatory and inflexible United States Sentencing Guidelines--the grin-and-bear-it approach to sentencing (9)--to advisory Guidelines with the return of significant, but not unbridled discretion. Shockingly, given the substantial judicial displeasure and even hostility toward the Guidelines, the return of substantial discretion has not significantly altered the length of most defendants' sentences. I suggest that this is due primarily to the anchoring effect. Computing the advisory Guideline range so early in the sentencing process strongly anchors a judge's sentence to that range, or close to it. This is true even when compelling factors suggest a significantly lower or, on rare occasions, higher sentence. (10)

This Article is organized as follows. Part I comprehensively examines the anchoring effect in a variety of intriguing settings through the lens of classic cognitive anchoring studies. Part II focuses on cognitive anchoring studies in several judicial contexts that involve actual judges, including some from Germany, but mostly federal and state court judges in the United States. Together, these first Parts provide a more thorough and in-depth analysis of the robustness of the anchoring effect than any prior scholarship discussing judges and anchoring.

Part III presents an overview of the federal sentencing revolution, from the implementation of the mandatory United States Sentencing Guidelines in 1987 through the Booker (11) and Gall (12) shockwaves arising from the ApprendP upheaval leading to the now -advisory Guidelines. These advisory Guidelines restore substantial, but not unlimited, sentencing discretion. Part IV examines the statistical trends of federal sentencing, showing that the Guidelines, even in their current advisory role, continue to exert a strong gravitational pull on federal sentencing. This Part also explains that the result of this pull is that very little has changed in terms of the length of federal judges' sentences, even with their new, broad discretion. Part V argues that the most likely culprit as to why federal district court judges have remained so tethered to the Guidelines, post Booker and Gall, despite their wide dissatisfaction with them, is the anchoring effect.

Part VI offers a modest, sententious but meaningful and straightforward proposal to help reduce the undesirable anchoring effect of the Guidelines. The proposal reorders the information in the presentence report (PSR) prepared by the U.S. Probation Office in every federal sentencing. Rather than disclosing the often complex Sentencing Guidelines calculations early in the PSR (where the anchoring effect comes in), the information about the defendant's personal history and other factors that a judge must consider and may use to vary downward or upward from the Guidelines would be disclosed first. The judge could then note on the PSR a preliminary sentencing range based on everything the judge is required to consider pursuant to 18 U.S.C. [section] 3553(a) and reach a tentative sentencing range before the PSR discloses the advisory Guidelines sentencing range. This reordering would greatly help in reducing the anchoring effect of the Guidelines. The judge would first have to confront why the initial sentencing range he or she wrote down, unencumbered by the actual computed Guidelines range, was different. The judge would then decide if the gravitational pull of the Guidelines unfairly influenced his or her [section] 3553(a) analysis or vice versa. Also, other highly relevant numerical sentencing information not currently included in the PSR should be presented in the latter portions of the PSR to counteract the anchoring effect of the Guidelines. Unlike prior unrealistic proposals offered by law professors to reduce the effect of anchoring in federal sentencing, (14) this proposal requires no further action by the U.S. Supreme Court or Congress and is easily implemented by any federal district court judge that chooses to adopt this recommendation.

1. THE POWER AND ROBUSTNESS OF THE "ANCHORING EFFECT"

   1. BACKGROUND

      Virtually all judges strive to be as fair and rational as possible when sentencing. But what if there are hidden psychological processes quietly at work that undermine their best efforts to be fair? Psychologists label such processes "cognitive biases." (15) These biases--which can lead to serious mistakes in decisionmaking, judgment, and reasoning--can cause judges to hold on to certain preferences and beliefs regardless of contrary, persuasive information. (16)

      Anchoring is a cognitive bias that describes the human tendency to adjust judgments or assessments higher or lower based on previously disclosed external information--the "anchor." (17) Studies demonstrate "that decisionmakers tend to focus their attention on the anchor value and to adjust insufficiently to account for new information." (18) Cognitive psychology teaches that the anchoring effect potentially impacts a huge range of judgments people make. This includes people who have developed expertise in their fields, like experienced real estate agents, (19) auto mechanics, (20) and physicians. (21) In discussing cognitive biases among specialized experts, Jeffrey Rachlinski and his colleagues observe: "Research on some experts--including doctors, real estate agents, psychologists, auditors, lawyers, and judges--shows that they often make the same kinds of mistakes the rest of us make." (22) Amazingly, repeated studies show that the "anchor" produces an effect on judgment or...