Skip to main content

Disentangling the Influences of Defense Attorney Plea Recommendations

January 10, 2023

Quattrone Center Research Fellow Johanna Hellgren has co-authored a paper that examines how defense attorneys make plea recommendations.

Despite a constitutional right to a jury trial, between 95 and 98 percent of criminal convictions in the United States result from guilty pleas, a number that has climbed steadily in recent years. In a paper recently published in Psychology, Public Policy and Law, Johanna Hellgren, a Research Fellow at the University of Pennsylvania Carey Law School’s Quattrone Center for the Fair Administration of Justice, and Saul M. Kassin, Professor of Psychology at John Jay College of Criminal Justice in New York, examined the extent to which defense attorneys’ plea recommendations are driven by their beliefs in the defendant’s guilt or innocence and by estimates of the probability-of-conviction at trial — and implications for defendants.

“As part of responsible advocacy, criminal defense attorneys communicate plea offers, investigate the case, and advise their clients accordingly, generally based on the apparent strength of evidence against their client,” said Hellgren. “Through our research, we sought to untangle whether strong evidence in an individual case drives the recommendation to accept a plea offer because it suggests that the defendant is factually guilty or innocent, because it signals a conviction or acquittal at trial, or, due to a combination of these factors.”

Since the mid-1920s, plea bargaining — a process that incentivizes defendants to plead guilty to obtain significant “plea discounts” – has served as the standard method through which criminal cases in the United States are resolved, with only about 2–3 percent of cases resulting in a full jury trial. Despite theory and research suggesting that defendants are often advised to plead guilty “in the shadow of the trial,” a theory that hinges on the expectations of trial outcomes, Hellgren and Kassin found that attorneys base their plea recommendations largely on pragmatic considerations.

In their paper, the researchers conducted two experiments in which attorneys received a case file containing highly inculpatory or exculpatory evidence that was either admissible or not admissible at a potential future trial. Participants read a case file containing a police report; a defendant’s written confession or denial; a motion and ruling to suppress evidence (an inculpatory confession in Study 1; an exculpatory eyewitness statement in Study 2) that was granted or denied; defense counsel’s interview notes in which the defendant proclaimed innocence or not; and the terms of a plea offer. Participants then made a plea recommendation and indicated their beliefs regarding the defendant’s guilt and likely trial verdict.

Hellgren and Kassin found that while a confession increased the perceptions of the defendant’s guilt, attorneys were more likely to recommend accepting the plea offer only when that confession was ruled admissible at trial. In the second study, the exculpatory eyewitness testimony increased perceptions of innocence, but attorneys were more likely to recommend rejecting the offer in favor of trial only when that testimony was ruled admissible at trial. The defendant’s assertions of innocence to defense counsel had little effect. They concluded that the system often leaves defense attorneys with little choice but to rely on pragmatic considerations when assisting their clients and made policy recommendations related to structural reform on two fronts.

First, the high rates of the so-called “plea discount” — that is, when defendants who invoke their right to trial face a “penalty” in sentencing for doing so. For example, the National Association for Criminal Defense Lawyers using data from the U.S. Sentencing Commission (2017) found that defendants convicted at trial received sentences that were 3.27 times higher than those who pled guilty to comparable crimes. Second, they noted that defense attorneys often lack access to case file information that would enable smart assessments of factual guilt and innocence as well as the probability of conviction independent of their client.

Read the full paper.