On “Coercive” Plea Bargaining

In his response to Barry Latzer’s An Incarceration Nation?, the Cato Institute’s Clark Neily argues that a significant contributor to over-incarceration is the number of improperly convicted individuals who were coerced into plea bargains. This argument, which Mr. Neily made with great rhetorical force, is one his writings have prompted me to explore. The issues he raises are important and worth engaging when writing and thinking about criminal justice policy. It is in that spirit which I, as the one respondent in agreement with Mr. Latzer, offer for consideration the following thoughts not raised in Latzer’s reply.

In making his argument, Neily points to the existence of a “trial penalty,” which he defines as “the differential between the time a defendant will serve if he agrees to plead guilty and the time he will serve if he exercises his right to trial and loses.” But such a differential doesn’t necessarily connote a “penalty” (as opposed to, say, a “discount”). The phrase “trial penalty” is a loaded one, which argues something specific: that the sentencing differential between what’s handed down after a plea bargain and what’s handed down after a guilty verdict at trial is vast enough to regularly coerce criminal defendants—even factually innocent ones—to plead guilty.

In making his argument, he points to a report published by the National Association of Criminal Defense Lawyers (NACDL) purporting to show that “those differentials can be staggering, with those convicted after a trial serving sentences two to three times longer than those who pleaded guilty.” But, the analytical value of the NACDL report is severely limited by its exclusive focus on federal data, as well as its lack of any meaningful controls to isolate the role that factors like criminal history play in explaining the sentence disparities its authors decry. Indeed, the report’s authors acknowledge that their analysis “does not take into account every factor in each individual case that may have led to a higher sentence.”

More precise analyses of this issue indicate—at the very least—that the empirical evidence regarding whether the data Neily presents reflects a “trial penalty” rather than a “sentence discount” is more mixed than one might guess after reading his essay. This is the first of three critiques I’ll make below. The second is that Neily’s argument doesn’t seem to leave room for other compelling-yet-benign explanations for the ubiquity of guilty pleas in modern American criminal prosecutions; and the third is that it’s far from clear that a reduction in plea bargaining wouldn’t be offset by the impact of longer sentences on incarceration rates.

In assessing whether and to what extent plea bargaining in the United States is by and large a coercive process, it might be instructive to look to some of the work done by David S. Abrams, whose empirical analyses of plea bargaining have found, for example, that the dominant “view of the trial penalty… is completely at odds with economic prediction.” In a 2011 study of plea bargaining in Cook County, Illinois, Abrams found that “a risk‐neutral defendant seeking to minimize his or her expected sentence would do substantially better by rejecting a plea bargain.” For those inclined to simply dismiss Abrams as a scholar whose studies are geared toward a predetermined result, I would note that in one assessment of plea bargaining in North Carolina, Abrams did find a trial penalty of about six months—though that finding was complicated by the state’s structured sentencing regime.

Abrams—whose work is not without its detractors—isn’t the only scholar whose findings run counter to the standard story regarding plea bargaining in the United States. A study published in Crime & Delinquency in 2018 found that “Defendants [in NYC] charged with more serious crimes, persons crimes, crimes with victims, and represented by private attorneys were more likely to go to trial than to be pleaded out.” That finding squares with a 2019 study, published in Legal and Criminological Psychology, which “suggest[s] that, given similar discounts, the likelihood for pleading guilty may be higher for defendants facing less serious charges that carry less severe sentences.” In other words, the prospect of a long prison sentence may not weigh on defendants as heavily as one might assume at the bargaining stage, leading many individuals facing long sentences to take their chances in court.

To be sure, whether modern plea bargaining has reached a point at which factually innocent defendants are regularly induced to plead guilty to crimes they didn’t commit is a serious question—one with which policy analysts ought to be genuinely concerned. But a recent study published by the University of Utah’s Paul Cassell provides reason to believe that wrongful convictions are extremely rare. Rarer still, per Cassell’s study, are wrongful convictions resulting from guilty pleas as opposed to trials. Yet there’s no getting around the fact that plea bargaining has become the overwhelmingly dominant resolution mechanism of criminal cases.

This takes me to the basis for my second critique, which is that the increased rate of plea bargaining in the United States might also reflect the impact of technological advancements on the likelihood of a guilty verdict. For example, it is now estimated that 80 percent of crimes involve video evidence, which is largely a function of the increase in the number of both public and private security cameras in use, as well as the increase in the use of body-worn cameras (BWCs) (which, by 2016, were used by nearly half of all law enforcement agencies in the U.S.). Interestingly, a meta-analysis of the literature on BWCs noted a study finding that 93% of prosecutors reported using BWC footage “primarily to prosecute citizens.” That meta-analysis also noted empirical support for the proposition that the presence of BWC footage increases the chances of guilty pleas (in addition to guilty verdicts, and the filing of criminal charges).

Beyond cameras, a wide range of evidence attributable to technological advances—think DNA analysis, facial recognition, cell site location data, computer databases, and wire taps—is now available to prosecutors. That sort of evidence strengthens cases in ways that would not have been possible back when trials were more common. The presence of such evidence makes prosecutors more likely to bring charges and make more punitive charge offers at the bargaining stage.

Finally, whether a substantial reduction in plea bargains would actually translate to a substantial reduction in America’s incarceration rate is also unclear. A 2017 paper found that though “more than 20 percent of defendants who currently receive prison sentences would be acquitted” if plea bargaining were banned, “the aggregate time convicts would spend behind bars would increase substantially.” The author of that paper concluded that “without the opportunity to bargain for a plea deal, defendants would face the prospect of much longer sentences,” such that “the increase on the assigned sentences would more than offset the lower probability of a conviction.”

To be sure, none of this is to say that there isn’t some subset of America’s prison population whose incarcerations were the products of coercive plea bargains. Surely, there is. But, it’s far from clear that such outcomes represent the center of the distribution of criminal cases prosecuted in the United States, let alone that they drive the U.S. incarceration rate. Whether the U.S. locks up too many people may be a complicated question; but whether America’s incarceration rate is driven by coercive plea bargaining is one I think we’re much farther from being able to answer.