Skip to main content

Cracking the Black Box

One way to keep prosecutors accountable and check their carceral impulses is by shedding some light on their vast discretion to charge crimes. Here’s a start.


In July, a sheriff arrested a 19-year-old woman after he witnessed her “stomping on a ‘Back the Blue’ sign” next to where he was conducting a traffic stop. Not only was the arrest a surprise, but the local prosecutor’s response made national news. The Utah prosecutor charged this teen with a criminal mischief charge with a hate crime enhancement for destroying the sign in an attempt to “intimidate law enforcement.” She now faces up to a year in jail or a fine of up to $2,500. Both the prosecutor’s decision to charge and to seek a severe penalty for this minor act grabbed headlines. The case is just one illustration of how prosecutors can use their discretion to amplify criminal injustice. 

Prosecutors have long held a position of great power in the criminal legal system. And as this power has come under increased public scrutiny, prosecutors have been criticized for their lack of transparency and oversight, for contributing to mass incarceration, for exacerbating class and racial disparities, and for destroying families and neighborhoods in their pursuit of so-called victimless crimes. Since 2015, in jurisdictions around the country, self-described “progressive prosecutors” have sought to break that mold, promising a more just criminal justice system and replacing the prevailing tough-on-crime approach with a more merciful approach towards defendants. That remains to be seen: As the United States continues to hold the title of the country with the largest incarcerated population in the world, it is far from clear whether most prosecutors are using their power for progressive aims or instead furthering injustice in American society.  

Prosecutors’ greatest power is arguably their vast discretion to decide whether American defendants will face criminal proceedings. Prosecutors decide whether to charge a crime, the number of criminal charges, the severity of the charges, the potential penalties, and applicable sentencing recommendations. Since the vast majority of cases end in plea bargains, where prosecutors use the threat of more severe sanctions for those who do not play along, judges and juries are no real check on prosecutors. Nor is the Constitution or courts, since those who plead guilty largely waive the few legal rights they may have.  

This form of unreviewable decision-making raises real concerns about the predictability and legitimacy of the criminal legal system. If prosecutors across the nation can freely decide Americans’ fates with limited restraint or oversight, how can the public be sure prosecutors are truly “doing justice,” as they like to claim they are? To answer that question, we need a way to crack open the “black box” of prosecutorial discretion to understand how prosecutors make these life-changing decisions. 

We need a way to crack open the “black box” of prosecutorial discretion to understand how prosecutors make these life-changing decisions. 

In our latest research, we attempt to do just that, by asking actual prosecutors how they would charge a specific hypothetical case and then assessing their answers. Specifically, we asked thousands of prosecutors from across the country how they would respond to an incident in which an emotionally distraught and seemingly “unstable” man loudly approached other passengers at a train station to demand their help with train fare, all while holding a knife in a nonthreatening manner. To reach the prosecutors, we combed through state bar association web pages and submitted freedom of information requests to produce an email list of 4,484 attorneys in state and local prosecutor offices. We got responses from about 540 of them, hailing from every region of the country and from different sized offices and jurisdictions. On average, they had been prosecutors for a little over 12 years. Two thirds of them were men and roughly ninety percent were white. (You can read more about the sample in our article.)

Supplemental Materials: National Prosecutor Survey

The following supplemental materials accompany the articles Inside the Black Box of Prosecutor Discretion and Race and Class: A Randomized Experiment with Prosecutors.

We asked all of the prosecutors to evaluate the following incident, written in the voice of the arresting officer:

At approximately 9:25pm, two citizens approached me to report an unstable man of moderate height and build in the 8th Avenue train station…. Once we entered the station, I immediately identified the unstable man, Johnson. Johnson was yelling obscenities, including the F-word, and banging on station turnstiles with the bottom of a buck knife. He appeared angry and slightly intoxicated. ….

According to witnesses, Johnson entered the station around 9:10pm and began begging for change to purchase a train ride. After repeated denials from patrons, Johnson started shouting obscenities, including “F**k all you!” and “F**k this damn city!” At one point, Johnson reportedly grabbed a woman’s arm when she refused his repeated requests for money. Although Johnson did not threaten the woman, he did dangle a knife by his side with his other hand. After interviewing approximately six witnesses, I concluded that nobody was physically injured from the incident, but that most all who were present became very afraid once Johnson started yelling with his knife in his hand.

When Johnson noticed our presence, he became noticeably less incensed and ceased yelling and banging on the turnstiles. He explained to us that his girlfriend had just broken up with him that afternoon, and he did not have money for the train…. After hearing Johnson out for a few minutes, I directed him to slide the knife in our direction and submit to an arrest. He complied without protest. 

In addition to the police reports, the prosecutors were also given a hypothetical sentencing scheme that we copied from a mid-sized jurisdiction, along with a list of potential charges that they could consider filing.      

The charges they chose were surprising. Prosecutors simply charged more crimes than we expected to our fictional defendant — without regard to whether he was Black or white, rich or poor, or some other status or identifier.

Note for starters that we intentionally designed a fact pattern in which the man’s behavior, while frightening to a large number of people, was not clearly criminal. With the possible exception of grabbing a woman’s arm (which the reports expressly described as nonthreatening) the police reports did not indicate that the man actually assaulted anyone, nor do they indicate that his possession of the knife was unlawful. Shouting and cursing in public are generally not crimes. On the other hand, the fact pattern is one in which many observers would see probable cause for an arrest — and thus the possibility of filing charges that would not be summarily dismissed in court. There are also facts that allow either an aggravated assault (felony) or a simple assault (misdemeanor). In short, the reports create a genuine zone of prosecutorial discretion.   

Within that zone, notice not only how varied prosecutors’ decisions were for the number of charges, but also how many prosecutors made the decision to bring charges at all.

If the prosecutors in our study were left to their own devices, they would like to charge a crime in every instance. Almost every prosecutor brought at least one charge against this defendant, with only 3% declining to charge him with any crime. These findings are in line with national research showing that prosecutors are continuing to charge great numbers of crimes, despite the fact that crime rates have fallen nationwide, and arrest rates have dropped too.

Then there is the question of sentencing. While most opted for misdemeanor charges or a suspended sentence, 16 percent of prosecutors selected a felony charge for the disruption, despite the fact that the man had no criminal history, and no people were injured. As for recommended sentences, while over 70 percent of participants recommended no term of confinement for the man in the hypothetical, the remainder recommended jail time, generally for a duration less than 30 days, though some recommended up to two years. Recommendations for monetary sanctions were similar: A substantial minority of prosecutors (roughly 40 percent) sought a monetary penalty, typically for an amount of less than $500, though some prosecutors sought a penalty as high as $5,000

Finally, there is the question of the number of charges the prosecutors filed. As scholars have noted, this practice of “charge stacking” is a significant driver of plea bargaining, as defendants facing a large number of charges for the same basic incident are more likely to plead guilty or to be convicted at trial. The decision of how many charges to file, in other words, is indirectly a decision about the outcome of the case. And here, too, our survey findings are significant: Almost 80 percent of prosecutors chose to bring multiple charges with an average of about three charges against the man. A few prosecutors chose to seek seven or more. 

Taking together these various findings, the overall severity of the prosecutors’ charging decisions is concerning. With about six out of 10 Americans without $500 in savings — the most commonly imposed penalty in this case — the implications can be devastating for the vast majority of criminal defendants who do not have the resources to pay off these fines. Additionally, with 30 percent of prosecutors suggesting jail time, criminal defendants would also face serious consequences beyond the confinement itself. Missing even a few days or more of work could cause the individual to lose his job, stable housing, or family support despite having only committed a minor crime as a first-time offender. Even having a criminal charge placed on an individual’s record can be devastating to a defendant. And almost every prosecutor in our study decided to charge a crime, in this case where no victim was physically harmed. We recognize, of course, that our hypothetical man certainly could have caused emotional harm to the individuals waiting for the train that day, and ideally prosecutors could recommend counseling services for any lasting trauma caused by this incident.

We asked the prosecutors why they decided to file charges as they did, and here, too, their responses were illuminating. Most said they considered the necessity of punishment despite this being, at most, a minor offense. They considered the financial state of the offender, mental health, using jail as a learning opportunity, and using initial charging as a strategy for plea bargaining. They also claimed to have considered the severity of a victim’s injury, harm caused, weapons used, the number and age of victims, prior criminal history, and involvement or use of drugs. These factors, however, were perhaps most notable for how they were invoked — to support filing charges even though the defendant caused no physical injury, had no prior criminal record, and seemed to be experiencing mental health issues, all facts that should have weighed against the severe charging decisions we observed.

Beyond the severity, the vast differences in treatment are also concerning. We learned that prosecutors do not reach the same conclusions about charges to bring or penalties to seek, even when given the same facts and the same law. Depending on the specific line prosecutor assigned to a case, criminal defendants could face two years in prison, six months in jail, or no confinement at all for the exact same conduct, with similarly wide ranges for monetary penalties. 

It could be that variability in these decisions may be tied to lack of formal guidelines for prosecutors in this space. Participants reported that only about half of their offices had guidelines to help them make charging decisions. For those that did note the presence of guidelines, many stated that the standards were not written (or were confidential). Less than 10 percent of prosecutors reported mandatory guidelines. 

Another way to potentially achieve uniformity is through supervision within a prosecutors’ office. But nearly three-quarters of the prosecutors said that they made these decisions on their own as opposed to with the help of a supervisor. 

Prosecutors have utmost discretion to decide whether and how to charge the very same case — with unfairly variable, and harsh, outcomes in the mine-run of charging decisions.

In short, it seems evident that these prosecutors have utmost discretion to decide whether and how to charge the very same case — with unfairly variable, and harsh, outcomes in the mine-run of charging decisions.

Our foray into the black box of prosecutor discretion gives us distressing insights into the reality of charging practices nationwide. While this study provides awareness about severity and variability in prosecutor decision-making, more work is needed to study the charging practices of prosecutors, their power to use bail as a punitive measure, cultural and structural features of their offices that may be resistant to change, and how individual prosecutors might be stalling any broader progress towards ending mass incarceration. What we know now is that simply providing prosecutors with charging guidelines and expecting them to reform themselves by reducing their charging practices is impractical at best — and irresponsible at worst. Left to their own devices, prosecutors have an outsized role in sustaining mass incarceration — and thus the public needs mechanisms for holding them to account for reducing and eliminating the charging of minor crimes. As the man in our hypothetical illustrates, alternative resolution is often appropriate. 

Our study also left us somewhat skeptical of the progressive prosecutor movement — and whether it will have large-scale influence on the everyday charging practices of prosecutors far removed from cities or those in smaller jurisdictions. Our findings suggest that, when given a choice, almost every prosecutor chooses to charge multiple offenses against individuals with no criminal record in situations where they are engaging in disruptive behavior that need not be treated as criminal. That’s dispiriting. And yet there are some signs for hope: Most prosecutors in our study looked past the presence of a knife and the reaction of concerned passersby and opted to seek a noncarceral sentence for an incident that some might rush to label a “crime of violence.” This suggests that, perhaps, slowly, the Overton window is beginning to shift. To shift it further, and check prosecutors’ discretion ever more, there is a lot more work to be done.

Image: Unsplash