An integrated model of prosecutor decision-making

Abstract

Sociolegal scholarship has long noted the many ways in which the law is interpreted and selectively applied by human decision-makers. Yet, the processes underlying one of the most significant discretionary waypoints in the criminal legal process – prosecutorial charging decisions – remain opaque. Using data from interviews and focus groups with prosecutors in three midsized jurisdictions, we propose a model of charging that integrates legal considerations, social identity, and organizational constraints. We find that felony prosecutors weave together legal and extralegal factors, often relying heavily on criminal history, to evaluate defendants’ moral character. Based on their evaluation of a defendant’s character, prosecutors charge strategically to secure a final disposition and sentence they view as appropriate for the defendant. Prosecutors’ identities and experiences act as lenses through which they interpret case facts in their evaluation of defendants’ character. However, the level of discretion provided by their chief prosecutor and the culture of the court community in which they work condition the process by which prosecutors achieve their desired outcomes for cases.

Sociolegalscholars have long noted the myriad ways in which the law is not enacted as conceived or intended in both criminal and civil contexts (Croyle 1979; Feeley 1983; Gould and Barclay 2012; McEwen and Maiman 1984; Muir 1967; Wald 1967). At the heart of this gap between the law-on-the-books and the law-in-action is the reality that the law is interpreted and differentially applied by individual and organizational actors – human decision-makers with agency (Pound 1910; Ulmer 2019). Across American society, there are numerous examples of routine nonenforcement of some laws (e.g., certain labor laws or laws on adultery) and selective nonenforcement of others (e.g., loitering or drug possession) (Calavita 2010; Western 2006). One of the most prominent arenas showcasing the importance of discretion is criminal prosecutions. Routine and selective nonenforcement in the context of prosecution emerged as a contentious and politicized topic in the early 2020s. At its core, this controversy is about discretion and a political battle over who interprets and selectively applies the law. It is perhaps best illustrated in Florida where Governor Ron DeSantis suspended the twice-elected, reform-minded Hillsboro State Attorney, Andrew Warren, over Mr. Warren’s, alleged, blanket nonprosecution policies. DeSantis argued that Mr. Warren’s comments about not enforcing current or future laws criminalizing abortion or transgender health care represented a “neglect of duty” (Lisciandrello 2023). Mr. Warren appealed his suspension, arguing his removal was unconstitutional, and noting that his office prioritized discretion, not mandatory charging policies (Lisciandrello 2023; Warren v. DeSantis 2024). This controversy over the extent of prosecutorial power is so widespread in fact that the American Bar Association launched a task force on prosecutorial independence to educate the public on prosecutors’ role in ensuring the integrity of the justice system and the importance of prosecutorial independence (American Bar Association 2024; Kanu 2023).

In 1971, long before the current controversy on prosecutorial discretion, the American Bar Association emphasized the importance of discretion, stating, “[t]he public interest is best served and evenhanded justice best dispensed not by a mechanical application of the ‘letter of the law’ but by a flexible and individualized application of its norms through the exercise of the trained discretion of the prosecutor as an administrator of justice” (as quoted by Mellon et al. 1981). Yet, the decision-making processes that underly some of prosecutors’ most consequential decisions – charging and plea-bargaining – remain opaque. As Calavita (2010) writes, “the law-on-the-books is almost always ambiguous, and this commodious quality of the law is exploited to construct legal meanings consistent with ideological, institutional, economic, or practical agendas” (114).

In this study, we seek to illuminate how human decision-makers – in this case, prosecutors – interpret and apply the law on the books. Drawing on extensive interviews and focus groups with prosecutors from three jurisdictions across the United States, we develop an integrated model of prosecutorial charging decisions that speaks to the aforementioned ideological, institutional, and practical considerations that influence prosecutors’ application of the law. We find that prosecutors navigate a web of intersecting incentives and pressures by engaging in a process of character construction – making moral judgements about who defendants are, what motivates them, and their ultimate culpability. Our model weaves together the multiple, intersecting factors at play when prosecutors must decide how to translate the law on the books into workable charging decisions.

Background

When pressed about their decision to prosecute or decline a case, prosecutors often recite some version of the line, “I am simply following the law.”¹ This adage downplays the role of prosecutorial discretion when, according to the U.S. Supreme Court, “[s]o long as the prosecutor has probable cause to believe that the accused committed an offense defined by statute, the decision whether or not to prosecute, and what charge to file or bring before a grand jury generally rests entirely in his discretion” (quoted in Spohn 2014: 169). A key aspect of prosecutorial discretion is the authority to decline to prosecute (Culp 1969). Even after prosecutors exercise their discretion to bring a case, their powers remain broad. Prosecutors select both the charge and the number of counts, which act as an anchor for plea negotiations. Since over 95 percent of convictions in the United States are the result of a plea agreement, rather than a trial, prosecutors play a direct role in determining the severity of the outcomes defendants will face (Reaves 2013).² The “shadow of the trial” model suggests that a prosecutor’s decision to offer a plea, and how much of a discount to offer in exchange for a guilty plea, is determined by the strength of the evidence (Bushway et al. 2014).

Contrary to both “the shadow of the trial” model and the claim that prosecutors simply apply the law, there is a large body of sociolegal scholarship suggesting that legal and state actors interpret laws and policies in highly contextualized ways (Calavita 2010; Maynard-Moody and Musheno 2003; Ulmer and Kramer 1996). Research in the criminal court context has focused on the difference between legal statutes and legal actors’ application of them (e.g., Bontrager et al. 2013; Wu 2016; Mitchell 2005). Moreover, scholars emphasize that criminal court operations do not reflect a purely objective application of the law but instead prioritize efficient case processing (Feeley 1979; Sudnow 1965) and the rapid categorization of defendants, often based on stereotypical notions of race and class (e.g., Clair 2020; Gonzalez Van Cleve 2016; Kohler-Haussman 2018). If, as sociolegal scholarship suggests, law’s power is partially determined by the actors who implement it, it is especially important to understand prosecutorial discretion because state prosecutors have a direct and substantial role in shaping the criminal justice landscape (Davis 2007; Pfaff 2017; Wright 2017: 397).

Theoretical frameworks for understanding prosecutorial discretion

Scholars from a variety of disciplines, including law, criminology, and sociology have studied prosecutorial discretion. From these bodies of research, a variety of disjointed predictor variables emerge at different levels of analysis to explain prosecutorial discretion at the charging stage. These include the following: assessments of defendants’ blameworthiness, culpability and moral character, prosecutors’ social identities, and the organizational context of courts (Galvin and Ulmer 2022; Levine and Wright 2012; Lowrey-Kinberg et al. 2022; Lynch 2019; Spohn et al. 2001; Steffensmeier et al. 1998; Ulmer 2019). We describe each of these theoretical approaches in turn, and then use them as building blocks for an integrated model of prosecutorial decision-making at the charging stage.

Assessments of defendants

The focal concerns perspective, which drives much research in prosecution, provides an individual-level framework for understanding the use of discretion. According to the focal concerns perspective, court actors consider defendant blameworthiness, the danger to the community, and the social and practical implications of a sentence when making discretionary decisions (Steffensmeier et al. 1998). Since prosecutors must rely on incomplete information to assess these focal concerns, they are likely to draw upon extra-legal factors as proxies for legally relevant information (Spohn et al. 2001). Problematically, prosecutors may draw on racial and ethnic stereotypes in evaluating blameworthiness and dangerousness and, as a result, “minorities – and particularly those who are young, male, and poor – may be treated more harshly than Whites” (Kutateladze et al. 2014: 519). Criminal history is another factor prosecutors rely on when assessing whether a defendant poses a danger to the community (Kutateladze et al. 2014). The third focal concern, practical considerations, which encompasses the likelihood of conviction and how cases will be viewed “downstream” by judges and juries, weighs heavily in prosecutors’ case assessments (Frohmann 1991; 1997; Spohn et al. 2001; Spohn and Tellis 2019).

Recently, scholars have argued that defendants’ rehabilitative potential constitutes a fourth focal concern. “Salvageability” (Galvin and Ulmer 2022) or “redeemability” (Ulmer et al. 2022) refers to whether prosecutors believe defendants are capable of reform, thus making them good candidates for treatment programs. Here too, prosecutors may evaluate Black or Hispanic defendants more harshly than White defendants due to stereotypes linking minorities to violence and criminality (Galvin and Ulmer 2022).

Categorization, moral judgements, and stereotypes

Like street-level bureaucrats examined in prior sociolegal research (Maynard-Moody and Michael Musheno 2003), court actors often operate based on moral judgements. In the felony context, Sudnow (1965) argues that, rather than evaluating a defendant’s actions according to whether they fit a given statute, defense attorneys instead assess the degree to which these actions fit or deviate from their ideas of “normal crimes.” In misdemeanor courts, this categorization of defendants often has even less to do with the law and facts in individual cases than in felony courts (Kohler-Hausmann 2018; Natapoff 2013). Both Feeley’s (1979) and Kohler-Hausmann’s (2018) work on lower courts emphasizes that, rather than adjudicating guilt and ensuring due process, misdemeanor courts are primarily focused on supervision and regulation, which they achieve through “procedural hassles” (Kohler-Hausmann 2018: 183). Commonplace aspects of criminal case processing, including arrest, court appearances, and fines, are, in part, tools to evaluate how well defendants can conform to the demands of the criminal legal process. Often these hassles come in the form of a conditional dismissal, where an individual can earn a case dismissal if they comply with specified conditions (Kohler-Hausmann 2018). If that person has subsequent criminal justice contact, their previous compliance is reinterpreted as a squandered second chance. Thus, according to this account of lower courts, actors make moral judgements about defendants based on prior criminal justice contacts and their ability to navigate procedural hassles.

When morally evaluating defendants and deciding who are “bad guys and who will remain bad guys” (Maynard-Moody and Michael Musheno 2003: 91), prosecutors often draw heavily on defendants’ criminal history. At sentencing, federal prosecutors use criminal history as a window into a defendant’s character, in some instances even using vacated charges to paint a defendant as greedy, a troublemaker, or inherently criminal (Lynch 2019). Beyond criminal history, prosecutors have limited factually relevant information upon which to determine a person’s threat and recidivism risk. This makes prosecutors vulnerable to basing their decisions around race, gender, and class-based stereotypes about criminality and dangerousness (Albonetti 1991; Steffensmeier et al. 1998; Wilson et al. 2017).

Indeed, prior research suggests that stereotypes shape evaluations of credibility, responsibility, and guilt. For example, Chung (2009) argues that gendered stereotypes about caretaking responsibilities lead to the charging of mothers more often than fathers in child fatality cases. Frohmann (1997: 541) finds that, when evaluating the likelihood of conviction, prosecutors rely on the racial and economic makeup of the victim’s neighborhood and consider how this will be viewed by potential jurors. Together, these studies highlight how culturally based images, moral evaluations, and stereotypes play a prominent role in prosecutors’ decisions.

On an institutional level, Van Cleve (2016) points to the language used by judges, prosecutors, and defense attorneys as evidence of how considerations of race permeate criminal courts. In their ethnographic investigation of the Cook County criminal court system, Van Cleve found that court actors’ comments about defendants and their families draw on stigmas associated with race and ethnicity. It is within this social world that judges and prosecutors then must deliberate on the morality and criminality of defendants. Thus, Van Cleve’s work emphasizes how racialized moral judgements can become entrenched in institutions.

Prosecutor traits and role orientation

Individual prosecutors are influenced by their personal identities and experiences as they evaluate cases (Lowrey-Kinberg et al. 2022; Robertson et al. 2019; Stemen and Escobar 2018). For example, Baker and Hassan (2021) find that experienced female prosecutors are more likely to charge domestic violence and sexual assault cases (which most often involve female victims) than their male counterparts. Prosecutors’ race may also be salient in their decision-making. In a randomized experiment using vignettes, for example, Robertson et al. (2019) find that non-White prosecutors recommended significantly fewer days of confinement than White prosecutors. In a similar vein, Meldrum et al. (2021) find that White prosecutors, in general, hold more punitive attitudes than their colleagues of other races and ethnicities.

Prosecutors, like probation and parole officers (Bolin and Applegate 2018), homicide detectives (Dabney 2020), and police officers (Gau and Paoline 2017), adapt their roles on-the-job based on their personal identity and priorities. Lowrey-Kinberg et al. (2022) identify three orientations to the role of a prosecutor: the Enforcer, the Reformer, and the Advocate. While prosecutors displaying the Enforcer orientation emphasize applying the law, Reformers focus instead on defendant rehabilitation and Advocates prioritize retribution for the victim. The authors found that prosecutors of color and those with defense experience were more likely to be Reformers, while female prosecutors were more likely than men to orient as Advocates. Conversely, Enforcers held a strict sense of morality and were generally those who had wanted to become prosecutors all their lives. Prosecutors in the same office, but with different orientations, interpreted case facts and defendants’ backgrounds differently. For instance, whereas an Enforcer might view a long history of drug offenses as evidence a defendant was beyond redemption, a Reformer might view this same history as proof the justice system had repeatedly failed the defendant. Existing research, therefore, suggests that prosecutors draw upon their individual experiences and priorities when deciding how to apply the law.

Court communities

Prosecutors – with their role orientations, stereotypes, and moral judgements – operate within broader court communities (Eisenstein et al. 1988). Through frequent interactions, legal actors within a court community become highly interdependent and have significant interest in maintaining productive working relationships (Eisenstein et al. 1988). Although prosecutors have discretion at the individual case level, organizational pressures from within the court community can act as a constraint on their discretion (Levine 2006; Mellon et al. 1981).

Courts are inhabited institutions where laws and regulations are brought to life by the individuals who work within the institution (Ulmer 2019). As such, any organization’s culture is “constantly produced, reproduced, and changed through the actions of participants” (Ulmer 2019: 490). Illustrating this phenomenon within prosecution, researchers have noted office-specific patterns in dismissal rates (Frederick and Stemen 2012), plea bargaining (Metcalfe 2016), and the use of monetary sanctions (Kirk et al. 2022). Other work examines how institutional norms and local context shape the nature and goals of prosecution work, for example contributing to the degree of turnover among prosecutors (McWithey 2020; Wright and Levine 2017), work team cohesion (Levine and Wright 2012), and even “collective understandings about which cases are prosecution-worthy and why” (Levine 2006: 746). Ideas about “going rates,” not “rock[ing] the boat” (Eisenstein et al. 1988: 30), and even what constitutes a crime worth pursuing (Ulmer 2019) appear to be context specific. These localized norms may, in turn, influence how prosecutors use their discretion in charging.

Present study

Existing literature offers a variety of disparate, sometimes overlapping, theories of prosecutorial charging decisions. It is unclear how, if at all, the various predictors identified in these theorical frameworks relate to one another and how, when, or why different factors might drive decision-making. Examining how the various factors outlined in the literature on charging interact, build upon, and condition one another has the potential to illuminate how actors within inhabited institutions, such as prosecutors’ offices, respond to multiple (sometimes competing) considerations and influences. Considering these factors in isolation leaves the field’s understanding of institutional actors’ discretion – including its boundaries – incomplete. Instead, we develop an integrated model that accounts for the considerations listed in current literature and elaborates how these factors jointly shape the terrain of prosecutorial discretion. Moreover, there tends to be a disconnection between theory and methods in prosecution research. Whereas theories in prosecutorial decision-making rely on “interpretation, culture, and processes” (Ulmer 2019: 483), researchers often draw on quantitative methods that may not be best suited to understanding decision-making. This is particularly true of scholarship working within the focal concerns perspective. As Lynch (2019) argues, most studies supporting the focal concerns perspective do not contain “direct measures of legal actors’ thoughts and processes,” (1154) but rather infer support based on defendant-level data. Instead, qualitative work may be better suited to illuminating underlying decision-making processes (Lynch 2019; see also Ulmer 2019). The present study addresses these gaps in theorizing on prosecutorial decision-making by leveraging qualitative data where prosecutors unveil their processes for making charging decisions.

Method

Data collection

The data for this project come from the Deason Criminal Justice Reform Center’s Prosecutorial Charging Practices Project (PCPP). As part of the PCPP, the authors conducted semi-structured interviews and focus groups with prosecutors in three mid-sized, geographically diverse jurisdictions in the United States. Site selection was, in part, based on the chief prosecutors’ willingness to participate in the research. We also sought to research jurisdictions that were of a similar size, yet geographically diverse in order to examine the role of local legal culture in the decision-making process. Further, the three offices were led by chief prosecutors who were committed to adopting evidence-based policies and saw value in partnering with researchers to understand office operations, but fit the criteria to be considered traditional prosecutors rather than reform-minded (Mitchell et al. 2022).

All data were collected in 2018 and 2019. Although much of the research on prosecution is based on findings from large urban jurisdictions (e.g., Abrams 2011; Cossyleon et al. 2017; Kutateladze et al. 2014; Spohn and Tellis 2019), the vast majority of prosecutors’ offices in the U.S. serve populations of less than one million (Perry and Banks 2011). Our findings come from these more typical jurisdictions where the dynamics of prosecution are likely to differ from those in large, urban areas (Wright and Levine 2017).³ Franklin County employs approximately 80 prosecutors and leans Democratic in national elections. Hazelton County employs just fewer than 50 prosecutors and leans Democratic in national elections while Springfield employs nearly 30 prosecutors and leans Republican. Given that penal intensity varies between states, and within states based on urbanicity (Beckett and Beach 2021), this multi-state examination of mid-sized jurisdictions can speak to the role of local patterns in punishment while broadening the focus of prosecution research beyond major metropolitan areas.

Across our three research sites, we conducted 47 interviews and 15 focus groups with supervisors and line prosecutors. In compliance with the university’s institutional review board, we obtained informed consent from participants before conducting the interviews and focus groups. We interviewed 14 prosecutors in Franklin, 18 prosecutors in Springfield, and 15 prosecutors in Hazelton. We recruited participants of all ranks and responsibilities with the help of the chief prosecutor in each office. Interviews lasted 45–60 minutes and addressed prosecutors’ approaches to charging. We also interviewed the chief prosecutor in each jurisdiction and conducted focus groups involving three to five prosecutors, each lasting 60–90 minutes. We conducted five focus groups in Franklin, four in Springfield, and six in Hazelton. The composition of the focus groups was based on participants’ availability. Prior to the start of the focus groups, we asked participants to bring a case they found challenging to charge. During the focus group, we asked prosecutors to describe their thought processes as they handled the case. We then invited discussion from other focus group members on how their approach would have compared. Occasionally, a prosecutor discussed a case that evoked divergent decisions from different prosecutors. When this occurred, we shared that case with later focus groups in the same office to elicit opinions on how the prosecutors in later groups would handle the case.

Our goal in conducting the interviews and focus groups was to be informal and conversational while executing a semi-structured protocol. To achieve informality and promote honest responses, we did not record the interviews or focus groups (see Dunlea 2022 for a similar approach). Consequently, we do not have transcripts of prosecutors’ full responses. The researchers wrote notes during the interviews and focus groups and produced memos synthesizing patterns and trends as the research progressed. When possible, the researchers recorded direct quotations but often the notes were paraphrased summaries of the prosecutor’s explanation. Throughout the manuscript we use direct quotations when available to illustrate themes.

Analysis plan

We analyzed the notes from the interviews and focus groups using a general inductive approach that involved an initial deep reading of the raw data to identify common or important themes (Thomas 2006). This process is closely related to grounded theory (Charmaz 2006; Glaser and Strauss 1967; Strauss and Corbin 1994) but did not divide coding into an open and an axial phase (Thomas 2006). After discussing the common and important themes as a full research team, the lead author used Dedoose (2021) to review all notes and coded for these themes, adding child codes for subthemes and counterexamples to the broader themes. From those themes, we then developed a model of prosecutorial decision-making. We created the model as a tool to stimulate theorizing on how prosecutors make charging decisions (Briggs 2007). By focusing on “the most explanatory aspects of a wider range of behaviors” (Segal and Spaeth 2002: 45) we are able to illuminate some of the extralegal considerations and pressures that shape the content and process of prosecutors’ decisions.

Results

Our model of prosecutorial decision-making is built around three observable points in the charging process: the facts of the case as reflected in the initial case file, the initial charges the prosecutor files, and the final charges at disposition (see Figure 1). In discussing their initial charging decisions, prosecutors’ responses frequently included a discussion of plea offers. Therefore, our model of prosecutorial charging decisions includes the final negotiated charges since that is a key consideration for prosecutors when making initial charging decisions. Although prosecutors consider the governing law at charging (represented by the dashed line running from the facts to the initial charges on Figure 1), the decision-making process involves additional strategic considerations and a longer-range view of the case. Even at the earliest stages of charging, prosecutors had a downstream orientation toward the eventual outcome of the case (labeled “desired outcome” on Figure 1). Thus, rather than charging based strictly on the governing law, they charge with an end in mind, a theme we describe more fully below.

Figure 1. Model of prosecutorial charging decisions. Notes: This model of prosecutorial decision-making is built around three observable points in the charging process: the facts of the case, the initial charges filed, and the negotiated charges (i.e., plea-bargaining). The dashed line from the case facts to the charging decision represents the governing law, but prosecutors decide how to apply the law through a process of character construction. Based on their evaluation of a defendant’s character, prosecutors determine what they view as an appropriate outcome for a given defendant (i.e., desired outcome) and then charge accordingly. This largely internal/individual process is shaped by prosecutors’ social identities and prior professional experiences. How prosecutors act on their desired outcome, however, is constrained by organizational factors such as the priorities of the chief prosecutor and the court community context in which they work. This model integrates insights from the focal concerns and court communities’ perspectives with prior findings on the relationship between prosecutor traits and decision-making to create an integrated model of prosecutorial charging decisions.

We found that prosecutors determine the “end” they have in mind through a process of character construction (labeled as such on Figure 1). When prosecutors spoke about how they handled cases, their explanations involved a deep description of the type of person the defendant was, a conclusion they based on the limited facts available in the case file. These inferences about defendants’ morality were replete with character development, often drawing heavily on the defendants’ criminal history.

We found, however, that individual prosecutors arrive at vastly different constructions of defendants’ character, both across study sites and within the same office. In explaining this variation, we identify the many influences that shape how prosecutors interpret the facts of a case and arrive at charging and plea-bargaining decisions. We separate these into “Internal/Individual” and “Organizationally Constrained” portions of the charging process. As part of the internal or individual factors, we find that prosecutors’ identities and prior experiences shape how they envision the nature and character of defendants. While we observed character construction and charging with an end in mind across our three research sites, the process by which prosecutors achieve their desired outcome is conditioned by the chief prosecutor and the local community context (incorporated as part the “Organizationally Constrained Process” on Figure 1). Our model integrates concepts from previously siloed legal, criminological, social psychological, and sociological frameworks to explain how identity and experience shape prosecutors’ interpretations of case- and defendant-level information and how organizational factors help explain when and how prosecutors act on their interpretations through charging.

Below, we elaborate upon our model. We begin by describing two core ideas that run through the entire charging process – charging with an end in mind and character construction. Following the description of these two key themes, we explain how prosecutors’ identities and experiences shape their charging decisions. Finally, we turn to the organizational constraints within which prosecutors must work: the policies and priorities of the chief prosecutor and the community context. Throughout, we share examples from our research that are emblematic of broader themes we identified. To preserve anonymity, we identify interviewee and focus group members’ responses with a code (F#/S#/H# for Franklin/Springfield/Hazelton interviewees and FFG, SFG, and HFG for the focus groups).

Charging with an end in mind

Consistent with prior work on prosecutors’ downstream orientation (Frohmann 1991; 1997), we found that, starting at the earliest stages of a case, prosecutors assess what they believe is the optimal outcome for that defendant (i.e., diversion, probation, custodial time, or dismissal). One Franklin prosecutor summarized this approach by quoting Stephen Covey’s advice to “begin with the end in mind” [F7]. As another prosecutor put it, “pick your battles and look at what the end result” should be [H13]. Charging is, therefore, part of a strategic process to achieve a desired outcome, not simply a process of selecting a charge that most closely maps on to a particular statute. In Franklin, prosecutors in focus group discussions often recited a version of the following if-then statement when describing their process for charging: if the charge is a class [X], the defendant will have to serve [X%] of the sentence, which gives them [X] amount of jail or prison time. Prosecutors then focused on what charge they could select that would achieve the amount of custodial time they thought the defendant deserved. In one example of how this outlook manifested in charging decisions, prosecutors explained that even when the elements for attempted murder were met, they would rarely charge it because specific intent to kill or cause gross bodily injury was hard to prove. Instead, they would select a series of lesser charges that were easier to prove in order to reach a similar sentence of the defendant [FFG].

While some prosecutors denied bringing extraneous charges [F3, F4, F13, H7, H11, S1], many other prosecutors’ calculus in achieving their desired end for a defendant consisted of adding charges to “incentivize the defendant to plead” [FFG; also S5, S18, HFG, H5, H6, H14, F10, F11, F12, F14]. Several prosecutors described using punitive leverage because they were focused on achieving retribution for victims. To achieve this end, they chose to add charges that were easier to prove but carried a lighter sentence, instead of or in addition to charges that carried a heavier penalty but were harder to prove. In just one example of several with the same theme, a Hazelton prosecutor shared a case of child sexual assault in which they charged the defendant with four felonies. Of these four charges, three carried a life sentence given the defendant’s criminal history. When asked why they added the fourth charge (which did not carry a life sentence), the prosecutor explained it was a strategic decision.⁴ If the case went to trial and the young victim was not specific enough in describing the acts that took place, they could still prove the fourth charge and achieve some retribution for the victim [HFG].

Charging aggressively was, at times, viewed as necessary to secure treatment for the right defendant. In Franklin, for example, both the chief prosecutor and the supervisor of the charging unit (F1) explained that it might be necessary to charge a defendant with a felony in order to secure access to a treatment program. They would then allow the defendant to plead to a misdemeanor once treatment was completed. Conversely, in a handful of cases, prosecutors chose to charge leniently when they viewed the defendant as redeemable (as in Galvin and Ulmer 2022). Examples of defendants viewed as worthy of leniency included a chronic petty offender who had been failed by the system and needed housing and psychiatric care [HFG], an honest and cooperative defendant with minimal priors [SFG, HFG], and a defendant who was remorseful during a video recorded confession and who had himself been abused as a child [HFG].

Thus, although there was variation in interviewees’ approach to charging, prosecutors across all three sites clearly considered how selecting charges upfront could lead them to their desired end for a defendant. Crucial to the process of charging is how prosecutors arrive at the desired “end” of a case. Reminiscent of other street-level bureaucrats’ evaluations (Maynard-Moody and Michael Musheno 2003), sorting defendants in this way is a highly subjective process that is closely tied to character construction of the defendant.

Character construction

Like court actors in misdemeanor courts (Kohler-Hausmann 2018) and other street-level bureaucrats (Maynard-Moody and Michael Musheno 2003), we found that felony prosecutors attempt to divine a defendant’s character, what motivated them, how dangerous they were to the community and, consequently, the appropriate punishment. In doing so, prosecutors wove together criminal history (which, for most crimes, is not a legally cognizable factor for charge selection), with legally cognizable factors such as offense severity and intent.

Criminal history

A defendant’s prior record was used as an indispensable element of character development at the time of charge selection as prosecutors decided whether the defendant was a habitual offender [SFG] and a “bad guy” [FDA]. With very few exceptions, prosecutors began their description of a case with a note about the defendant’s prior record, or lack thereof. For example, in Franklin, prosecutors began their case review by turning to the back cover of the case file where criminal history information was recorded [FFG]. Before even reading the facts of the case, they appeared to search for clues as to the defendant’s character. In one example that exemplifies this broader theme, a prosecutor in Hazelton began their explanation of a case by saying: “Black female, late 30s, didn’t have an extensive record but had a felony possession charge and charges for use of a controlled substance and she hadn’t served much time for her priors. The defendant had previously been involved in multiple alternative programs but flunked out” [HFG]. These details about prior offenses and prior attempts at alternative programs were presented proof of an unsympathetic defendant, setting the stage for charging the defendant harshly in an attempt to secure the highest possible penalty.

In an example of how criminal history was used to sort defendants, Springfield prosecutors described making charging decisions in solicitation cases. The prosecutors explained that they wanted to determine if the offenders were “predator[s] or dumb guy[s]” [SFG]. Recitations of a defendant’s criminal history went hand-in-hand with descriptions of the defendant being “creepy” or saying “some gross things.” On the other hand, a lack of prior convictions suggested that the contact was most likely a “one-time deal,” and the defendant simply took advantage of an opportunity, making him – in their minds – less morally culpable. Thus, prosecutors appeared to use criminal history as a proxy for elements of defendants’ character and, subsequently, how they should be charged.

A Franklin focus group discussion of a young man arrested for a revoked driver’s license provides an additional illustration of prosecutors’ reliance on criminal history to signal morality. The defendant was stopped with one pill of Adderall and a trace amount of a Schedule I drug (which he admitted to using). The defendant had a criminal history that consisted of some drug and theft charges and one assault. The assigned prosecutor conceded that the defendant was young, but, in the end, charged the defendant with a felony, citing the defendant’s prior record: “I would look at his background and think, ‘he doesn’t give a fuck about any break you’re going to give him’” [FFG]. By contrast, another prosecutor in the same office interpreted the defendant’s same criminal history as masking a drug addiction that needed treatment and justified a misdemeanor charge. While, ultimately, each prosecutor came to a different moral judgement, each drew upon criminal history in evaluating the defendant’s character [FFG].

Criminal history was especially important in cases where prosecutors were torn on how to proceed. For example, prior record would tip the scales toward pursuing the case in non-violent crimes where there was not a significant public safety concern or where intent was unclear or hard to prove [S3, SFG]. In a Hazelton focus group, three prosecutors discussed a weak drug possession case where none of them felt they had the evidence to prove the defendants were in possession of the drug. One prosecutor said they would consider trying to leverage a plea but for the fact that the defendants did not have any priors [HFG]. In this case, the defendant’s priors, not the facts of the case, were the deciding factor in the prosecutor’s decision-making process. In these borderline cases, criminal history filled in the gaps in the prosecutor’s story about the defendant.

Further, prosecutors shared stories of other defendants whose lack of or minimal criminal history was a consideration in amending existing charges down [S2, F3, FFG, HFG]. A Franklin prosecutor [F3] offered an example of a retail theft case in which the value of the item was just over the monetary threshold for a felony. Although the case met the elements required by statute, the defendant had no criminal history and therefore, “[This type of case] always goes misdemeanor” because that is “more just,” they explained. The specifics of defendants’ criminal records were also important. Prosecutors considered recency of the convictions [S2, SFG], number of prior justice system contacts (even if not resulting in a conviction) [HFG, SFG], and even how other prosecutors had handled prior cases in the defendant’s file [SFG]. For example, one Springfield prosecutor explained that if they saw, in reviewing the defendant’s records, that a defendant originally had a felony charge but pled to a misdemeanor in a previous case, they viewed this as the defendant “got a chance here” and wasted it, thereby making him or her not deserving of leniency again [SFG]. Criminal history could even supersede victims’ wishes [S5, FFG]. For example, one prosecutor described a domestic violence case they handled where the victim wanted probation but, because the defendant had three prior felonies (over a decade old), the prosecutor wanted the defendant to serve jail time [FFG]. In all these ways, criminal history was central to constructing the character of the defendant.

There were, however, a few counterexamples of prosecutors who resisted using criminal history to inform their initial charging decisions [F5, F6, FFG]. The strongest of these counterexamples was provided by a gang prosecutor in a Franklin focus group. This prosecutor previously served in the military and was narrowly interested in whether the facts of the case fit the statutory definitions. Consequently, they had little interest in moral evaluations of defendants and were the most likely to charge strictly based on the elements.⁵ Despite these few counterexamples, for the majority of prosecutors criminal history was a primary method of defendant evaluation.

Offense severity and intent

Prosecutors used severity of the offense as character development in deciding who was a “bad guy” even in the absence of a long or violent criminal history [FFG, HFG, SFG]. One Franklin prosecutor emphasized this point using the story of a 19-year-old who took a victim to an ATM under threat of force. Although the defendant’s prior history was sparse – only one juvenile conviction for burglary – the prosecutor still charged aggravated robbery and kidnapping due to the egregiousness of the circumstances [FFG]. Likewise, a prosecutor in Hazelton described a sexual assault case where they found the facts so heinous that they said they would have asked for a life sentence regardless of the defendant’s prior record [HFG]. In another example, a Springfield prosecutor shared the case of a defendant who had beaten his elderly wife while drunk. The victim was hospitalized with injuries to her internal organs. The defendant had a long history of arrests for driving while intoxicated (DWI), but few convictions and no prior assault charges. Still, the prosecutor intuited a violent personality to the defendant, saying, “you don’t get here without a violent history” [SFG]. The prosecutor reasoned that previous juries must have overlooked the DWI, or the police must have conducted poor investigations.

Prosecutors used intent to animate defendants’ character [F8, F12, FFG, S4, S12, SFG, HFG], in addition to its relevance to whether an act meets the elements of a criminal offense. For instance, a Franklin prosecutor explained the importance of thinking, “Why is [the defendant] doing this?” when, for example, trying to differentiate between a defendant who was in possession of drugs for personal use and a defendant who intended to sell the drugs [FFG]. A prime example of the importance of intent was provided by a Springfield focus group. The case involved a defendant who was a legal immigrant to the United States. The defendant left her sick young daughter at home asleep to buy medicine. While the mother was away, the girl woke up and was found wandering around outside the apartment building. The mother returned and was visibly distressed but was arrested for child endangerment. The prosecutor who handled the case explained that the defendant was clearly not a career criminal or monster (i.e., she did not have a long criminal history). They recognized that the defendant had good intentions and left to obtain medicine for her daughter, “not for Pilates” [SFG]. The prosecutor characterized the defendant as good mother who made a mistake, weaving together intent and criminal history in their evaluation of the defendant.

Individual-level influences on character construction

Prosecutors’ experiences as members of certain sociodemographic groups as well as their professional backgrounds inform their approach to the work of prosecution (Lowrey-Kinberg et al. 2022). We find that identity and prior professional experiences are lenses through which prosecutors filter the factors described above (criminal history, offense severity, intent). Thus, the mechanism by which prosecutors arrive at a desired end for a given case – character construction – is a process in which prosecutors’ own social identities and prior experiences interact with case- and defendant-level factors.

Identity

Black Americans are overrepresented in the criminal justice system as both victims and defendants (Ghandnoosh 2015) but are underrepresented in the legal field (American Bar Association 2022). These realities seem to create a complex relationship between race and charging philosophy. For example, a Black female prosecutor in Springfield, S7, acknowledged that she finds it “really, really hard to [trust the police] because I know there is over-policing of Black and Brown communities.” She argued that even if a case technically met the elements of a crime, she did not feel the need to always charge the defendant, if she thought the defendant did not deserve formal punishment. She described a common scenario that she viewed as worthy of dismissal rather than charging: defendants being charged with trafficking drugs into prison (a felony) when they merely had drugs in their pockets as they were being booked into jail. As she explained, “if I know the case is crap, why charge? Is that justice?”

Similarly, a prosecutor in Franklin talked extensively about being one of only a couple Black prosecutors in the office and how his approach to charging differed from others in the office. He had previously been incarcerated for a crime he did not commit and described himself as being less likely than his colleagues to move forward with every possible charge in a case. He noted that he was always mindful of the effect a case can have on a defendant’s life and avoids felony charges when possible [F7]. The other Black attorney we interviewed in Franklin, F12, was raised in poverty with a negative view of police and the prosecutors’ office. She similarly believed she charges differently than her colleagues by examining the totality of the circumstances, including police behavior and defendants’ personal circumstances. As examples, she described wanting to dismiss a case in which police used jaywalking as a pretense for making a stop, as well as a drug possession case in which the defendant was in possession of a single Xanax pill. Both these prosecutors described their experiences as lenses through which they interpreted case facts and selected charges.

However, it was not the case that all the Black prosecutors we interviewed were uniformly more moderate in charging than their White colleagues. For example, H5, a Black prosecutor in Hazelton, believed that she secures more convictions than most of her colleagues and bemoaned the liberal leaning jury pool in her county. In a similar vein, S11, another Black female prosecutor, became a prosecutor to serve as an “advocate for the community.” She viewed defendants as having significant free will and was described by the chief prosecutor in Springfield as having a more punitive philosophy than others in the office. These counter-examples may, in part, result because race is just one component of social identity that interacts with other aspects of a person’s identity and experiences (Burgess-Proctor 2006; Crenshaw 2012).

Prosecutors’ gender also constituted a lens through which they viewed defendants’ actions and inferred their intentions. Nationally, women are more likely than men to be subjected to sexual harassment and experience sexual assault (United States Department of Justice 2019) which, we argue, may influence female prosecutors’ attitudes toward charging. For instance, in one Springfield focus group of female prosecutors, the discussion focused extensively on defendants or crimes they viewed as “predatory,” “creepy,” “gross,” and “scary” and how they charge these defendants more harshly [SFG]. Female prosecutors’ repeated use of these adjectives, which were not as present in male prosecutors’ accounts, reflect how these female prosecutors focused on crime details they found particularly disturbing to paint a picture of the type of defendant with whom they were dealing. These observations suggest that moral evaluations of defendants are not divorced from prosecutors’ gendered experiences.

Female prosecutors were also the only ones to discuss being able to directly put themselves in the position of victims or be victim focused [H5, H10, H11, H12, H14, F2, F9, S7, S14], even going so far as to describe themselves as “vessels of vengeance” for victims [SFG]. This, in turn, appeared to affect their charging decisions in crimes with a victim. H12, for example, views herself as more lenient on drug cases than others in the office, but more punitive on crimes of personal violence because she can see herself in the role of a victim and has “a problem with violation of personal space.” Similarly, F2 described herself as being acutely concerned about violent offenses and the need to protect children and the elderly. Consequently, she choose to aggressively charge sex offenses and those cases with children and elderly victims. In contrast, she described charging thefts and property crimes less aggressively.

Prior professional experience

Prior professional experiences also condition how prosecutors make charging decisions, with most veteran prosecutors sharing that they charged differently than they did as novices. While newer prosecutors tended to be focused on the elements of the offense, experience generally moved prosecutors away from strict application of the law to a more contextualized decision-making process where they were able to charge with an end in mind [S3, S17, H11, H12, F1, F6] (see also Wright and Levine 2014).⁶ Experienced prosecutors also described “making better decisions” now by declining cases when there is not enough evidence to win [F1], and generally working faster [H4] and being more confident in dismissing cases when warranted [F11].

Moreover, experience teaches prosecutors that some offenses may not be worth pursuing or worth giving a defendant a felony record. Emblematic of this broader theme, one prosecutor described how now, with more perspective on what deserves a charge, he asks, “is it worth proving it?” [S3]. S17 explained that when they first started, they “went overboard” when selecting charges and charged as many counts as possible, whereas time has tempered this approach. S17 gave the example of charging a defendant with talking back to the police. While they may have added a charge for this behavior earlier in their career, they came to realize the additional month of punishment was not worth the resources because it was unlikely to have a significant impact on the defendants’ future behavior. Other senior prosecutors described experience as having given them “more compassion” for defendants [S16, F2]. One of the most senior attorneys in Franklin, F1, described how earlier in their career they “[didn’t] give a damn that an 18-year-old got tagged with a felony.” With greater awareness of the consequences of a felony conviction, they are now more likely to agree to a misdemeanor sentence for a felony for the right defendant.

In Hazelton, which has the highest violent crime rate of the three jurisdictions in our study, some prosecutors attributed their waning aggressiveness to the seriousness of the cases they see [H10, H14]. H14, for example, had been a prosecutor in Hazelton for 14 years and explained, “We all become somewhat desensitized the longer we stay here. We become less worked up by the smaller crimes.” They also felt that they charge less now than as a new prosecutor because they have developed the ability to look at the “big picture.” Similarly, H10, a prosecutor with 10 years of experience, described how, over time, they started to see even serious cases and think “I’ve seen worse.” Thus, although the mechanism differed from other prosecutors, the tempering effect of experience remained.

In contrast, a group of Springfield prosecutors said experience had the opposite effect. With experience, they had become more secure and confident in their judgments, which allowed them to add counts and take risks [S4, S5, S16, S18, SFG]. S16, a prosecutor of 24 years explained that their experience allowed them to resolve more cases by plea bargain than their colleagues because their cases were “priced to move.” In other words, because they had more experience than anyone in the office, they had a better sense of the plea deals defendants would take. As discussed in greater detail below, this differing pattern may be partially due to the context in which Springfield prosecutors operate: a conservative community with a “tough-on-crime” orientation and a chief prosecutor who constrains line prosecutors’ discretion.

Due to their extensive direct experience with defendants, prosecutors who had been defense attorneys tended to integrate humanizing details about the defendant into their case evaluations and subsequent charging decisions, with greater focus on rehabilitation [F1, F3, F8, H1, H8, S8]. For example, H1 described their frustrations in the role of a public defender, citing the many unmet needs of clients, and explained that they became a prosecutor in order to have greater leverage in offering alternative outcomes to better serve defendants. F3, another former public defender, extolled the value of this experience, arguing they were better able to “look at both sides” of a case. Their charging philosophy, especially toward drugs, reflected this experience, as they were focused on keeping first-time drug offenders with small amounts of drugs out of prison. To achieve this, they charge cocaine residue as a misdemeanor, rather than a felony, for young defendants with no priors. S8, a prosecutor with experience in private and public defense, took particular issue with charges for driving with a suspended license, describing these as an issue of “driving while poor,” rather than for being a danger to society. They attributed their ability to craft a reasonable plea offer (i.e., sending the defendant to a driver’s license clinic) to their background in defense work. Thus, prior experience as a defense attorney also seemed to shape the stories prosecutors created about defendants.

Organizational constraints

We find that that the priorities of the chief prosecutors and the community context of the jurisdiction conditioned the degree to which prosecutors have the discretion to achieve their desired case outcomes. As such, our findings emphasize the localized nature of discretion, the power of court community norms and county legal culture in shaping criminal case outcomes (Eisenstein et al. 1988), and the utility of conceptualizing the prosecutor’s office as an inhabited institution (Ulmer 2019).

Chief prosecutor

The policies and priorities put in place by the chief prosecutor dictate the degree to which line prosecutors felt empowered to use their discretion. Whereas line prosecutors in Springfield perceived “mixed messages” [S9] from the chief prosecutor about which cases to prosecute and which to dismiss, line prosecutors in Franklin and Hazelton described having almost complete discretion to handle cases as they saw fit.

The Springfield chief prosecutor had a stated goal of dismissing “bad cases” (i.e., cases they viewed as not being worth the resources to pursue) and, at the time of our research, had recently sent an email stating this new policy [SDA]. Yet line prosecutors felt that this message was not being implemented in practice [S7, S9, S11, S12, S13, S14, S15, S18]. For example, S18 described a case they inherited from another prosecutor who had left the office in which a bar owner was constantly getting into fights. The previous prosecutor on the case had been denied when he asked the chief prosecutor for permission to dismiss the case, only for S18 to have the dismissal approved a few months later. Another prosecutor described how the message from the chief prosecutor was to “not bring weak-ass tampering cases,” but, in practice, the chief prosecutor’s second-in-command always wanted to find a way to bring the case anyway [S11]. Another prosecutor estimated that they would be justified in dismissing close to 15 percent of their cases but felt that they were only “allowed” by supervisors to dismiss about five percent [S13].

A striking example of this tension between the chief prosecutor’s priorities and line prosecutors’ discretion came from a focus group discussion of a case, described above, where a mother was arrested for child endangerment for leaving her sick daughter at home while she went to buy medicine [SFG]. Despite the prosecutor’s evaluation of the defendant as a good mother who made mistake, they did not feel empowered to use their discretion to drop the charges. The defendant was ultimately convicted of a felony, an outcome the prosecutor felt was unjust but unavoidable given the environment the chief prosecutor cultivated. The perceived contradiction between the Springfield chief prosecutor’s message and the directives from leadership on individual cases disempowered prosecutors in using their discretion to handle cases.

The Springfield chief prosecutors’ approach differed markedly from that of Franklin and Hazelton’s chief prosecutors, who instead gave line prosecutors a wide degree of discretion to charge and resolve cases as they saw fit [H7, H8, H9, H10, H12, H13, H15, F6, F9, F10, F11, F13, FFG]. “I hired you, and you continue to be on staff, because I trust your judgement” was how H13 explained the Hazelton chief prosecutor’s message to line prosecutors. As a result, a given defendant’s fate rested heavily on the line prosecutor’s characterization of him or her. A focus group discussion of a police stop in Hazelton depicts the implications of chief prosecutors giving line prosecutors significant latitude to exercise their discretion. The police stopped a vehicle with two occupants. During the stop, the officers smelled marijuana and found nearly $2,000 worth in the center console [HFG]. All three prosecutors in this focus group agreed that it was a weak case. They described the officers as being excited about the case, but the prosecutors agreed they could not prove that the occupants of the vehicle were aware of the drugs. Based on the same set of facts, however, the three prosecutors described approaches that would have led to very different outcomes for the defendants. One of the prosecutors said they would nolle prosequi the charges because it would be impossible to prove, saying “if I’ve got nothing, I’m not going to prosecute.” On the opposite end of the spectrum, another attorney said they would have been able to get the defendants to plead to a misdemeanor with a suspended sentence. The third prosecutor in the focus group said they would have considered leveraging a plea but ultimately would have nolle prosequied the charges because the defendants did not have a criminal record [HFG].

This between-prosecutor variation was especially evident in Franklin where the chief prosecutor took pride in having diversified the staff during their tenure. They had recently recruited attorneys who were Black, older, and had prior defense experience [FDA]. Yet, given our findings concerning the role of social identity in charging decisions, described above, the combination of the diverse staff and the hands-off approach of the chief prosecutor led to divergent characterizations of defendants, and the freedom to charge according to those assessments. Resisting arrest cases showcased how prosecutors’ ideological differences translated into differences in charging decisions. While some prosecutors identified resisting arrest [F5] and disorderly conduct [F10] as the rare offense types where they routinely added charges because they felt the police under-charged the case, others cited these two types of charges as routinely problematic because officers over charge the number of counts [F6, F11]. The fact that these widely differing approaches existed in the same office was viewed by prosecutors in the office as a natural consequence of discretion, rather than as indicating a problematic lack of consistency.

Community context

The three jurisdictions in our study varied in regard to the ideological leanings of the surrounding community. Whereas the Springfield community was conservative and “tough-on-crime,” Franklin and Hazelton were both in jurisdictions that lean Democratic. The community context in each jurisdiction conditioned how and to what extent prosecutors were able to use their discretion to act according to their construction of defendants’ character.

Due to punitive bent of the community in Springfield, prosecutors were almost always inclined to take cases to the grand jury for fear of community “backlash” [S12, also SFG] if they exercised their discretion, even when their characterization of the defendant would have suggested a declination or dismissal as the appropriate outcome [S2, S5, S6, S7, S9, S11, S12, S13, S14, S15, S18]. As one prosecutor put it, “if [a case] technically meets the elements but it’s a shitty case, I leave it to the grand jury. I don’t want to be the bad guy. Let the grand jury be the bad guy” [S9]. Another prosecutor described a similar thought process: “I was always taught to bring a case to the grand jury if the elements exist beyond probable cause [if] there’s no legal basis to dismiss” [S5]. A “bonus” [S18] of this approach was being able to point the finger at the grand jury if a victim called to complain about a case not being pursued.

Even in the case of “stupid felonies,” cases that meet all the elements but are not in the interest of justice to pursue [SFG], prosecutors in Springfield agreed their only option was to present the case to the grand jury. An example of a case meeting these criteria was a case involving a man who shoved his mother-in-law during a heated fight. Because the victim had recently turned 65, she was considered elderly, and the defendant was charged with a felony [S17]. The prosecutor was conflicted because the victim had provoked the defendant, and if the incident had taken place a month earlier the defendant would have been charged with a misdemeanor. Despite their own interpretation of the case, S17 decided to let the grand jury be the “conscience of the community” and saw it as only fair for them to decide whether the case should be prosecuted. Some prosecutors described “brow beat[ing]” the grand jury into issuing a no bill in these types of cases, yet always felt compelled to at least present the case [SFG]. In the event the grand jury indicted a case that the prosecutor hoped they would not, prosecutors described having to be creative in how they achieved their desired outcome for the case by either quietly dismissing it after months of delays [S7] or offering a particularly lenient plea deal to the defendant [S5, S6].

Prosecutors’ view of and relationship with local law enforcement also conditioned their discretion. In Springfield, for example, prosecutors viewed the local police as zealous, yet incompetent with a “negligible grasp of what the law is” [S15; also S1, S6, S7, S8, S12, S17, SFG]. Referring to the police department, one prosecutor said, “it’s a shit show over there” and bemoaned their futile attempts to educate officers on the law, saying, “we do training over there and nothing changes” [S8]. Tampering with evidence cases were often named as particularly problematic [S1, S6, S11, S14, S15, SFG]. In a case cited by several prosecutors, an officer pulled over a car in which the 19-year-old passenger had been drinking and dropped a beer bottle out of the window as the officer approached the driver’s door. Rather than charging the passenger with underage drinking, the officer accused the passenger of tampering with evidence [S1, S11, SFG]. Prosecutors also cited law enforcement requesting charges for drug free zone enhancements in cases where this enhancement did not apply [SFG], as well as overcharging family assaults as serious bodily injury cases when the elements were not met [S1, S8, SFG]. As a result, prosecutors were frustrated by the “crap cases” [SFG] the police would bring because there is “a presumption that they should continue” [S12] once filed. Yet, due to the tough-on-crime orientation of the community, prosecutors did not feel empowered to use their discretion to amend charges and would almost never decline to charge cases outright.

In stark contrast to the strained relationship Springfield prosecutors had with law enforcement, in Franklin prosecutors had close working relationships with law enforcement [FDA, F1, F2, F3, F4, F5, F6, F7]. Rather than feeling community pressure to accept charges requested by law enforcement, as in Springfield, prosecutors in Franklin said charging was collaborative [FFG, F2, F3]. The significant amount of face-to-face time between officers and prosecutors in this office was partially due the jurisdiction’s felony review process, which required officers to call or come to the office to discuss what charges, if any, were appropriate before making a felony arrest. Several attorneys even said their philosophy in charging was to ask the officer what charges they thought were appropriate and then follow their lead, rather than unilaterally deciding a charge [F2, F3]. Importantly, the charging unit is staffed by senior attorneys who will not be “a doormat for police” [F1]. Franklin prosecutors’ frequent interactions with officers gave them the opportunity to explain their rationale for not bringing charges in cases they felt did not warrant prosecution and to help officers catch missing information [F1].

Whereas prosecutors in Springfield and Franklin repeatedly mentioned the police department as an important factor in how they handled cases, prosecutors in Hazelton more often cited a juror pool and judges they felt favored defendants [H1, H3, H5, H9, HFG]. Prosecutors in Hazelton responsible for serious felonies reported adding charges to indictments to encourage defendants to take a plea offer. They defended stacking charges (“loading the deck”) on the basis of “pro-defendant judges” and a “defendant-friendly” juror pool that tended to mistrust police [HFG]. For example, in a drug deal turned non-fatal shooting, prosecutors charged the defendant with two charges. The elements were also met for a third, more serious charge, but they explained that they were confident the defendant would plead to one of two original charges, and they feared losing credibility with the liberal judge if they added too many charges. Due to inconsistent testimony between witnesses, prosecutors also believed that a local jury would “100 percent” return a not-guilty verdict if the case were taken to trial, so they offered a plea deal for deferred prosecution [HFG]. Given the liberal leanings of the local community, prosecutors in Hazelton were highly motivated to secure plea agreements to achieve their desired outcomes and would frequently charge with that goal in mind.

Taken together, interviewees’ descriptions of their decision-making processes in the context of their workplaces highlight how there is variation, not only in how prosecutors evaluate the moral character of defendants, but also in case processing norms between jurisdictions. In all three jurisdictions, the court community context conditioned the degree to which prosecutors could enact their desired “end” for cases. Chief prosecutors, too, have a proximate influence on prosecutors’ work. Theoretically, they are able to moderate the impact of the court community on the discretion of their line prosecutors. However, in our research, they chose either to bend to the will of the local culture (as in Springfield) or defer to the discretion of their line staff (as in Hazelton and Franklin).

Discussion

Based on extensive interviews and focus groups with prosecutors in three mid-sized jurisdictions, we propose a model of prosecutorial charging decisions that integrates existing theoretical frameworks and identifies the mechanisms through which extralegal factors come to influence charging decisions. We find that prosecutors charged with an end in mind and determined that “end” through a process of defendant character construction. In prosecutors’ minds, the purpose of charging was to set the groundwork for getting a final disposition that fits, not just the crime, but the defendant. How prosecutors interpreted the available case facts, however, was related to their own identities and professional experiences. Further, the extent to which prosecutors felt empowered to carry out their desired outcome was conditioned by the priorities of the chief prosecutor and the local court community.

Theoretical contributions

Our charging model integrates existing theoretical frameworks to explain how prosecutors use their discretion at charging. This model moves beyond aggregation and elaboration to instead detail how the various factors described in the literature interact and build upon one another. For instance, while moral evaluations of defendants, drawing strongly upon criminal history, were predominant in interviews and focus groups, prosecutors’ identities served as lenses through which they interpreted case facts and determined appropriate dispositions for defendants. Further, as a backdrop to discretionary decision-making, the priorities of the chief prosecutor could heavily influence the extent to which prosecutors could make decisions based on their evaluation of defendants. A primary contribution of our work, therefore, lies in illustrating how prosecutors weave together varied considerations in selecting a charge.

Additionally, we address the gap noted by Lynch (2019) and Ulmer (2019) in their critiques of the current methods used in most courts research. Although much can be deduced from quantitative analyses of charging decisions, by focusing on the mechanisms underlying how prosecutors evaluate defendants and the consequences of these moral evaluations, our findings speak to how the focal concerns of prosecution (Galvin and Ulmer 2022; Spohn et al. 2001; Ulmer et al. 2022) work in practice. For instance, many quantitative studies of case processing decisions control for defendants’ criminal history (e.g., Frederick and Stemen 2012; Kutateladze et al. 2014; McCoy et al. 2012; Shermer and Johnson 2010), yet few explanations exist for why prosecutors would weigh criminal history in initial case evaluations beyond its ability to signal that a defendant is a danger to the community (Kutateladze et al. 2014; Lynch 2019). The prosecutors in our study clearly articulated how they used criminal history to determine a defendant’s character, precisely what they look for in prior criminal history (number, recency, and severity of priors), and the way it drives charging decisions. In this way, prosecutors use criminal history as an essential aspect of a defendant’s “moral makeup” (Lynch 2019: 38) and charge them according to this characterization.

In a similar way, our research demonstrates how practical constraints and consequences affect charging decisions in practice. We found that prosecutors are mindful of the chief prosecutor’s priorities as well as the local community context, primarily the political leanings of the community and prosecutors’ relationship with local law enforcement. These practical constraints and consequences conditioned the degree to which prosecutors were able to exercise their discretion to make the charging (or dismissal) decisions they thought appropriate. We find that it is primarily the discretion to be less punitive that seems to cause friction with police departments or the local community, whereas prosecutorial discretion to be punitive (i.e., tough on crime) has traditionally gone unquestioned by other court actors (Davis 2007). Together, our findings emphasize the importance of broader professional and organizational considerations in understanding the totality of the practical constraints and consequences prosecutors weigh during charging.

Scholars have pointed to the focal concerns of dangerousness and blameworthiness as one vehicle by which racial and ethnic stereotypes seep into court actors’ decision-making, resulting in inequitable outcomes (Hawkins 1981; Kutateladze et al. 2014; Ulmer et al. 2022). Our results suggest that prosecutors’ evaluation of redeemability (Galvin and Ulmer 2022; Ulmer et al. 2022), is another possible way in which stereotypes surrounding race may be folded into the charging process. In selecting a charge, prosecutors subjectively categorize defendants into either being inherently “bad guys” or someone with a fixable problem. Reminiscent of deciding which defendants are capable of being “salvaged” (Galvin and Ulmer 2022: 1333), this process is vulnerable to bias and may contribute to less favorable outcomes for Black and Hispanic defendants who may be more likely to be seen as violent or dangerous (Galvin and Ulmer 2022; Hawkins 1981) rather than deserving of help and support. Unsurprisingly, none of our interviewees volunteered that they consider Black or Hispanic defendants more likely to be “bad guys.” However, stereotypes linking young, Black or Hispanic, and male individuals with notions of dangerousness and threat may inadvertently fill in the gaps in prosecutors’ assessments of defendants’ character (Albonetti 1991; Levinson et al. 2010; Spohn and Holleran 2000; Wilson et al. 2017). Further, the overrepresentation of young, Black and Hispanic men among those arrested (Beckett et al. 2006; Kochel et al. 2011; Lytle 2014) may reinforce biased cues about who fits the profile of a criminal.

Prosecutors’ reliance on criminal history may also perpetuate the over incarceration of Black and Hispanic men. With the exception of a small number of offenses that have escalating charge codes for repeat offenses (e.g., DWI), prior criminal history is legally irrelevant to the consideration of what offense occurred. Nevertheless, prosecutors’ moral judgements of defendants relied heavily on criminal history. Simultaneously, sentencing procedures in most jurisdictions make prior record an explicit factor in determining the allowable sentence for a convicted defendant (Tonry 1987; United States Sentencing Commission 1993). Thus, considering criminal history twice – first at charging, then in sentencing – presents a potential double penalty for defendants. This is especially problematic given that criminal history is not a race-neutral measure (see Omori and Petersen 2020; Shannon et al. 2017). If a prior record signals a greater likelihood of charges being filed, more severe charges, and more severe sentences for those charges if convicted, the result is compounding disadvantage for Black and Hispanic defendants.⁷

While our work elaborates the mechanisms underlying the focal concerns perspective, our findings also suggest that prosecutors’ social identities and backgrounds influence the charging process via prosecutors’ own “value system” (as in Maynard-Moody and Michael Musheno 2003: 3, see also Lowrey-Kinberg et al. 2022). Prior work emphasizes how empathy is discouraged in criminal justice organizations, in part, “as a function of the demographic distance between criminal justice actors and offenders/defendants” (Lynch 2011: 184). We found that prosecutors use their intuitions, judgements, and values stemming from their group membership and life experiences to interpret case facts and decide which defendants should be punished as opposed to rehabilitated. Prosecutors’ varied approaches suggest that divergent outcomes are a reality, not only between offices (e.g., Beckett and Beach 2021; Kautt 2002; Wright and Levine 2017) but also within the same office, which contributes to variability in how justice is manifested. Thus, the “inhabitants” of the prosecutor’s office are centrally important (Ulmer 2019) and a defendant’s fate may turn on which prosecutor is assigned their case, particularly in jurisdictions where there is limited oversight. Consequently, as our findings illustrate, the focal concerns perspective paints an incomplete picture of how prosecutors select charges because it does not account for the role of prosecutor identity and experience.

Importantly, the model we present is descriptive, not prescriptive. We have highlighted how the process of character construction may create space for extralegal considerations to influence charging decisions and how offices with chief prosecutors and community contexts that allow for broad prosecutorial discretion can produce divergent outcomes for similar cases within the same office. However, full adoption of a strict legal model is not necessarily preferable. Joining with Malcom Feeley (1979), we caution against full adoption of the legal model of decision making as greater fidelity to “the Law” could lead to even harsher outcomes, as court actors sometimes use their discretion to “correct” for laws and policies they view as unduly punitive (Kramer and Ulmer 2002; Mitchell et al. 2023; Travis et al. 2014; Ulmer 2014). Discretion is, therefore, an indispensable aspect of prosecutors’ work (Davis 2007: 6).

Future research

Interviewees’ descriptions of their decision-making processes in the context of their workplaces build upon the court communities’ perspective (Eisenstein et al. 1988). Our analysis revealed that the chief prosecutor and community context conditioned the degree to which prosecutors could exercise their discretion. However, “American prosecution is highly localized,” and we examined prosecutorial discretion in just three offices (Miller and Caplinger 2012: 267). Therefore, the model we proposed needs to be tested in a wide range of court communities to determine whether the components we identify are relevant to prosecutorial decision-making in other contexts. Specifically, office dynamics and decision-making processes differ in jurisdictions of different sizes (McWithey 2020; Wright and Levine 2017). Thus, future research should investigate whether the model we describe also applies to large jurisdictions with hundreds of prosecutors and smaller jurisdictions with just a handful of prosecutors.

Additionally, future qualitative research examining offices with a spectrum of approaches to prosecution, including those with explicitly reform-minded goals (Mitchell et al. 2022) could significantly elaborate upon the role of organizational factors in conditioning prosecutorial discretion. Finally, due to the limited case processing data available in the jurisdictions we partnered with, we were unable to compare prosecutors’ stated rationales for charging against actual case processing decisions. Substantiating prosecutors’ accounts of their thought processes in charging against administrative data would be a valuable direction for future research.

Conclusion

Prosecutorial discretion provides a central example of how the law on the books is filtered through individual and organizational lenses and applied according to “ideological, institutional, economic, or practical agendas” (Calavita 2010: 114). Given prosecutors’ extensive discretion to restrain individuals’ liberty, it is important to understand the process by which prosecutors make these important decisions. Our model of charging emphasizes that prosecutors’ identities and experiences shape how they interpret case facts in their evaluation of defendants’ moral character and, consequently, their determination of what a defendant deserves in terms of a disposition and sentence. Yet, organizational factors in the form of the chief prosecutor and court community shape how prosecutors’ charging decisions unfold in practice.

As suggested by the recent controversy around prosecutors’ authority to selectively enforce – or routinely not enforce – the law, discretion is not devoid of ideology or identity. The focal concerns of prosecution and sentencing are often discussed jointly with the conceptualization of courts as communities (Mitchell et al. 2022; Shermer and Johnson 2010; Ulmer 2019). However, prosecutors’ identities and experiences are not typically integrated with these existing theoretical frameworks to make sense of how prosecutors move from a series of limited case facts to initial charging decisions, and subsequent plea offers. Much of the research on prosecution focuses on differences in case outcomes between court communities (Arazan et al. 2019; Franklin 2010; Kautt 2002; Martin et al. 2018) and between defendants within the same court communities (Jawjeong 2016; Steffensmeier et al. 2017). The more limited body of research that examines variation in case outcomes between court actors operating within the same court community, often emphasizes the power of professional and organizational socialization (Bowman et al. 2023; Spohn 1990; Steffensmeier and Britt 2001; Steffensmeier and Hebert 1999; but see Johnson 2014; Wooldredge 2010). In the discussion of organizational conformity, the uniqueness of “inhabitants” is often minimized. Although we find that organizational considerations related to the chief prosecutor and broader court community constrain prosecutors’ processes for achieving their desired outcomes on cases, who prosecutors are – their social identity and experiences – plays an important role in how they interpret case facts and the stories they construct of who a defendant is based on those interpretations. As such, our integration of the individual/internal process of charging with the broader organizational constraints at play during charging reflects the nature of criminal courts as inhabited institutions where law on the books comes to life.

Using prosecutorial charging decisions as a guiding example, our model emphasizes the multilevel nature of discretion in law. Although many disciplines examine decision-making, they tend to do so at a particular level of analysis. Whereas criminologists and legal scholars often focus narrowly on legal- and individual-level factors, social psychologists more often examine how group membership and identity shape decision-making. Sociologists take a more macro-level perspective of how people make decisions within a particular social world. The model of decision-making we propose highlights that each of these levels of analysis interact to create a dynamic decision-making process. Discretion, in any context, is best understood with a multilevel, and cross-disciplinary, perspective.