Police discretion is the latitude officers have to decide, moment by moment, how to enforce the law. Should they stop this car or let it pass? Arrest this person or issue a warning? Use force or try de-escalation? These choices accumulate into patterns that shape who gets drawn into the criminal justice system and who walks away. Research shows that the individual officer who responds to a call can matter enormously: swapping one officer for another changes the likelihood of arrest by a striking margin, even when the underlying situation is identical. Understanding what drives those choices, and where they go wrong, is one of the central questions in policing research.
How Much the Individual Officer Matters
One of the clearest demonstrations of discretion’s power comes from studying what happens when different officers respond to similar calls. Research tracking calls for service found that a one-standard-deviation increase in an officer’s personal arrest propensity corresponds to roughly a 40 percent increase in the likelihood that the call ends in an arrest, or about one additional arrest per hundred calls.1PubMed Central. “Whose help is on the way?” The importance of individual police officers in law enforcement outcomes Put differently, two officers arriving at the same domestic disturbance or the same public intoxication call can reach entirely different outcomes depending on their personal tendencies. The situation matters, but so does the person wearing the badge.
This is not just an academic finding. It means that staffing decisions, shift assignments, and which officer a dispatcher sends to which call all feed into whether someone gets arrested, warned, or referred to services. Departments that treat officers as interchangeable units miss a major source of variation in how justice is administered on the ground.
What Shapes a Decision in the Moment
Officers do not make choices in a vacuum. A constellation of situational factors pushes them toward one outcome or another. Among the best-studied is the suspect’s demeanor. Research consistently finds that how a person acts during an encounter heavily influences what officers do next. Calm suspects are less likely to experience force, while upset, angry, or noncompliant suspects are more likely to face both low-level and more serious force.2Journal of Criminal Justice. Use of force during stop and frisks: Examining the role of suspect demeanor and race The link between hostile demeanor and arrest has held up across multiple re-analyses of classic policing datasets.3Criminology. DEMEANOR, CRIME, AND POLICE BEHAVIOR: A REEXAMINATION OF THE POLICE SERVICES STUDY DATA
This makes intuitive sense: an officer facing a cooperative person in a minor incident has room to exercise leniency, while hostility or aggression narrows the perceived options. But the finding also has an uncomfortable edge. Demeanor is partly a product of the person’s emotional state, mental health, intoxication level, or prior experiences with police. Someone who has been stopped repeatedly without cause may not greet the next officer with calm friendliness. The demeanor effect, while real, can entrench cycles where the people most likely to be policed are also most likely to react in ways that escalate the encounter.
Information officers receive before they even arrive can also skew their judgment. A randomized experiment using a firearms simulator and over 300 active officers found that dispatched information heavily influenced the decision to pull the trigger when confronted with an ambiguously armed person. When that dispatched information was wrong, it led to a significant increase in shooting errors.4Police Quarterly. Dispatch Priming and the Police Decision to Use Deadly Force Officers were, in effect, primed to see a weapon that was not there. This kind of confirmation bias is not a moral failing; it is a well-documented feature of human cognition under stress. But it has lethal consequences.
Racial Disparities in Discretionary Stops and Searches
Discretion becomes especially contentious when racial patterns emerge. A growing body of evidence shows that Black and Latino drivers are treated differently during traffic stops, particularly in the kinds of stops that give officers the most latitude. A systematic review of the empirical literature found that Black and Latino drivers received more traffic tickets than White drivers, that Black individuals were disproportionately targeted for discretionary searches, and that African American drivers were more likely to experience investigative stops involving more invasive inquiries and formal complaints.5PubMed Central. Racial Prejudice and Police Stops: A Systematic Review of the Empirical Literature
The distinction between high-discretion and low-discretion stops is important here. A stop for running a red light is low-discretion: the violation is objective and hard to dispute. A stop for an equipment violation, such as a dim license-plate light or tinted windows, gives the officer far more room to decide whom to pull over. Research on traffic stops has found that Black drivers are more likely than White drivers to be stopped for these high-discretion equipment violations. Community members also perceive these stops differently; Black participants evaluated recordings of high-discretion stops more negatively than White participants did, while evaluations of low-discretion stops did not differ by race.6Journal of Social Issues. Racial Disparities in the Discretionary Context of Traffic Stops: How Organizational Practices Shape Institutional Interactions
One common rebuttal is that racial disparities in stops simply reflect deployment patterns: officers are concentrated in high-crime areas that happen to be disproportionately Black. The evidence partially supports this, but only partially. A study of traffic stops found that nearly half occurred within a quarter mile of crime hot spots and more frequently involved Black drivers, consistent with the deployment theory. However, stops involving Black drivers were still more likely to include discretionary searches even after accounting for proximity to hot spots, and that disparity actually increased with distance from the nearest hot spot.7Race and Justice. The Impact of Police Deployment on Racial Disparities in Discretionary Searches In other words, deployment explains part of the picture, but individual officer decision-making adds its own racial dimension on top.
What Happens When Discretion Is Removed
If discretion introduces bias and inconsistency, the obvious fix seems to be removing it. Mandatory arrest laws in domestic violence cases are the most studied example. These statutes require officers to make an arrest whenever they find probable cause of intimate partner violence, eliminating the choice to warn, mediate, or walk away. The results have been more complicated than reformers hoped.
An influential early study found that mandatory arrest laws actually increased intimate partner homicides, potentially because victims became less willing to call police if they knew an arrest was guaranteed, and because some abusers retaliated violently after being arrested and released.8Journal of Public Economics. Does the certainty of arrest reduce domestic violence? Evidence from mandatory and recommended arrest laws That finding alarmed policymakers. A later reanalysis challenged it, finding no evidence that mandatory arrest laws increased homicides. Instead, the reanalysis found some evidence that discretionary arrest statutes, which allowed but did not require arrest, were associated with decreased spousal homicides.9Journal of Public Economics. Revisiting the effect of warrantless domestic violence arrest laws on intimate partner homicides
The disagreement between these studies has not been fully resolved, and the domestic violence field continues to debate how much discretion officers should retain. But both sides of the debate share an implicit recognition: discretion is not simply good or bad. Removing it can have unintended consequences, especially when the people the policy aims to protect have their own strategic reasons for not wanting police involvement. This is one of those areas where the evidence genuinely points in multiple directions, and anyone who tells you the answer is simple is selling something.
Mental Health Encounters and the Gray Zone
Perhaps nowhere is discretion more consequential than in encounters with people experiencing mental health crises. Officers responding to these calls face a choice with no clean option: arrest the person, transport them to a hospital, or try to resolve the situation informally on the spot. A study of mental health-related calls in Chicago found that informal interventions, with no legal action or hospitalization, were the most common resolution. Hospital transports happened more often than arrests, but both were less frequent than the officer simply talking the person through the moment and leaving.10PubMed Central. Everyday police work during mental health encounters: A study of call resolutions in Chicago and their implications for diversion Researchers described this as the “gray zone” of mental health policing, a space where formal categories like arrest and hospitalization do not capture what actually happens.
Pre-booking diversion programs, where officers redirect people to services instead of arresting them, have been introduced in many jurisdictions to formalize better options in this gray zone. A systematic review found that while most studies did not show significantly different arrest rates between officers trained in diversion and those who were not, trained officers consistently made more referrals to mental health services.11PLoS ONE. Evidence for the effectiveness of police-based pre-booking diversion programs in decriminalizing mental illness: A systematic literature review A separate review found that police diversion programs were associated with reduced reoffending and lower costs, though they did not consistently improve the person’s behavioral health outcomes.12Criminal Justice and Behavior. Police Diversion at Arrest: A Systematic Review of the Literature
The picture that emerges is cautiously optimistic but incomplete. Training officers to use their discretion differently, steering them toward referrals rather than arrests, does seem to change the flow of people through the system. Whether it changes their long-term well-being is less clear. The mental health system those referrals lead to is often underfunded and overloaded, which limits how much any police-level intervention can accomplish.
Discretion in Juvenile Justice
When officers encounter young people suspected of minor offenses, the stakes of discretion are especially high. A formal arrest can saddle a teenager with a record that follows them into adulthood, while a warning, a call to parents, or a referral to a youth program might produce a better outcome for everyone. A systematic review of police-led diversion programs for youth found a modest but real benefit: diverted low-risk youth reoffended at a rate of roughly 44 percent, compared to about 50 percent for youth processed through traditional channels.13PubMed Central. Police‐initiated diversion for youth to prevent future delinquent behavior: A systematic review That six-point difference adds up when spread across thousands of encounters per year.
Australia’s experience with expanding juvenile diversion options offers a useful cautionary note. Legislation gave officers a wider range of discretionary and diversionary tools to accommodate individual differences in young offenders. The intent was flexibility, and officers did use the tools. But an unintended consequence was greater unpredictability of outcome, meaning that similar young people in similar situations could receive wildly different responses depending on which officer they encountered.14International Journal of Police Science & Management. Policing Young Offenders: What Role Discretion? More options did not automatically mean better or fairer outcomes. Without clear guidelines about when to use which option, expanding the menu of choices can amplify the same inconsistencies that discretion always produces.
Training Officers to Make Better Split-Second Decisions
If discretion is unavoidable, the question becomes whether training can make officers use it better, especially in the highest-stakes moments. Research using police firearms simulators has explored this in detail. One pilot study tested a training method that combined scenario-based practice with physiological stress management. Officers who received the intervention showed significantly better physiological control, situational awareness, and overall performance, and made more correct use-of-force decisions than officers in a control group.15Sage Open. A Training Method to Improve Police Use of Force Decision Making
Even something as seemingly mechanical as where an officer points their weapon can change outcomes. A randomized experiment with over 300 officers found that taking a lower muzzle position when confronting an ambiguously armed person significantly improved shoot/no-shoot accuracy without meaningfully sacrificing response time to genuine threats.16Police Quarterly. “Engineering Resilience” Into Split-Second Shoot/No Shoot Decisions: The Effect of Muzzle-Position The idea is to build small buffers into the decision environment, giving the officer a fraction of a second more to assess the situation before committing to lethal force. Researchers have described this as “engineering resilience” into the encounter rather than relying solely on the officer’s judgment under extreme stress.
These findings are promising, though they come from simulator environments. Real encounters involve more chaos, more variables, and fewer controlled conditions. Still, the direction of the research is clear: how officers are trained to use their discretion in high-pressure moments can be improved with deliberate practice and procedural changes. Most departments do not invest nearly enough in this kind of training, often offering only a few hours of simulator time per year.
Body Cameras and Predictive Algorithms
Technology has been billed as a check on discretion, but the evidence is more mixed than the marketing suggests. Body-worn cameras were widely adopted in the 2010s with the expectation that being recorded would make officers more restrained and accountable. A Campbell systematic review found that body cameras reduced citizen complaints against officers by about 17 percent. But the same review found that cameras did not significantly affect arrest behaviors or officers’ self-initiated activities like discretionary stops.17PubMed Central. Body‐worn cameras’ effects on police officers and citizen behavior: A systematic review The reduction in complaints could mean officers behaved better, or it could mean civilians filed fewer complaints knowing they were on camera too. The studies could not disentangle those explanations.
Predictive policing algorithms, which use historical crime data to direct officers to geographic “hot spots,” present a different kind of interaction with discretion. In theory, these tools constrain where officers patrol and what they prioritize. In practice, officers retain enormous latitude over whether to follow the algorithm’s recommendations. Ethnographic research on predictive policing found that compliance varied widely by division, shift, and individual officer. During busy periods, officers spent almost no time in the algorithmically designated zones. Even during slow periods, some officers would drive to a predicted hot spot, briefly check in on their laptop, look around, and leave, while others ignored the boxes entirely.18PubMed Central. Technologies of Crime Prediction: The Reception of Algorithms in Policing and Criminal Courts The algorithm suggested; the officers decided. Technology did not eliminate discretion so much as add another input for officers to weigh or discard.
Legal Accountability and Its Limits
The legal framework around police discretion creates a paradox. The Supreme Court’s 1968 decision in Terry v. Ohio validated the practice of stop and frisk, requiring only that an officer have “reasonable suspicion” to justify the action.19Annual Review of Law and Social Science. The Law and Social Science of Stop and Frisk Reasonable suspicion is a deliberately flexible standard, lower than probable cause, and it effectively grants officers wide discretion to initiate encounters based on factors that are difficult to challenge after the fact.
When officers are sued for constitutional violations, they can invoke qualified immunity, a defense that hinges on whether a “reasonable officer” would have acted the same way under the circumstances.20Policing: An International Journal. The reasonably unreasonable police officer: A paradox in police civil liability jurisprudence The practical effect is that officers are rarely held personally liable unless their conduct violates a clearly established legal precedent, and courts have interpreted “clearly established” very narrowly. An officer who does something objectively harmful but novel, meaning no prior court case addressed that exact scenario, often receives immunity. This structure means discretion is legally insulated in ways that make external accountability difficult.
Procedural Fairness and Public Trust
Even when discretionary decisions produce the same outcome, the process by which officers exercise that discretion affects how the public perceives them. Survey research with over 1,600 residents of a metropolitan area found that procedural justice evaluations, whether officers treated people with respect, explained their actions, and allowed people to voice their concerns, were a primary driver of trust in police. Neighborhood context mattered too; residents of disadvantaged neighborhoods tended to start from a lower baseline of trust. But across settings, how officers used their authority mattered at least as much as what they decided.21Crime & Delinquency. Trust in the Police
This finding reframes the conversation about discretion in a useful way. Much of the debate focuses on outcomes: who gets arrested, who gets searched, who gets shot. Those outcomes matter immensely. But the process of discretion, the way an officer communicates during a stop, whether they explain why they pulled someone over, whether they listen before acting, shapes public willingness to cooperate with and support the police. Departments that invest in procedural fairness training are not just being polite; they are building the institutional legitimacy that makes effective policing possible in the first place.
Fatigue, Overtime, and the Erosion of Good Judgment
Officers are not making discretionary choices in ideal cognitive conditions. Shift work, mandatory overtime, and chronic sleep disruption are endemic to policing. Research on long work hours in law enforcement has found that fatigue-related impairments to officer performance and decision-making generate unexpected social and economic costs.22American Journal of Industrial Medicine. Impact of long work hours on police officers and the communities they serve An officer nearing the end of a sixteen-hour shift is making discretionary calls with degraded attention, slower reaction times, and reduced emotional regulation. The training findings described earlier, showing that physiological control and situational awareness improve use-of-force decisions, imply that anything undermining those capacities makes bad outcomes more likely.
This dimension of discretion rarely enters public debate, which tends to focus on bias, policy, and accountability. But the officer who makes a terrible split-second decision at 3 a.m. after twelve hours on duty is operating under conditions that would impair anyone’s judgment. Staffing levels, shift schedules, and overtime policies are, in a real sense, use-of-force policies. Departments that run officers ragged and then expect flawless discretionary judgment are setting up both officers and the public for failure.