Lum and Nagin

Reinventing American Policing by Cynthia Lum and Daniel S. Nagin

Abstract

Two principles form the bedrock for effective policing in a democratic society:
- Crimes averted, not arrests made, should be the primary metric.
- Citizens’ views about the police and their tactics matter independently of police effectiveness.
Both principles are important and must be balanced case-by-case.
These principles should guide efforts to reinvent American policing.
Seven steps are essential to reinvent democratic policing:
1. Prioritize crime prevention over arrest.
2. Monitor citizen reactions to the police and report results.
3. Reform training and redefine the “craft of policing.”
4. Recalibrate organizational incentives.
5. Strengthen accountability with greater transparency.
6. Incorporate the analysis of crime and citizen reaction into managerial practice.
7. Strengthen national-level research and evaluation.

Introduction

Deadly use of force by police has led to protests and questioning of police tactics.
Citizens and politicians have called for changes to make the police more accountable and transparent.
President Obama convened a task force to make recommendations for police reform.
There have been upticks in crime in some cities.
A recurring question: How can police maintain community trust and confidence while effectively preventing crime and keeping citizens safe?
Both objectives are fundamental and neither should trump the other.
Reorienting police practices requires fundamental changes in the functions, values, and operations of law enforcement.
A seven-point blueprint for reinventing American policing, guided by two principles grounded in research and experience, is put forth.

Principle 1: Crime Prevention Is Paramount

Crimes averted, not arrests made, should be the primary metric for judging police success.
This principle follows from Cesare Beccaria’s observation that “it is better to prevent crimes than to punish them.”
Punishment is costly to society, the individual, and the police.
Arrests signify a failure of prevention.
Principle 1 does not imply that police should stop making arrests, as they are important to bring perpetrators to justice.
Proactive prevention activities are more effective than reactive arrests.
Proactive policing focuses on people, places, times, and situations at high risk of offending, victimization, or disorder.
Proactive policing contrasts with reactive approaches by addressing problems before they beget further crimes, often without emphasizing arrest.

Principle 2: Citizen Reaction Matters

Police in democracies are responsible not only for preventing crime but also for maintaining their credibility with all segments of the citizenry.
Citizen reaction to what the police do is critical, independent of their success in preventing and solving crime.
Citizen trust and confidence may facilitate police effectiveness.
The overriding objective should be to create a safe democratic society, not a safe police state.

Lawfulness as a Constraint

Lawfulness is treated as a constraint, meaning police should not engage in illegal practices.
Illegal practices should not be condoned even if they advance principles 1 or 2.
Police training, procedures, and metrics of success have been tailored to achieving legality but with little attention paid to their implications for advancing principles 1 and 2.

Interdependence and Complexity

The three core functions—preventing crime, bringing perpetrators to justice, and maintaining credibility and trust—are significant and interdependent.
Ineffectiveness in apprehending perpetrators may erode preventive effectiveness.
Commitment of time and resources to apprehending perpetrators may come at the expense of crime prevention.
Maintaining trust may be tied to the ability of the police to prevent crime and bring perpetrators to justice, but trust may be eroded by too many arrests for minor crimes.
Some police-citizen encounters are hostile through no fault of the officer.
Policing's complex organization and many responsibilities not directly related to crime control add difficulty in pursuing these two principles.
How police perform in all of their responsibilities may influence their effectiveness in preventing crime and maintaining public trust and credibility.

Blueprint Summary

The seven-point blueprint is anchored in relevant research evidence.
Section I discusses the function and costs of arrest.
Section II analyzes evidence on police effectiveness in preventing crime and maintaining order, emphasizing proactive problem-solving rather than reactive response and arrest.
Section III discusses evidence on citizen reactions to police activity and strategies to improve trust and confidence.
Section IV discusses controversial strategies that challenge the two principles: broken windows policing and stop, question, and frisk.
The essay concludes with seven proposals for reinventing American policing.

Proposal 1: Prioritize Crime Prevention over Arrest

Police should focus efforts, reforms, and resources on activities that prevent crime and avert the need for arrests and their ensuing costs.

Proposal 2: Monitor Citizen Reactions

Create and install systems to monitor citizen reactions to the police and routinely report results.

Proposal 3: Reform Training and Redefine Policing

Reform training and redefine the “craft” of policing to include prevention and community relations.

Proposal 4: Recalibrate Organizational Incentives

Recalibrate organizational incentives to incorporate measures of effective crime prevention and maintenance of citizen confidence and support.

Proposal 5: Strengthen Accountability with Transparency

Strengthen accountability with more transparency by increasing the availability of data and policies related to police-citizen interactions.

Proposal 6: Incorporate Analysis

Incorporate analyses of crime and citizen reaction into managerial practice.

Proposal 7: Strengthening Research

Strengthening National-Level Research and Evaluation: A robust infrastructure of research and its dissemination is essential if major advances are to be made.

The Function and Costs of Arrest

Arrest serves to bring perpetrators to justice and to deter crime through the threat of apprehension.
The legal authority to arrest is a defining feature of policing, institutionalized in deployment, training, metrics, and organization.
Arrests are costly for society, the person arrested, and the police.
Emphasizing crime prevention tries to avert the need for arrest.
Prevention averts losses and suffering, avoids expense to society, and reduces costs associated with offenders’ reentry into society.
Police cannot prevent all crime, making some arrests inevitable.
Most arrests are not for serious crimes; in 2013, only 4.3% were for violent offenses and 13.8% for property offenses.
The vast majority of apprehensions are for less serious offenses, open to greater discretion, and often result from policy choices.
Arrests for minor infractions do not come without costs.
Time spent processing an arrest takes police off the street and away from activities that might prevent serious crime.
Informal surveys suggest that between 2 and 4 hours are spent by officers per arrest for minor offenses.
In 2013, there were 9.2 million arrests for nonindex, part II offenses.

Criminal Justice Costs

Police officers’ time and lost crime prevention opportunities are only the beginning of the criminal justice costs associated with arrest.
Few arrestees serve time in state prisons because of the low severity of the arrest charges, but many end up serving time in local jails, particularly before trial, including for cases ultimately dismissed.
Less well recognized is the comparable growth in jail populations.
In 1980, the jail population was about 182,000. By 2011, it was over 730,000.
Adjusting for general population growth, the increase in the jail population was about the same as that for state and federal prisons.
Nationwide this translates into billions of dollars per year of increased spending by local governments to pay for a half-million-person increase in the jail population.
The increase in arrests for drug possession was one driver of the growth in jail populations.
Chauhan et al. (2014) report that about 50 percent of misdemeanor arrests in urban areas result in a conviction and about a third of convictions involve a sentence with incarceration.
Between 1983 and 2002, the share of inmates in jails held for violent and property crimes declined from 69.3 to 49.4 percent.
By contrast, the percentage held for drug offenses grew from 9.3 to 24.7 percent.
In 1983, possession accounted for 4 percent of the jail population. By 2002, that had grown to 10.8 percent.
A more recent Bureau of Justice Statistics (BJS) sponsored survey suggests that the share of jail inmates incarcerated for drug offenses has remained at about 25 percent (Beck et al. 2013, table 9, p. 19).
The percentage of jail inmates held for public-order offenses remained steady at about 20 percent from 1983 to 2002, but the number of people being held grew enormously because of the growth of the total jail population.
Assuming that 20 percent of the jail population are incarcerated for public-order offenses, the jail population for drug and public-order offenses grew from about 66,000 in 1983 to about 327,000 in 2011.
Corrected for the growth in the general population, that is a fourfold increase over three decades.

Costs for Arrestees

Other costs include the consequences of misdemeanor arrests for arrestees themselves.
These include the costs of legal representation, bail, and time spent, possibly at the expense of work, attending legal hearings (Kohler-Hausmann 2013).
Misdemeanor arrests also affect eligibility for student loans, public housing, and professional licensure (Natapoff 2012).
A misdemeanor arrest record may also carry a stigma in the labor market (Uggen et al. 2014).

Benefits of Apprehension

Balanced against the cost of arrests are two benefits of apprehension: bringing perpetrators of crime to justice and reinforcing police capacity to deter crime through the threat of apprehension.
Crime clearance rates for serious crimes have remained largely unchanged for more than four decades despite significant fluctuations in the index crime rate (Braga et al. 2011).
That stability reflects the reality that most crimes are solved by the apprehension of the perpetrator at the scene of the crime or by eyewitness identification (Greenwood, Chaiken, and Petersilia 1977).
Capture at the scene might be improved by more rapid responses to service calls, but efforts to do this have not been successful (Kansas City Police Department 1977; Spelman and Brown 1981).
Eck (1992) concluded that “it is unlikely that improvements in the way investigations are conducted or managed have a dramatic effect on crime or criminal justice” (p. 33).

Deterrence

Arrest may be viewed by the police and citizens as providing a deterrent to the person arrested and to others.
Criminologists refer to the former as “specific” deterrence and the latter as “general” deterrence.
Concerning specific deterrence, there is some evidence that being arrested causes juveniles, especially those who are inexperienced offenders, to increase their perception of the risk of apprehension and thereby deter minor offending (Smith and Gartin 1989; Nagin 2013).
There may also be a specific deterrent effect of arrest for certain types of domestic violence offenders (e.g., those who are employed) but not others (Sherman and Eck 2002).
Concerning general deterrence, the story is more complicated.
In the early years of empirical deterrence research, investigators correlated or regressed crime rates on arrests per crime.
These studies found evidence of deterrent-like effects but had fatal statistical flaws that made them uninterpretable (Blumstein, Cohen, and Nagin 1978).
Another flaw first pointed out by Cook (1977), and later elaborated on by Nagin, Solow, and Lum (2015), is that arrests per crime, like its close cousin the clearance rate, measures only apprehension risk for crimes that actually occur.
Would-be offenders do not pick targets for crime randomly.
Deterrence research shows that even though apprehension itself may not have strong deterrent effects, perceptions of apprehension risk figure heavily in offender decision making.
Targets with a low risk of apprehension are preferred (Wright and Decker 1994; Nagin 1998).

Targeted and Proactive Policing

The targeted and proactive policing strategies are intended to make crime targets less attractive by heightening apprehension risk.
Nagin, Solow, and Lum (2015) illustrate that targeted, proactive policing can strategically alter apprehension risk across opportunities for crime in a way that reduces both crime and arrests.
Reductions in crime and arrests may be accompanied by a reduction in the clearance rate.
This may occur because offenders are victimizing only the remaining small fraction of total criminal opportunities that are characterized by very low risk of apprehension.
Arrest-based metrics such as the clearance rate are flawed measures of police effectiveness in preventing crime.

Conclusion on Arrests

Some circumstances, of course, require police to make an arrest, particularly if the arrestee is suspected of committing or is known to have committed a serious crime.
There are significant costs to arrests and that their crime control effectiveness is often not demonstrable.
These costs can be avoided and affect crime control by preventing crimes in the first place (Durlauf and Nagin 2011).

Preventing Crime and Disorder

Nagin (2013) distinguishes two distinct crime control functions of the police.
The first requires police to act as “apprehension agents” for perpetrators of crimes.
The second involves the police preventing crimes from occurring in the first place.
In this role police can be described as “sentinels,” who provide guardianship for potential targets, thereby mitigating opportunities for crime.
Much of the literature on how police prevent, control, and deter crime and disorder (consistently with principle 1) focuses on their sentinel role.
One area of research addresses the effect of police numbers on crime rates.
A second examines the effects of police deployment tactics on crime—especially in targeted geographic areas and on repeat offenders.
The third concerns police efforts to reduce crime by proactive policing against disorder, sometimes called broken windows or zero-tolerance policing.
This section focuses on the latter two literatures rather than on changing the number of police.
The consistent conclusion of research on police numbers and crime is that increases in police presence decrease crime, and decreases in police presence are associated with crime increases (Sherman 1992; Apel and Nagin 2011; Nagin 2013).
Police presence—an essential element of the sentinel role of policing—can affect crime rates.

A. Police Deployment Tactics to Deter and Prevent Crime

Proactive targeting more effectively prevents crime than do traditional react, investigate, and arrest approaches.
“Hot spots” policing provides one example of a targeted, proactive deployment strategy for which there is good evidence of effectiveness.
Hot spots policing involves increased police presence at places in which crime concentrates.
The idea stems from Sherman, Gartin, and Buerger (1989) when they found that only 3 percent of addresses and intersections (“places”) in Minneapolis produced 50 percent of all calls to the police.
Numerous studies since then have found similar evidence of hot spots (e.g., Weisburd and Green 1995; Eck, Gersh, and Taylor 2000; Roncek 2000; Weisburd 2015) and their stability over time (Weisburd et al. 2004).
The rationale for concentrating police in hot spots is to create a prohibitively high risk of apprehension by increasing the level of guardianship at a place with high levels of criminal opportunities.
Sherman and Weisburd (1995) conducted the first test of the efficacy of concentrating police resources on crime hot spots.
In this randomized experiment, hot spots in the experimental group received, on average, a doubling of police patrol intensity compared with hot spots in the control group.
Declines in total crime calls in experimental spots ranged from 6 to 13 percent compared to controls.
In another randomized experiment, Weisburd and Green (1995) found that hot spots policing was similarly effective in suppressing disorder at drug markets without leading to displacement of crime elsewhere.
Braga, Papachristos, and Hureau (2014) summarize findings from 19 experimental or quasi-experimental evaluations of hot spots policing.
Twenty of 25 tests from these evaluations found significant reductions in crime and disorder, leading them to conclude, “Hot spots policing strategies generate small but noteworthy crime reductions, and these crime control benefits diffuse into the areas immediately surrounding the crime hot spots” (p. 633).
Subsequent evidence indicates that police do not have to stay in hot spots all day; intermittent and unpredictable stays of 12–15 minutes may optimize their residual deterrent effects (Koper 1995; Telep, Mitchell, and Weisburd 2014).
This is shown by the Evidence-Based Policing Matrix, in which Lum and her colleagues gather and annually update evaluations of police-involved interventions to deter and prevent crime that reach at least a moderate threshold of methodological rigor.
The Matrix shows that the greatest concentration of statistically significant positive effects is for interventions that increase police presence at crime hot spots and are focused, tailored, and proactive.
When police geographically target high-crime places and do so with tactics tailored to the problems at hand, they are more effective in preventing crime (Weisburd and Eck 2004; Braga and Bond 2008; Weisburd et al. 2010; Taylor, Koper, and Woods 2011; Braga, Welsh, and Schnell 2015).
Tailored and focused approaches often encompass proactive “problem-solving” or “problem-oriented” strategies.
Originally articulated by Goldstein (1979, 1990), problem-oriented policing involves careful police analysis of crime problems with direction of police and other resources to resolving underlying factors that contribute to those problems.
Eck and Spelman (1987) developed the “SARA” model, an acronym highlighting the process of problem solving (“scanning,” “analysis,” “response,” and “assessment”).
Changes might include improving lighting, closing problem bars, increasing closed-circuit television surveillance, or adjusting spatial and temporal patterns of police patrol.
These approaches are wide ranging and may include the use of situational crime prevention (Clarke 1980, 1995, 1997) and “crime prevention through environmental design” ( Jeffery 1971; Newman 1972; Crowe and Zahm 1994; National Crime Prevention Council 1997).
These approaches require the police to play a role in understanding the causes and nature of criminal opportunities at places and then engage in “opportunity mitigation.”
Tailored and problem-solving approaches can also include what Mazerolle and Ransley (2005, 2006) describe as “third-party” policing” in which police collaborate with regulatory authorities to alter or eliminate malleable features of the social and physical environment that facilitate crime.
Buerger and Mazerolle (1998) describe third-party policing as “police efforts to persuade or coerce organizations or non-offending persons, such as public housing agencies, property owners, parents, health and building inspectors, and business owners to take some responsibility for preventing crime” (p. 301).
Law enforcement can collaborate with private security officers, who in some places concentrate in sufficient numbers to disrupt opportunity structures that promote crime (Cook and MacDonald 2011).
Finally, proactive approaches can involve targeting problem people, a strategy sometimes described as “focused deterrence,” which has its origins in Operation Ceasefire, a Boston-based intervention designed to prevent gun violence among rival gangs (Kennedy, Piehl, and Braga 1996; Kennedy 2011).
Tactics involve a “pulling all levers” approach.

Focused Deterrence

Pulling all levers thus requires police to collaborate with multiple nonpolicing agencies (Braga et al. 2001; Braga, Kennedy, and Tita 2002).
A review of 10 focused deterrence studies that targeted “hot people” concluded that nine significantly reduced crime (Braga and Weisburd 2012).
The authors, however, were cautious about the strength of the overall findings because of weaknesses in the study designs.
Like other problem-solving strategies, focused deterrence approaches vary depending on the types of offenders and crimes being addressed.
More recent studies support focused deterrence. Braga, Hureau, and Papachristos (2014), Groff et al. (2014), and Papachristos and Kirk (2015) found significant effects of focused deterrence on violent crime.
Many forms of proactive and problem-oriented policing that have been shown to be effective involve police operating as sentinels rather than as apprehension agents, with a focus on “hot” places or people.
In hot spots where crime opportunities are abundant, the police may be the sole sentinels, as these places often lack effective guardianship by residents.
Reactive arrest is not the only option.
Police can affect crime when they act in ways that mitigate crime opportunities at places or with individuals, thus discouraging motivated offenders from acting on criminal opportunities in the first place.

B. Broken Windows Policing and Stop, Question, and Frisk

Being proactive involves making discretionary decisions about groups to focus on, problems to tackle, and behavior of police that can introduce bias, inequity, illegality, or questionable practices.
Honoring principles 1 and 2 requires that police not engage in practices that are illegal.
Proactive approaches could also lead to more, not fewer, arrests.
One area of proactive policing in which police roles as apprehension agents and sentinels become blurred is cases in which police make large numbers of arrests, usually for minor crimes such as disorderly conduct or drug possession, with the ultimate goal of preventing serious crime.
Sometimes this approach is used at crime hot spots.
While maintaining order and reducing public nuisance have been a long-standing function of the police, a seminal article by Wilson and Kelling (1982) tied this function to the prevention of serious crime.
Wilson and Kelling likened disorder and public nuisance to a “broken window” that attracts further acts of vandalism or crime.
They argued that when left unchecked, disorder and public nuisances may likewise set the stage for serious crime.
What has come to be called “broken windows” or “zero-tolerance” policing is premised on this theory that proactive policing against disorder and public nuisances pays larger dividends in prevention of serious crime.
While Kelling and others have argued that policing disorder can involve problem-solving strategies and not just making large numbers of arrests (Kelling and Coles 1996; White, Fradella, and Coldren 2015), the latter form of broken windows policing has been widely adopted and has been the subject of much controversy.

Zero Tolerance Policies

Starting in the early 1980s and particularly in the 1990s, zero-tolerance policies grew popular with police departments, the general population, and elected officials.
This led to their diffusion to the United Kingdom and elsewhere (Jones and Newburn 2007; Martinez 2011) and can be seen in the increase of arrests for minor crimes since 1980.
Clearance rates for serious crimes have remained largely unchanged for decades, and arrest rates for part I index crimes have consistently followed crime rate patterns more generally.
*The data used by Kelling and Sousa and by Harcourt and Ludwig cover a period of numerous innovations in policing heralded by the “Compstat” era ushered in by William Bratton.
Several classes of less serious crimes have an enormous growth in arrests over the past three decades. Arrests rose disproportionately for drug possession. The arrest rate for drug possession or use doubled from 1980 to 2012.
Two other cat- egories of nonindex arrests also grew substantially.
The simple assault arrest rate increased by 77.6 percent, which might be explained, in part, by an increase in mandatory arrest policies for domestic violence.
*A much larger, catchall category labeled as “all other offenses (except traffic)” in- creased by 40.7 percent.
There is not a close correspondence between the ar- rest rates for nonindex offenses and for part I crimes.
*While the two rates have tracked each other since 2000, from 1980 to about 2000, the nonindex arrest rate rose even though the total index crime rate fell.
Another indicator of the increased emphasis on arrests for nonindex crimes is an increase in the ratio of part II arrests to part I arrests.
Figure 3 shows the ratio of the part II arrest rate (as adjusted in fig. 2) to the part I arrest rate. In 1980, the ratio was 3.4. By 2006, it was 5.6. Since then it has declined to 4.6 but remains well above the 1980 level.

Uncertainties

There is no uncertainty about the association between disorder and crime; places that have more disorder also have more serious crime.
What is uncertain is whether the correlation of crime and disorder across places and over time is a reflection of a common set of underlying causes such as poverty, social disorganization, or even ineffective policing, or whether the relationship is causal—specifically, that disorder leads to serious crime.
Empirically distinguishing these alternative explanations is difficult.
It is practically and ethically impossible to subject communities to different levels of disorder in order to observe whether manipulation of disorder affects levels of crime.
Some tests of the effects of policing interventions against disorder pro- vide indirect evidence. These studies come in two forms.
Kelling and Sousa (2001) used precinct-level data for New York City to examine whether higher rates of misdemeanor arrest were associated with lower levels of crime after taking account of other char- acteristics of the precinct.
They concluded that aggressive misdemeanor arrests substantially reduced serious crime.
Corman and Mocan (2002), who also analyzed New York City data, reached a similar conclusion. Rosenfeld, Fornango, and Rengifo (2007) found only a modest effect. Fagan and Davies (2003) and Harcourt and Ludwig (2005) found no ev- idence of an effect. The Harcourt and Ludwig study is notable because it includes an analysis of the data used by Kelling and Sousa (2001).

Hot Spot Policing and aggressive arrests

This then brings us to the hot spots policing experiments in which or- der maintenance policing and aggressive arrests for misdemeanors may have played a part.
The examples by Braga et al. (1999) and Braga and Bond (2008) do include aggressive misdemeanor arrest policing as part of a plethora of proactive police activities to reduce crime at hot spots. Both found crime prevention effects.
More speculative inferences, however, are possible with mediation analysis, which can be used to examine how each component of the total package of treatments might moderate the effects of the treatment (MacKinnon, Fairchild, and Fritz 2007).
*In an experimental study comparing problem-oriented po- licing with directed patrol, Taylor, Koper, and Woods (2011) found that problem-oriented policing created significant effects with no increase in field stops.
Most recently, Braga, Welsh, and Schnell (2015) conducted a system- atic review of high-quality research evaluating disorder policing.
Most recently, Braga, Welsh, and Schnell (2015) conducted a system- atic review of high-quality research evaluating disorder policing. They conclude that the strongest effects were generated by community and problem-solving interventions designed to change social and physical dis- order conditions at particular places and that aggressive order mainte- nance strategies targeting individual disorderly behaviors did not gener- ate significant crime reductions.
Another proactive policing approach sometimes associated with bro- ken windows policing is “stop, question, and frisk” (SQF).
(SQF) was the source of the landmark Supreme Court decision in Terry v. Ohio, 392 U.S. 1 (1968), that laid out the constitutional standard required for its use: police officers have to have a “reasonable suspicion” that the person or persons being stopped had committed a crime or was about to do so and that the individual was armed. In theory, SQF is distinct from broken windows policing in that its focus is not on preventing disorder. However, SQF has been coupled with broken windows policing in both practice and critique.

Stop, Question, and Frisk

In practice the tactic has been used to search not just individuals suspected of being armed but also those suspected of carrying drugs.
*Thus, it is not surpris- ing that SQF’s continued use and justification under the Terry standard remains controversial from both legal and social perspectives.
Most of the research is based on data from New York City. Results are mixed, but more studies conclude that it is effective than not. However, all of these studies are based on nonexperimental data and are vulnerable to problems inherent in the use of such data.
*The study by Weisburd et al. (2016) successfully satisfied standards laid out by Nagin and Weisburd (2013) for credible causal inference finding SQF was effective in preventing crime but that it does not account for other policing tactics used in conjunction with SQF that might contribute to crime prevention but without the noxious effects of SQF.

Conclusion on aggressive policing tactics

Suggesting that aggressive policing tactics involving stopping and questioning citizens and, when appropriate, arresting them have no place in policing in specific police tactics to reduce firearms violence (Koper and Mayo-Wilson 2006).
*These tactics were tailored to mitigate opportunities for firearms carrying in crime hot spots and were found to have positive effects (Sherman, Shaw, and Rogan 1995; McGarrell et al. 2001).
Most importantly, some might have very negative reactions from citizens and are viewed by many in minority communities as oppressive and contributing to police-community tensions (Straub 2008).

Maintaining Trust and Confidence of Citizens

The mandate to maintain trust and confidence of citizens, like the mandate to prevent and address crime, is not explicitly established in law, the Constitution, or standard police operating procedures.
Both mandates arise from the view that police in advanced democracies are responsible for the public’s well-being and welfare.
Democratic governance cannot allow police unfettered authority to achieve security; rather, police must do so in a manner that not only is within legal bounds but also is acceptable to citizens.
Police must be accountable, transparent, open, responsive, reliable, and fair.
Citizen reaction can take various forms, including complaints, lawsuits, negative survey responses, protests, noncompliance, resistance, and defiance to commands.

Trust and Confidence

Citizen trust and confidence in the police matter in their own right in democracies and, like crime prevention, must be gauged as a performance metric.
Citizens who distrust the police may not report situations that might lead to crime.
Police tactics and strategies should be judged in terms both of citizen reactions to policing activities and of their effectiveness.
Neither should be seen as trumping the other.

Survey results

The BJS Police-Public Contact Survey (PPCS) data and Gallup polls on attitudes toward the police are two important measures of citizen re- actions to police services.
They convey a consistent and important story: Police enjoy high levels of overall satisfaction, but levels of satisfaction are consistently lower among black and Hispanic communities, and especially among younger minority males who have greater contact with the police.
*However, differences between the attitudes of whites and blacks were stark, as table 1 shows. Among whites the police received the third-highest ranking among 17 institutions; only the military and small businesses ranked higher. Among blacks, police drop to seventh. Both groups have far greater confidence in the police than in the criminal justice system generally.

Traffic and Street Stops

*86 percent of persons involved in traffic stops believed that the police behaved properly and treated them with respect. Only 10 per- cent believed that police behaved improperly. In the case of street stops, these two percentages were, respectively, 66 percent and 25 percent.
*Regardless of the reason for the traffic stop, black (67 percent) and Hispanic (74 percent) drivers were less likely than white drivers (84 percent) to believe the rea- son for the stop was legitimate.
*Similar findings of differential perceptions of police behavior across races are found in the 1999, 2002, 2005, and 2008 PPCS surveys (Durose, Schmitt, and Langan 2005; Engel 2005; Durose, Smith, and Langan 2007; Eith and Durose 2011; Langton and Durose 2013).
*Studies examining both specific jurisdictions and national-level data show that blacks and His- panics are stopped, ticketed, and searched at higher rates than whites (Gaines 2006; see also Fagan and Davies 2000; Walker 2001; Lundman and Kaufman 2003; Reitzel and Piquero 2006; Warren et al. 2006), even when they are at no greater risk of carrying contraband (Engel and Calnon 2004).

Procedural Justice

Another dimension of citizen confidence in the police is whether people distinguish their overall confidence in police from their evaluations of specific police tactics such as the results on the approval of New York City residents of police use of SQF where only 45 percent of all residents approved. However, approval for the overall performance of the New York City Police Department was materially higher at 57 per- cent approved of this performance.
*Some research indicates that changes in policing and police policy can affect citizen trust and confidence.
Research on procedural justice tests the procedural justice theoretical perspective that emphasizes the importance of people perceiving that they have been treated fairly and with dignity even if the outcome is not to their liking.

Legal Compliance

A recent review of the literature on procedural justice and legal compli- ance by Nagin and Telep (forthcoming), concluded that while much research demonstrates an association between perceptions of pro- cedurally just treatment and legal compliance, very little research estab- lishes a connection between actual treatment in this regard and legal compliance.
Community policing review review of community policing interventions by Gill et al. (2014) found a small number of studies with even moderate methodo- logical rigor citing that community-oriented activities improve citizen perceptions of police but that the effects of these strategies on crime control are small and inconclusive.

Citizen surveys are important because

They create feedback loops for agencies to better target policy changes.
*Increase openness, transparency, and legitimacy, all hallmarks of democratic policing.

Problems to address with citizen actions

That large differences exist between racial and ethnic groups.
*Blacks in particular and to a lesser extent Hispanics have materially lower levels of confidence in the police than do whites.
*Some research suggests that decisions on police tactics and strategies may worsen or improve citizen reaction, trust, and confidence.

Race and Policing

Police mist