Criminal Justice

Forcible Stops: Police and Citizens Speak Out 181

Public Administration Review,

Vol. 77, Iss. 2, pp. 181–192. © 2016 by

The American Society for Public Administration.

DOI: 10.1111/puar.12684.

Forcible Stops: Police and Citizens Speak Out

Eli B. Silverman is professor emeritus

at John Jay College of Criminal Justice

and the Graduate Center of City University

of New York. He previously served with

the U.S. Department of Justice and the

National Academy of Public Administration.

His recent publications include The Crime Numbers Game : Management by Manipulation (CRC Press, 2012) and articles in Journal of Police Studies and Justice Quarterly . E-mail: [email protected]

Christine S. Barrow is assistant

professor in the Department of Criminal

Justice at Molloy College. She earned a PhD

in criminal justice from Rutgers University

in 2012. Dr. Barrow has spent several

years performing research on urban youth

involvement in community organizations.

She is author of a book chapter titled “A

Second Chance: Delinquency Prevention

among Special Education Students” and

coauthor of a chapter titled “Unreasonable

Suspicion: Youth and Policing in NYC.”

E-mail: [email protected]

John A. Eterno is professor and

associate dean in the Department of

Criminal Justice and Legal Studies at Molloy

College, New York. He retired from the New

York City Police Department as captain. He

has won awards for his teaching, research,

and service. He is author or editor of six

books, numerous book chapters, and

peer-reviewed and other materials. He has

appeared as an expert for media such as

CBS, National Public Radio, and ABC.

E-mail: [email protected]

Abstract : A recent federal district court decision found that the New York City Police Department engaged in unconstitutional policing. Officers stopped millions of minority young men over the span of several years. Most of those stopped were innocent of any wrongdoing. This article uses qualitative and quantitative research methods to investigate these inequities, examining the views of both the police and those who were stopped. The authors find that pressure on officers emanating from headquarters was responsible for much of the illegal behavior by lower-level officers. Additionally, youth in minority neighborhoods felt mistreated by the police; this led to distrust in the police. The authors suggest solutions to ameliorate this painful episode in police practice. The article concludes with specific recommendations such as reducing management pressure on officers and police developing community partnerships.

Practitioner Points • The New York City Police Department needs to adjust its style of policing so that it does not place excessive

pressure on officers to simply write stop-and-frisk reports. • These policing tactics have resulted in weak police–community relationships, particularly with male minority

youth. • It is recommended that police legitimacy be a priority. • Officers need to be more proactive and visible throughout their shifts and not focused on making numbers. • Police agencies need to consider community partnerships.

John A. Eterno Christine S. Barrow

Molloy College

Eli B. Silverman City University of New York

The public administration literature stresses the importance of collaborative partnerships between police and communities (Choi and Choi 2012 ). Yet for decades there has been a strained relationship between police and minorities in the United States, a focal point of which can be seen in forcible stop-and-frisk tactics.

This article analyzes data from both police and citizens to comprehend this relationship in the context of the stop-and-frisk activities of the New York City Police Department (NYPD). The research suggests that crime will decline if police–community partnerships are active and vibrant. However, police performance management has developed a very aggressive, numbers-driven style of management and police activities that has tended to alienate citizens and sow distrust between police and communities (Eterno and Silverman 2012 ). This management style has been very popular since 1994, when New York City substantially reformed its police managerial and operational systems. The keystone of those reforms was CompStat (“compare statistics”), a police performance management system devised to accurately measure, record, target, and combat crime.

CompStat was originally created to provide increased authority and discretion to local police levels. Some suggest that it gradually morphed into a top-down, centralized, bureaucratic managerial control system based on “broken windows” (Eterno and Silverman 2012 ) that, despite various interpretations, has generally been used by the NYPD to mean zero tolerance and enforcement of even minor violations (Eterno 2001 ). Many have questioned its effectiveness (Fagan and Davies 2000 ; Greene 1999 ; Harcourt 2001 ). A recent U.S. Department of Justice (2016) report on Baltimore policing raises similar concerns.

The CompStat system is not only important in New York but also has been widely imitated in policing elsewhere and has served as a model for a more generalized offshoot called CitiStat (Gerrish 2016 ; Radin 2015 ). Since CompStat was first unveiled by the NYPD in 1994, a survey by the Police Foundation for the National Institute of Justice (Weisburd et al. 2001 ) revealed, one-third of the nation ’ s 515 largest police departments had implemented a CompStat-like program by 2001 and 20 percent were planning to do so. The same survey found that about 70 percent of police departments with CompStat programs reported

182 Public Administration Review • March | April 2017

attending an NYPD CompStat meeting. At this point, the CompStat system is ubiquitous throughout the democratic policing world, in part because New York City is a flagship police department (Bayley 1994 ). It is critical, therefore, to understand what is happening in New York City in order to grasp policing developments throughout the democratic world.

Policing in a democracy is no simple matter. Officers must follow the law while they enforce it. Packer ( 1966 ) posits two useful criminal justice models for exploring both of these mandates. In the first model (which assumes that most people arrested are guilty), termed “crime control,” the police focus on curbing crime. Packer likens this to an assembly line in which the suspect is arrested, brought to court, and sentenced as quickly as possible. The second model, the due process model, is likened to an obstacle course and has a built-in assumption that suspects are innocent. Obstacles are placed before the police, such as reading suspects their rights, securing warrants before searching, and the like. Neither model is right nor wrong; rather, they serve as useful ways to analyze the criminal justice system and are relevant to this study, in which we examine the policy in New York City known as stop and frisk.

The controversial practice of stop and frisk in New York City has led to scholarly inquiry that has demonstrated disparities in terms of NYPD stops of minorities compared with the population of those same minorities (Fagan and Davies 2000 ; Fallon 2013 ; Ferrandino 2013 ). This is often labeled “racial profiling,” which refers to the discriminatory practice by law enforcement of targeting individuals for suspicion of crime based on the individual ’ s race, ethnicity, religion, or national origin. Yet few explanations are offered as to the mechanisms by which those disparities occur. One study references evidence in the Floyd stop-and-frisk case, in which the NYPD was found to be conducting unconstitutional stops, but provides no primary data (Margeson 2014 ). Another presents primary data from marijuana arrests but only surmises when it comes to policy issues (Johnson et al. 2008 ).

A major research concern regarding urban stop-and-frisk policing is the impact on neighborhood youth. New York City youths are disproportionally targeted for police stops in comparison to other age groups. Stoudt, Fine, and Fox ( 2011 ) examine the NYPD Stop, Question, and Frisk data set for the years 2008 and 2009. They conclude that stop and frisk is a policy heavily focused on younger minority citizens in New York City. Fratello, Rengifo, and Trone ( 2013 ) found that the same was true for 2012. Court decisions allow an officer to stop and frisk an individual without probable cause to investigate a situation, but only if the police officer has a level of proof termed reasonable suspicion. The officer must reasonably believe that the person has committed, is committing, or is about to commit a crime. 1 If reasonable suspicion is not present, then stop and frisk is considered unconstitutional. Other research also documents a wide range of harms to youth, particularly minority residents, including low trust in adults of public authority, being subjected to unjustified stops and physically intrusive searches, slow police response time, police harassment, verbal abuse, excessive use of force, and low levels of police legitimacy (Brunson and Miller 2006 ; Brunson and Weitzer 2009 ; Fine et al. 2003 ; Hinds 2007 ; Jones-Brown 2000 ; Mastroski, Reisig, and McCluskey 2002 ).

Much of this behavior impacts public perceptions of police. According to a recent Gallup poll, Americans’ confidence in the police is at a 22-year low of 52 percent (Jones 2015 ). New York City mayor Bill de Blasio has said that he supports stop and frisk, but, unlike his predecessor, he has publicly stated his opposition to the abuse of the practice. The practice of stop and frisk has left an impression on residents in New York City and beyond, with many believing that the practice is motivated by racial or age-based bias and that it limits the rights of citizens in public spaces (Evans and Williams 2015 ).

The official stop-and-frisk policy was clearly intended to reduce crime and, more specifically, according to then–police commissioner Raymond Kelly (2002–13), to remove guns from the streets (Paddock 2013 ). We examine stop and frisk using both quantitative and qualitative methods. We conducted two surveys of police officers to quantify a wide range of supervisory pressures as well as the importance of other key factors such as the race of the officer. We also provide qualitative research revealing the attitudes of neighborhood residents subjected to stop-and-frisk practices. Based on these unique data, we outline practices and policies that might better meet the needs of both police and citizens. In a democratic society, police derive their power from the people, working in partnership with them—with their approval in most cases—to help society function and offer peace and stability. This is not a simple task, nor is it meant to be. Officers cannot do as they please, as in a police state, but must follow the law even as they enforce it.

Background In New York City, Mayor Michael Bloomberg took office in January 2002 and appointed as his police commissioner Raymond Kelly, who inherited a 60 percent decrease in reported serious crimes. Additionally, stops were fairly low, about 100,000 per year. While forcible stops were of some concern (see Daniels v . City of New York 2 ), most New Yorkers were not worried. Indeed, the Center for Constitutional Rights (CCR), which argued for the plaintiffs in Daniels , settled the lawsuit against the city with a requirement that the city monitor its stop-and-frisk activities and CCR serve as a watchdog.

As Commissioner Kelly ’ s tenure continued, the police seem to have either totally forgotten or neglected the settlement in Daniels . Forcible stops skyrocketed, and the NYPD made little to no effort to monitor itself—even to the point of not providing data to CCR for its review. It was not until CCR sued the city in a successor case to Daniels—Floyd v. City of New York 3 — that the data eventually emerged as a result of a federal district court order. When the data were eventually released, the New York Civil Liberties Union (NYCLU) and other experts were able to analyze what was occurring. The number of forcible stops had skyrocketed to nearly 700,000 per year at the peak in 2011, with the vast majority of stops being conducted on young minority males. Based on belatedly released stop-and-frisk reports (also called UF250s), most of those stopped were completely innocent and never were arrested or received a summons. Additionally, taking guns off the streets, the city ’ s self-proclaimed main reason for using forcible stops, happened only rarely, in 0.1 percent of stops (NYCLU 2016 ).

Crime continued to decline, but not as precipitously as in previous years. By the time Commissioner Kelly left office on January 1,

Forcible Stops: Police and Citizens Speak Out 183

2014, reported serious crime was down 80 percent since its peak, an additional 20 percent from when he took office 12 years earlier. At least some of this decrease, however, may be attributable to the manipulation of crime reports (Eterno, Verma, and Silverman 2016 ). Additionally, if we assume that serious crime was down 80 percent, then the city should have been much safer. If this is true, a logic problem arises. With crime down so much, fewer suspects should be on the streets, but this inference is at odds with data indicating a substantial increase in forcible stops of suspects.

In our democratic society, under the provisions of the U.S. and New York State constitutions and the New York State Criminal Procedure Law, police must have some reasonable suspicion before they may forcibly stop someone. While this level of proof is not as high as probable cause—the level of proof needed for arrest—it does require more than a hunch or gut reaction on the part of the officer. Thus, if officers are required to have reasonable suspicion and crime supposedly was down 80 percent, there should have been far fewer criminal suspects on the streets. Yet with nearly 700,000 forcible stops (of ostensible criminal suspects), it appears that such suspects were ubiquitous.

Using Packer ’ s crime control/due process dichotomy, the NYPD clearly adopted a strong crime control stance, engaging in nearly anything to reduce crime. However, the police were playing close to the edge of legality, and, according to the Floyd verdict, crossed the line and violated the law. With millions of mostly minority young men being forcibly stopped in the span of a few years, it became clear to many that the police were going too far. Grassroots movements including Communities United for Police Reform and the Police Reform Organizing Project focused on this issue. As would be expected when the social contract is violated, a new mayor, Bill de Blasio, was elected in 2013 on a platform of changing police behavior.

The new mayor and his chosen police commissioner, William Bratton, are working under different rules. Mayor de Blasio did not challenge the federal district court ’ s ruling in Floyd, so the NYPD now has a federal monitor reviewing its forcible stop practices. Stops have plummeted to new lows, with fewer than 50,000 forcible stops in 2014 (NYCLU 2016 ). The city has also adopted a new inspector general appointed by the mayor who has oversight of the NYPD, in yet another attempt to review the police department ’ s activities and policies. This development is over and above the Floyd case ’ s court- appointed federal monitor, the existing Civilian Complaint Review Board, federal agencies, district attorneys, the state attorney general, and other state and city offices with oversight authority in various areas (e.g., city comptroller, public advocate, and state comptroller). Police leaders must also contend with unions, a sometimes aggressive and critical press, and popular antipolice movements.

Given this background, we examine two key questions. First, does the CompStat performance management system, with its focus on crime control and measureable police activity (e.g., stop-and-frisk reports, summonses, arrests), translate into pressure on officers of all ranks? We hypothesize that there is a direct relationship between CompStat and the pressure officers feel to conduct forcible stops. In addition to pressure, we examine other explanatory variables such as gender, race, and the education level of officers. Second, using our

qualitative data, we ask how pressure on officers influences, if at all, police–community interactions.

Our study uses a unique methodology with primary data from both the police and communities. We hypothesize that pressure, if it exists, has a deleterious influence on such relations. We analyze three unique data sets. Two of the data sets are surveys of police officers, and the third is from the field to allow citizen feedback. The data allow us to fill a research gap by offering primary evidence regarding the etiology of the huge increases in stop and frisk as well as how those increases were perceived on the street. Finally, we offer specific suggestions to address these key issues.

Methods Mixed-Methods Approach We note at the outset that the NYPD is not a transparent organization. Numerous freedom of information requests were routinely ignored by the NYPD (see New York City Public Advocate 2013 ). Our design had to overcome this by adopting an innovative method. Additionally, given the high level of police contact with youth in certain New York City neighborhoods, it is important to attempt to understand the dynamics of these contacts and their impact on youth. To understand the complexities of these situations, a mixed-methods approach is most appropriate. Quantitative surveys provide information on the factors that shape police behavior, while the qualitative data of citizen accounts provide a detailed understanding of the impact of police stops. The two approaches complement one another.

Quantitative Police Surveys The key aspects of our quantitative methodology are two self- reported, anonymous surveys conducted with retired New York City police officers as respondents. Importantly, this method offers two modes of protection to respondents. First, because the officers are retired, they have little to fear from retaliation. Second, and just as important, we designed the survey so that we could not determine the names of those who responded. Anonymity is a technique common to social science to increase validity (truthfulness) when asking sensitive questions. In this survey, we ask officers to reveal sensitive information about forcible stops, some of which may have been illegal. Anonymity and the use of retirees can provide officers with a sense of security and thereby facilitate truthful responses (Babbie 1989 ; Bradburn 1983 ; Dillman 1983 ; Neuman 2000 ). Both surveys were carefully designed using focus groups of officers and other researchers to help with question wording and other aspects of the survey before the full administration.

Given that the sampling frames in both surveys are retired police officers, memory can be an issue. We address this using techniques suggested in the scientific literature: a very limited number of questions (Bradburn, Rips, and Shevell 1987 ); a self-administration method that allows maximum time to consider answers (Bradburn, Rips, and Shevell 1987 ); less complex matrix-type questions for key variables (Bourbonnais, Meyer, and Theriault 1988 ); using only those employed for many years (Stewart, Tonascia, and Matanoski 1987 inform us that five or more years of employment leads to better recall—only one respondent had less than five years of experience); and asking questions that are pertinent to the study and to respondents (Bourbonnais, Meyer, and Theriault 1988 ).

184 Public Administration Review • March | April 2017

For the purposes of this article, both quantitative surveys are reviewed. However, our more recent 2012 survey, with nearly 2,000 officers of all ranks, is analyzed more extensively using multivariate methods. That is, both surveys show enormous pressure on respondents to conduct stop and frisk. Nevertheless, the second survey is more useful at explaining this pressure throughout the ranks. We explain both surveys’ methods.

The first survey used a self-report mail design and was sent to respondents in 2008. For this survey, we contacted the Captains Endowment Association and asked whether we could send our survey to all its retirees. These respondents were middle- and higher- ranking officers (captains and higher). The instrument was mailed to all retirees in September 2008; by the following February, we had 491 returned questionnaires, a response rate of 41 percent.

There were 263 (54.1 percent) captains, 98 (20.2 percent) deputy inspectors, 63 (13 percent) inspectors, 28 (5.8 percent) deputy chiefs, and 34 (7 percent) others. These percentages also approximate the ranks of those retired. These data increase our confidence that there is little, if any, nonresponse bias.

A similar method was used for the second survey. This allowed a large sample to be collected of 1,962 retired officers of all ranks. An e-mail survey was conducted in March 2012. The sampling frame was registered officers in the NYPD active retirees database. Only retired officers can register on the site, which is maintained by the NYPD. The purpose of the site is to offer information to retirees and tap them in emergency. It is completely voluntary for retired members to register. The site allows registered members to contact other registered retirees and the site supplies e-mail addresses and other contact information. We had access to a retiree, which allowed us to download the contact information. (Additional database information can be found at the website http://www.nypd2.org/ retirement/home.html .)

Using this database, we sent a request to participate in the survey—which included a unique link to the questionnaire— by e-mail to the 4,069 NYPD retirees with operational e-mail addresses. The survey was sent on March 26, 2012. If the unique link did not indicate a response, a follow-up was sent on April 3. A final follow-up was sent only to links that indicated no responses to the first or second invitations on April 16. In total, there were 1,962 responses, a return rate of 48.2 percent. The characteristics of respondents closely mirrored known NYPD trends. For example, the percentage of women steadily increases in our sample (1981–93, 2.1 percent female; 1994–2001, 10.7 percent female; 2002–12, 11.0 percent). There is also a pyramidal structure in the data mimicking the NYPD ranks (i.e., more police officers compared with sergeants and so forth). The large size of the sample, the corresponding characteristics of our samples to NYPD, and, as we shall see, the close correspondence of the results of the two separate surveys give us confidence that our results reasonably reflect the larger population of NYPD officers. In addition, while we focus on the quantitative data, the study ’ s interpretations are supported by anecdotal evidence, other widely available information (cited herein), and personal in-depth interviews with more than 30 commanders and other officers, both active and retired.

Qualitative Community Research This section examines the experiences of young men and women immediately impacted by stop-and-frisk practices. It also assesses how existing police practices shape youth ’ s perceptions of police in their communities. Data for this section were derived from a broader study of youth involvement in neighborhood organizations. Qualitative interviews were conducted with young men and women at a Brownsville (Brooklyn) recreation center. It should be noted that Brownsville is among the 10 highest stop precincts for the years 2003–12 (Jones-Brown et al. 2013 ).

Prior to setting up interviews with youth participants, a letter was sent to the head of the New York City Department of Parks and Recreation describing the purposes of the study. Permission was granted to conduct the investigation in June 2009. Data collection took place in two stages: a snowball sample of 15 respondents was selected during July 2009, and a random sample of 20 respondents was selected during December 2010 for this analysis. As a result, 35 participants were selected for this study.

Prior to each interview, the researcher explained the full purpose of the study and administered full-consent procedures with an age- appropriate consent form for their signature. The respondents were assured confidentiality and informed about the potential risks of the study. The questions asked were topical in nature, which allowed the researcher to ask specific questions that did not distract respondents from their activities for extended periods of time. The focus was on respondents’ perceptions and experiences with the police in their communities. Respondents were asked questions including, “Do you have respect for the police?,” followed by a probe, “Can you talk a little bit about any experiences or encounters you have had?” and “Do you feel the police keep your neighborhood safe?”

Each interview lasted between 45 and 60 minutes. Interviews were transcribed and reviewed immediately following each visit to the center, to guard against inaccurate interpretations of the data. The data were reviewed for clarity and continuity in the stories provided by the participants. Data analysis for this investigation was achieved using grounded theory methods ,where themes were systematically generated from the data (Glaser and Strauss 1967 ). The in-depth interviews from young men and women allowed for a substantive analysis of their interaction with police officers in their community.

Findings Quantitative Surveys of Police Tabular analyses . The two offi cer-administered surveys show remarkably similar results. Such replication is powerful evidence suggesting that the survey results are an accurate depiction of NYPD events. Both surveys show strong management pressure on respondents to conduct stop and frisk during the Kelly/Bloomberg era (2002–12). Respondents similarly noted high pressure to write summonses and make arrests in more recent years (not shown).

Table 1 shows the pressure on captains and higher-ranking officers during the pre-CompStat versus CompStat era. The variable pressure was measured on a typical 1–10 scale, with 1 being the least pressure and 10 being the most pressure. With respect to the variable on stop and frisk, only 5.1 percent of commanders before CompStat felt high pressure to get stop-and-frisk reports. Those who retired

Forcible Stops: Police and Citizens Speak Out 185

after CompStat, however, felt markedly higher pressure. Of the 452 commanders surveyed, 315 worked during CompStat. Of those 315, 28.3 percent felt high pressure to increase their stop-and-frisk reports. This is a 23.2 percent difference. Our measure of association shows this to be a strong and statistically significant relationship (gamma = .617, p < .0001). Two other tables from this survey (not shown) indicate similar results, namely, significantly more pressure on commanders during the CompStat era. These tables show higher pressure to write summonses (25.5 percent felt high pressure before CompStat, while 45 percent felt high pressure during CompStat; gamma = .431, p < .0001) and higher pressure to make arrests (16.5 percent felt high pressure before CompStat, while 42.5 percent felt high pressure during CompStat; gamma = .571, p < .0001).

Table 2 shows the results of the second survey administered in 2012 with a different sample of officers of all ranks. The results from this survey are remarkably similar to the first, showing a relationship that is statistically significant, strong and increased pressure during CompStat. However, in this survey, we are able to parse out different eras—pre-CompStat (before 1995), early CompStat (1995, the first full year of CompStat), the Giuliani era (1995–2001), and the Bloomberg/Kelly era (2002–12). We found that the pressure was clearly highest during the Bloomberg/Kelly era.

In table 2 , we see that high pressure from management/supervisors to increase stop-and-frisk reports increased from a low point of

9.1 percent before CompStat, to a midpoint of 19.1 percent during the Giuliani era, to a high point of 35.1 percent during the Bloomberg/Kelly era. This means that four times as many officers felt high pressure during the Bloomberg administration compared with officers who worked before CompStat, and about twice as many officers felt high pressure to increase stop-and-frisk reports compared with those during the Giuliani era. This result is fairly strong and statistically significant (gamma = .461, p < .0001) and comports well with the 2008 study. Additionally, pressure from management/supervisors to write summonses and make arrests was also significantly higher. Pre-CompStat high pressure to make arrests was 18.4 percent; during the Giuliani era, 35.3 percent; and during the Bloomberg/Kelly era, 49.3 percent (gamma = .422, p < .0001). Similarly, for summonses, high pressure, in ascending order, was 37.3 percent, 49.2 percent, and 61.6 percent, respectively (gamma = .322, p < . 0001). This, too, mimics the results of the previous survey.

In sum, the tabular analyses indicate that more than 2,000 officers in two separate surveys felt extensive pressure to conduct stop and frisk—a finding that is difficult to ignore. Officers were clearly feeling significant pressure from higher-ups to produce. Given the years of retirement (we asked those who retired after 1994 to respond to their experiences only after that year), the likely source of this pressure is CompStat. Because CompStat stands for “compare statistics” in the context of a performance management meeting, commanders

Table 1 2008 Survey Findings

Crosstab: Year * Pressure to increase stop and frisk

Did You Serve on NYPD after 1994? Total

No Yes

Increase stop and frisk categorized

Low Count 75 71 146

Percentage 54.7% 22.5% 32.3%

Medium Count 55 155 210

Percentage 40.1% 49.2% 46.5%

High Count 7 89 96

Percentage 5.1% 28.3% 21.2%

Total Count 137 315 452

Percentage 100.0% 100.0% 100.0%

Gamma = .617, p < .0001 Kendall ’ s Tau-b = . 347, p < .0001

Table 2 2012 Survey Findings

Crosstab: Year categorized * Stop pressure categorized

Year Categorized Total

Before 1995 1995–2001 2002–12

Stop pressure categorized

Low Count 297 127 205 629

Percentage 57.8% 36.7% 24.4% 37.0%

Medium Count 170 153 341 664

Percentage 33.1% 44.2% 40.5% 39.0%

High Count 47 66 295 408

Percentage 9.1% 19.1% 35.1% 24.0%

Total Count 514 346 841 1701

Percentage 100.0% 100.0% 100.0% 100.0%

Gamma = .461, p < .0001 Kendall ’ s Tau-b = .302, p < .0001

186 Public Administration Review • March | April 2017

at these meetings are pressed to defend what is occurring in their precincts. Over the years, they have been held strictly accountable for the numbers in the command, with crime numbers and anticrime activities as the focus of attention. From our in-depth discussions with numerous commanders (more than 30), stop-and-frisk reports morphed into something that could cover the commanders when questioned about a crime. For example, if a robbery occurred, commanders would try to show that forcible stops were done in the area at the time of the incident to attempt to capture the perpetrator. We further test our data to determine whether these results hold in other models.

Correlations . We conducted more in-depth analyses with the 2012 data because they contain information on all ranks and other relevant details not included in the earlier data. The coding of the variables is explained in table 3 . This is important to interpretation of the results. Table 4 contains the results of bivariate correlations

among key variables. The year that offi cers retired and the pressure to conduct stops have a statistically signifi cant, moderate, and positive association. This means that in more recent years, offi cers

are feeling more pressure to conduct stops. Given the previous crosstabs, this is expected.

The race of officers also shows a statistically significant relationship to pressure to conduct stops. However, this relationship is weak. Nonwhite officers are more likely to feel pressure compared with white officers. Because most stops are conducted in minority areas in New York City, it makes sense that nonwhite officers would feel more pressure. Nonwhite officers may feel more ill at ease

with the stop policy. Rank also shows a significant, weak, but positive association. This means that higher ranks are feeling more pressure. There is also a statistically significant and weak relationship with education. Educated officers feel more pressure; they may be aware that what they are doing is legally and ethically questionable (as the judge found in holding the NYPD liable in the Floyd case).

Pressure to obey the law is not significant. That is, officers perceived very little pressure to obey the law and hence the pressure did not significantly influence officer behavior. This is interesting because this pressure should cause some conflict in officers. It appears that the department is not putting much pressure on officers to obey legal guidelines. In a separate analysis (not shown), it appears that less pressure is placed on officers to obey the law, especially in more recent years. Gender is not significant, with men and women seeming to act similarly. Lastly, whether officers had a patrol assignment since CompStat began in 1994 has have a significant, positive, and moderate relationship. This makes sense given that patrol officers (those who work the streets) are most likely to feel the brunt of the policy to make forcible stops.

Table 3 Variable Coding

Variable Coding

Pressure to conduct stops 1–10 (1 = least, 10 = most)

Pressure to obey law/constitution 1–10 (1 = least, 10 = most)

Year retired 1 = earliest, 48 = most recent

White or not 0 = not white, 1 = white

Gender 0 = female, 1 = male

Education 1 = high school, 2 = some college, 3 = bachelor ’ s degree, 4 = some graduate school, 5 = graduate/law school

Patrol assignment after 1994 0 = no, 1 = yes

Rank 1 = police offi ce/detective, 2 = sergeant, 3 = lieutenant, 4 = captain or above

Table 4 Pearson Correlation Matrix

Increase Stop and Frisk

What Year Did You Retire?

White or Not

Rank Education Obey Legal/

Constitutional Rules

Patrol after 1994?

Gender

Increase stop and frisk Correlation 1

N 1830

What year did you retire? Correlation .357 ** 1

N 1701 1815

White or not Correlation –.095 ** –.162 ** 1

N 1762 1760 1870

Rank numerical Correlation .110 ** .094 ** .098 ** 1

N 1791 1801 1841 1907

Education level Correlation .062 ** –.009 .038 .414 ** 1

N 1811 1802 1867 1891 1928

Obey legal/ constitutional rules Correlation .011 –.058 * –.031 .010 .002 1

N 1797 1740 1792 1827 1845 1866

Patrol after 1994? Correlation .329 ** .558 ** –.081 ** .188 ** .013 –.040 1

N 1759 1799 1803 1846 1859 1797 1877

Gender Correlation –.025 –.134 ** .154 ** .063 ** –.032 –.065 ** –.061 ** 1

N 1809 1802 1865 1891 1917 1845 1856 1926

** p > .01 (two-tailed).

From our in-depth discussions with numerous commanders

(more than 30), stop-and-frisk reports morphed into some- thing that could cover the

commanders when questioned about a crime.

Forcible Stops: Police and Citizens Speak Out 187

Multivariate regression . Given that many of these variables have a statistically signifi cant infl uence on our dependent variable, we conducted a multivariate regression analysis. Regressing the pressure to increase stop and frisk onto this set of independent variables allows us to explain 16.0 percent of the variation in the pressure to increase stop and frisk (see table 5 ). Regression assumptions were tested and found to be well within acceptable parameters. We tested other models, and the one presented best represents the data. For example, we removed age of the offi cer in the presented model because of its strong relationship to year of retirement. This is because multicollinearity was a problem (a very high correlation and variance infl ation factor were seen). In our presented model, explanatory power is low, but the F -test of overall signifi cance indicates that our model predicts better than an intercept-only one. Many of the variables are statistically signifi cant. We surmise that because we do not have a variable that directly measures the impact of CompStat, this weakens our explanatory power. Nevertheless, slope coeffi cients (those that are statistically signifi cant and those that are not) show critically important fi ndings that fi t neatly with the scientifi c literature and our research.

The pressure from management to obey legal guidelines does not show a statistically significant relationship with the dependent variable—pressure to conduct stops—while controlling for the other variables in the model. This variable does not cross the .05 threshold. Additionally, the slope coefficient is small ( β = .039), showing that for every unit increase in pressure that officers feel to obey the law, we see a slight .039 increase in pressure to conduct stops while holding the other variables in the model constant. With the legal pressure variable not significant and the slope coefficient small, we note that officers were not heavily influenced by pressure to obey the law. This variable is not significant in the bivariate correlations as well. It could be that the NYPD was so focused on crime control that due process considerations (obeying the law) became less critical (see, e.g., Packer 1966 ). This variable may simply reflect the department ’ s ambivalence toward legal restrictions and its extreme focus on crime control. This is seen in the literature especially on NYPD ’ s “broken

windows” policy (Eterno 2001 ; Eterno and Silverman 2012 ; Fagan and Davies 2000 ; Silverman 2001 ).

Other variables in the model also show interesting results. While controlling for the other independent variables in the model, the slope coefficient for officers’ retirement year is positive and statistically significant ( β = . 087, p < .0001). For every year increase in retirement, the pressure to conduct stops increases .087, controlling for other variables in the model. We also note that the standardized slope coefficient indicates that retirement year has the largest influence on the pressure to commit stops compared with the other variables in the model ( β = .249). This fits nearly with our tabular analysis and means that officers in more recent years are feeling more pressure to conduct stops. Even while controlling for the other independent variables, supervisory pressure to conduct stops increases as the retirement year increases. In essence, what we saw in the bivariate (tables and correlations) remains in the multivariate model. Examining both the bivariate and multivariate results indicates that the most likely explanation for these results is the impact of the CompStat performance management system as used by the NYPD (Eterno and Silverman 2012 ). That is, the most recent years show the most pressure to conduct stops.

Another statistically significant variable—patrol assignment after 1994—supports the interpretation that CompStat is critical to understanding the pressure to conduct stops. Since CompStat began in 1994, it stands to reason that those who were working in

precincts on the streets (that is what “patrol” means in the NYPD) during the CompStat era felt more pressure to conduct stops. Patrol after 1994 shows a positive and significant impact on pressure to conduct forcible stops ( β = 1.053 , p < .0001) while controlling for the other variables in the model. Additionally, the standardized slope coefficient indicates that this variable also has a large influence on the model compared with the other variables

( β = . 177). This supports our hypothesis on CompStat as well.

How officers identify themselves racially—as white or not white—is also statistically significant ( p > .05). The negative slope coefficient ( β = − .423) indicates that nonwhite officers felt more pressure

Table 5 Ordinary Least Squares Regression

Unstandardized Coefficients Standardized Coefficients Sig.

β SE β t

(Constant) .527 .503 1.048 .295

Obey legal/ constitutional rules .039 .023 .038 1.654 .098

White or not –.423 .176 –.057 −2.406 .016

Gender .422 .252 .039 1.675 .094

What year did you retire? .087 .010 .249 8.813 .000

What is the highest level of formal education you have completed? .167 .069 .062 2.429 .015

Patrol after 1994? 1.053 .168 .177 6.259 .000

Rank numerical .131 .084 .041 1.571 .116

Dependent variable: Increase stop and frisk R 2 = 16.0 F = 45.539, p < . 000

Since CompStat began in 1994, it stands to reason that those

who were working in precincts on the streets during the Comp-

Stat era felt more pressure to conduct stops.

188 Public Administration Review • March | April 2017

to conduct stops even after controlling for the other variables. This makes sense, as minority officers may be more likely to feel pressure when they have to forcibly stop members of their own race/ethnicity. Some literature implies that minority officers have different experiences in their departments compared with white officers (Dowler 2005 ; Strokes and Scott 1996 ).

Regarding gender, the slope coefficient is positive, indicating that women feel more pressure to conduct stops. Importantly, however, this difference is not statistically significant, suggesting that men and women do not differ in their perceptions of management stress to conduct stops ( β = .422, p = .094). In a study comparing male and female officers with respect to acting within the law, differences in behaviors are explained by the situation (Eterno 2006 ). Management pressure, then, is not a likely source of variant legal behaviors previously observed between men and women.

Education level has a statistically significant influence on pressure to conduct forcible stops ( β = .167, p < .05). As education level increases, so does the pressure that officers feel to conduct forcible stops, controlling for the other variables in the model. It may be that educated officers have an understanding that what they are doing as wrong or even illegal (see Floyd v. City of New York ). Further, officers who are educated are generally taught to think critically, to go beyond what they are being told. Such thinking could account for the increased pressure. We note that some research indicates that college- educated officers feel more stress (Zhao, He, and Lovrich 2002 ).

The last variable in the model is rank. The slope coefficient for rank indicates a positive association controlling for the other variables in the model ( β = .131, p = . 116). However, rank does not have a statistically significant influence on management pressure to conduct stops. This result, then, may have occurred by chance. One might think, however, that as officers advance up the ranks, they would feel less pressure from management. Yet in the CompStat environment, pressure is expected on all ranks—perhaps even more on higher-ups, as they attend CompStat meetings regularly. In this study, there is no evidence to suggest higher-ups are receiving less pressure, which is in sync with our hypothesis. The slope coefficient indicates more pressure on higher ranks. Although this difference is not statistically significant, there is no evidence of the opposite (higher-ups feeling less pressure).

In sum, our quantitative findings indicate high pressure on officers to conduct forcible stops with no significant pressure to obey the law. The data suggest, furthermore, that pressure to conduct stops emanated from the performance management system (CompStat), in which commanders were held strictly accountable for what occurred in their commands. Officers at lower levels were then forced to deal with this pressure. Anecdotal evidence from the unions and whistleblowers supports this interpretation (Eterno and Silverman 2012 ). How did all this translate into what citizens were feeling in the streets? These findings constitute the next part of our study.

Qualitative Field Research in Communities As previously discussed, a total of 35 qualitative interviews, 24 male and 11 female, were conducted with the recreation center

participants. Participants ranged in age from 14 to 21, with a mean age of 18. Respondents’ ethnic backgrounds varied; many identified themselves as being of Caribbean origin from countries including Guyana, Haiti, Jamaica, Trinidad, and St. Lucia. Others identified themselves as African American, Latino, Cherokee Indian, Polish, Nigerian, and Kenyan.

As one of the research goals was to examine how respondents perceive different forms of control in their lives, youth perceptions of the police were formed through direct and indirect contact with officers. Three significant themes emerged from the respondents’ experiences with the police: negative police interaction, avoiding police contact, and police apathy.

Negative police interactions . Youth views of street stops in their community were often shaped by what they had experienced and witnessed in the neighborhood. The overall consensus was a feeling of being mistreated and belittled when interacting with the police. When asked, “Do you have respect for the police?,” Eric, an 18-year- old male, stated, “It depends, on the way they approach me. Like if they come to me in a subtle type of manner then I ’ ll present myself the right way. If they approach me in a negative type of fashion, I guess I ’ ll have to be negative about it.” (Note: Respondents’ answers are presented in their original form without corrections.)

When asked the same question, James, a 17-year-old male answered,

No because I don ’ t like the procedures in which they go about doing things and approaching people. I don ’ t like how they approaching people telling them to shut up and don ’ t let us speak or nothing . They try too hard and they abuse their power. Like they can see a group of people walking the street, you won ’ t be doing anything but they ’ ll stop you and search everybody for no apparent reason and they ’ ll make sarcastic jokes and remarks about it like it ’ s funny to them.

Eric ’ s response suggests that, depending on how the police speak to him, he may respond in a hostile manner, which may put him at risk for a confrontation with the police and a possible arrest. According to James, part of the frustration comes from not knowing why the police are making the initial stops. That frustration is compounded when residents felt subjected to belittling remarks or

that their voices were being suppressed.

Other respondents reported being asked for identification in addition to being searched. For example, William stated, “Plenty of times they harass me. They see you with anybody and they think you up to something, they ask for ID, I give them and then they leave me

alone.” Gregory shared,

Every three months I get stopped. They search me. I ’ ll be close to my house, coming from the store. I had my earphones in my ear and driving slow next to me, and all of a sudden he told me to stop and just searched me. I don ’ t know if they were waiting for me to look at them. I didn ’ t pay them no attention. When I looked they asked if I had anything. And then searched me.

Our quantitative fi ndings indicate high pressure on offi cers to con-

duct forcible stops with no signifi - cant pressure to obey the law.

Forcible Stops: Police and Citizens Speak Out 189

In both of these instances, the police did not explain to the youth why the young men were being stopped or why they needed to provide identification.

When stopped without being informed of their wrongdoing, youth in this sample also tried to make sense of police stops and the rationale behind the procedures. Consequently, another frustration expressed was the behavior or appearance that warranted the stop. Joe, a 19-year-old male, described in detail an encounter with the police in his neighborhood:

It was hot, I had my shirt off, I was walking, this was in the nighttime, I was walking to the store. I was walking across the street and seen the van, I took a step back and I looked and then they looked at me and I kept walking and they seen that. So I kept walking. I was to the store and saw that it was closed so I turned around and walked back and I seen the van on the corner. Then they pulled up beside us and jumped out and put me in handcuffs and ask me my name. I said why you want to know my name I don ’ t know you? What did I do? He asked me for ID and I said I ain ’ t have none. He said they looking for someone and I fit the description. I said you don ’ t have the right to ask me my name and put me in handcuffs.

I told him my name is Mike [not his real name]. He said Mike what? I said “Mike.” He kept looking at the paper and then looking at me. He said we got a call that someone did something and it was on this exact block. He asked what building I live in. I told him on this block. Some girl called my uncle she told my uncle that they got his nephew in handcuffs. And then I see my uncle and my grandmother coming down the stairs. I ’ m walking in handcuffs. My grandmother was cursing them out and my grandmother identified who I was so they let me go.

In this instance, residing on the street where a criminal incident had taken place put this young man at risk for police contact. He was stopped and handcuffed because he was told he fit the description of the suspect. Vague descriptions of criminal suspects put law-abiding individuals at risk for illegal police stops. For example, Derrick, an 18-year-old male, expressed his annoyance with these actions. He shared, “Sometimes cops may go overboard with the description. ‘We have an African American male, 5–6, black hair.’ That ’ s everybody. That ’ s not a good description.”

Avoiding police contacts . Other respondents made note of which characteristics increased the likelihood of being stopped. Ian, an 18-year-old male, shared, “Sometimes when you wear a specifi c color in a neighborhood. Like I live in a blood neighborhood…and if I wear red, they would stop and search you because of the color.”

Stacey, an 18-year-old female, discussed the tactics she used to avoid contact with the police. She shared, “I don ’ t hang out with a lot of people that attract the cops. Like the cops start with anybody so I guess the guys. They start with people that have their pants down low, they see them with a blue or red flag, I don ’ t hang around those people.”

Because of such persistent tactics, young men and women developed ways to avoid police contact such as refraining from wearing certain

clothing or associating with individuals who do so. However, being male, being a minority, and living in an area where a recent criminal incident had occurred were unavoidable characteristics for many in this sample. Vague descriptions of alleged perpetrators increase the risk for unwarranted police stops.

Police apathy . In addition to respondents’ feelings toward police stops, there are other consequences of police practices. Offi cers are rarely found when true criminality takes place. This was a recurring theme among the respondents. When asked, “Do you think the police do a good job at keeping your neighborhood safe?,” a number of young men and women felt police tactics did not lower crime and victimization. Kevin, a 19-year-old male, explained his views on policing particularly in public housing:

Its these projects two blocks from my house, Pitkin projects, you hear shooting now and then, there is this boy his name is Henry, he got shot right there on Pitkin, that ’ s where the police came and the next day they are gone. You hear shooting now and then and they are not there. Usually they have a police stand, they don ’ t have it there, they put it like two blocks down like across from the junior high school. They don ’ t put it right there where the problem is. Its like they are afraid of the kids doing the shooting.

Kirk, a 17-year-old male, expressed similar views about police and neighborhood safety. He shared,

They [the police] are too laid back for me. There ’ s not enough in the activities going on around here. There ’ s a lot of trouble that comes around here. The glass bottles being thrown at people at cars, if we are walking to the store we see people throwing glass bottles or just retarded things. If there are cops around of course they may just try and stop it, there ’ s never that many cops around here to prevent any of it.

Respondents in the sample also expressed experiences with personal victimization of themselves and their loved ones. Again when asked, “Do you feel the police do a good job at keeping your neighborhood safe?,” Kevin shared,

No not really because one of my older friends got stabbed in the chest and they haven ’ t found the guy. One of my friends got shot around the corner from me and they still ain ’ t find the guy even though everybody still yelling his name out there who did it, they still ain ’ t looking for the guy. Now he ’ s probably somewhere out of NY state.

Joe shared his experience of being victimized in his neighborhood. He explained, “I got jumped a couple of weeks ago. I ran up to the cop car and told them what happened and they road off like it was nothing. They road around the block with their sirens but that doesn ’ t help.”

Lastly, Derrick shared a similar instance about police presence after a neighborhood shooting occurred. He stated: “Well an incident happened a couple of weeks ago over by 275 school a shooting. Its 2 blocks down its right next to coop. It must have been serious because after that incident cops were lined up over here, down there,

190 Public Administration Review • March | April 2017

by coop, so you know…but that should be a regular. A cop on every corner period. Don ’ t wait for something to happen when you can be there already and prevent something from happening.”

These qualitative findings suggest that existing police tactics do little to prevent violent crime from taking place, including assaults and shootings. Individuals who are not engaging in criminal activity are subject to police contact. Research on policing in urban communities has documented a wide range of harms, particularly for minority residents, including slow police response time, police harassment, disrespectful treatment, verbal abuse, and excessive use of force (Brunson 2007 ; Brunson and Miller 2006 ; Jones-Brown 2000 ; Mastroski, Reisig, and McCluskey 2002 ).

Youth experiences with police contact appear to be consistent with the research on race and policing in urban communities. In a study conducted by Brunson and Weitzer ( 2009 ), black respondents expressed hopelessness regarding their situations with police because they felt that officers would never see them as anything other than symbolic assailants, even when they were engaged in entirely lawful activity. Several of the respondents in this qualitative study shared these feelings; hence the need to employ tactics to avoid police attention.

Gau and Brunson ( 2010 ) also concluded that young men in a socioeconomically disadvantaged urban area harbored ambivalent feelings about the police. Police judged them based on their clothing, accessories, friends, and/or the neighborhoods in which they resided. They felt that police would use the inferences they drew from these surface characteristics as justification for stopping, questioning, and/or frisking them even when they were not engaging in crime. The young men and women in Brownsville reported efforts to avoid certain individuals, certain clothing, and parts of the community.

Finally, another negative impact of unwarranted police stops is lower police legitimacy. Young men and women see police as more legitimate when they were viewed as acting fairly and lawfully. Lowered legitimacy can influence both law-abidingness and the willingness to cooperate with legal authorities (Penner et al. 2014 ; Tyler, Fagan, and Geller 2014 ). Respondents in Brownsville were not informed of any wrongdoing when they experienced police contact. Such experiences can shape their perceptions of law enforcement as a legitimate institution.

Conclusion These findings, both qualitative and quantitative, are consistent with the judge ’ s decision in the Floyd case, which held that the NYPD ’ s practice of stop and frisk constituted “indirect racial profiling.” These data put us in a position to understand the consequences of behaviors by police officers in New York City and, by extension, any police agency using these aggressive tactics. Indeed, numerous departments throughout the world have emulated the NYPD model (Eterno 2001 , 2003 ; Eterno and Silverman 2012 ; Silverman 2001 ). Based on our findings, we offer the following conclusions and recommendations.

First, the variables representing patrol after 1994 as well as retirement year clearly show that CompStat is influencing officers of all ranks. CompStat, as practiced by the NYPD and departments elsewhere, must be adjusted. It places excessive pressure on officers to conduct stop and frisk without reason—or reasonable suspicion. Commanders are under great pressure at CompStat meetings to ensure that they demonstrate their activities to the upper echelons. For example, if a robbery occurs, officers are expected to do some forcible stops near the robbery, regardless of the level of suspicion. When a commander attends a CompStat meeting, he or she will refer to the stop-and-frisk reports as evidence that officers were monitoring the situation with the robbery. In fact, however, officers were often just stopping anyone in the area regardless of their legal right to make such stops. Our analyses confirm the powerful combination of pressures that officers felt from supervisors. There was increased pressure to conduct forcible stops and, at the same time, no measurable pressure to obey legal guidelines. These combined results, we argue, led to a substantial increase in illegal stops. Such unconstitutional stops had adverse effects on the community.

Based on anecdotal and whistleblower reports, other data analyses of the two surveys that show comparable pressure to write summonses and make arrests (not reported in this article) and direct interviews with officers and commanders, we further suggest that this pressure on commanders often translates into quotas for the rank and file. One commander, for example, told us, “You are out of your mind if you want to be a captain. NYPD should not deny quotas. Cops losing discretion, that ’ s what Hitler did, a very dangerous thing.” Another stated, “Eventually CompStat became just generate numbers. We even got pressure from City Hall… Upper echelon just wanted more and more.” This quota pressure becomes particularly worrisome in Impact Zones, where rookies were assigned to crime-ridden areas. Rookies are least able to deal with the conflicting pressures of quotas from superiors to make stops, their legal limitations (following the law while enforcing the law), community expectations, and imprecise and sometimes counterfactual media portrayals. Such conflicting expectations are bound to have unpredictable consequences, especially for neophytes. The first step for police is to revise CompStat so that pressures are more reasonable, management/supervisors take a team approach

(rather than a top-down style common to CompStat), and meetings involve due process considerations such as obeying legal restrictions. Constitutional protections for all citizens must rank with crime control as an integral part of the CompStat performance management system.

These policing tactics have resulted in weak police–community relationships, particularly with male minority youth. Data from the qualitative interviews suggest that youth in these communities are subjected to frequent forcible stops, many of which are unconstitutional. Most of these youth are completely innocent of any wrongdoing (NYCLU 2016 ). Most youth expressed resentment that police stopped them based on little or no evidence. Police rarely explained to them why they were being stopped and searched. Such feelings of resentment seem warranted given the level of intrusion and lack of any explanation. Officers rarely apologized to these youths for these stops. Officers need to follow the law,

Constitutional protections for all citizens must rank with crime control as an integral part of the CompStat performance manage-

ment system.

Forcible Stops: Police and Citizens Speak Out 191

and even when legally sound stops are conducted, they should, at a minimum, explain their actions and apologize to innocent suspects.

Research suggests that youth who experience the justice system as fair may be less likely to offend. According to Sarre and Langos ( 2013 ), youth are more likely to comply with the law if they view police as being legitimate agents of the state. Therefore, based on what police have been doing, some people, especially in minority communities, do not see police as a legitimate institution. It is recommended that establishing and maintaining police legitimacy become a priority.

Our research suggests officers are aggressively conducting forcible stops. However, based on the qualitative part of our study, once officers reach their activity targets there is little to no police presence in the community as officers disengage. These officers are rarely found when true criminality is taking place. Officers on paper appear proactive. In actuality, however, they may be causing more problems with the communities they serve. They become reactive and nonresponsive to community needs. Again, they are not seen as a legitimate agency of the state. Officers need to be more proactive and visible throughout their shifts and not focused on making numbers. One respondent felt that there should be more officers on the street to help lower crime; however, other respondents suggest that if new officers are hired, they need to work with the community.

Police agencies need to consider community partnerships. For example, Choi and Choi ( 2012 ) suggest that community partnerships with police will make the police more effective at fighting crime. Citizens who see something will rarely say anything if they feel victimized by the police. Basic crime detection intelligence will slip through the cracks as officers alienate minority youth. Additionally, labeling theory suggests that once youth are stigmatized as criminal (through forcible stops), they are more likely to become deviant (Eterno and Silverman 2012 ). For example, a forcible stop leading to a minor violation such as drinking alcohol in a public place makes it more difficult for youth to establish themselves as law abiding. In fact, the police themselves are now less likely to hire such individuals as officers.

Nonwhite officers felt more pressure compared with white officers. This also coincided with more educated officers feeling more pressure. It is possible that nonwhite and more educated officers are more aware of the questionable (or even illegal) nature of these practices. While more research is required, the data suggest that the hiring of nonwhite and more educated officers may allow more critical thinking at all levels rather than mindless following of what appear to be illegal orders from superiors. Officers should be problem solvers, not automatons (Goldstein 1977 , 1990 ). This may also help with community partnerships as well.

Forcible stops can be a useful tool for police if used properly. A recent analysis indicates that the court-appointed monitor in the Floyd case found that officers are still not following the basic rules of

stop and frisk (Baker 2016 ). For example, they rarely give receipts to those stopped, as now required. This may indicate a more profound problem with the NYPD beyond training. It may mean the police culture itself needs to be addressed. This can be problematic, as research on the police culture indicates it is a powerful influence on officers (Crank 1995 ; Drummond 1976 ; Manning 1989 ). Nevertheless, the data-driven recommendations contained here specifically address both police and community concerns and deserves urgent attention.

These conclusions and recommendations stem from the importance of the CompStat performance management system. The way this likely works is that commanders are held directly responsible for what occurs in their commands. Some areas known for high crime were ravaged by the police with forcible stops (see Floyd ) without regard to the health of the community so that commanders could cover themselves at the CompStat meetings (with stop-and-frisk reports). Lower-level street officers also felt the high pressure to produce numbers. Anecdotal evidence suggests that this became problematic when evaluations, vacations, and assignments were doled out based on officer ’ s activity. If officers did not have good activity (e.g., a significant number of stop-and-frisk reports), they might be given poor evaluations, denied vacation, or given unpalatable assignments. This led ultimately, and perhaps inevitably, to poor police–community relationships.

Notes 1 . See Terry v. Ohio, 392 U.S. 1 (1968). 2 . See the Center for Constitutional Rights’ summary of the case at https://

ccrjustice.org/home/what-we-do/our-cases/daniels-et-al-v-city-new-york (accessed September 29, 2016).

3 . Floyd v. City of New York (2013), Opinion and Order of United States District Court Southern District of New York, 08 Civ. 1034 (SAS), http://ccrjustice.org/ sites/default/files/assets/Floyd-Liability-Opinion-8-12-13.pdf (accessed September 29, 2016).

References Babbie , Earl . 1989 . The Practice of Social Research , 5th ed. Belmont, CA : Wadsworth . Baker , Al . 2016 . New York Police Still Struggle to Follow Street Stop Rules, Report

Finds . New York Times , February 17 . Bayley , David H. 1994 . Police for the Future . New York : Oxford University Press . Bourbonnais , Renee , Francois Meyer , and Gilles Theriault . 1988 . Validity of Self

Reported Work History . British Journal of Industrial Medicine 45 : 29 – 32 . Bradburn , Norman M. 1983 . Response Effects . In Handbook of Survey Research , edited by Peter Rossi , James D. Wright , and Andy B. Anderson , 289 – 328 . New York : Academic Press . Bradburn , Norman M. , Lance J. Rips , and Steve K. Shevell . 1987 . Answering Autobiographical Questions: The Impact of Memory and Inference on Surveys . Science 236 : 157 – 61 . Brunson , Rod K. 2007 . Police Don ’ t Like Black People: African-American Young Men ’ s Accumulated Police Experiences . Criminology & Public Policy 6 ( 1 ): 71 – 102 .

Brunson , Rod K. , and Jody Miller . 2006 . Young Black Men and Urban Policing in the United States . British Journal of Criminology 46 ( 4 ): 1 – 28 .

Brunson , Rod K. , and Ronald Weitzer . 2009 . Police Relations with Black and White Youths in Different Urban Neighborhoods . Urban Affairs Review 44 ( 6 ): 858 – 85 .

While more research is required, the data suggest that the hiring of nonwhite and more educated offi cers may allow more critical thinking at all levels rather than

mindless following of what appear to be illegal orders from

superiors.

192 Public Administration Review • March | April 2017

Choi , Cheon Geun , and Sang Ok Choi . 2012 . Collaborative Partnerships and Crime in Disorganized Communities . Public Administration Review 72 ( 2 ): 228 – 39 .

Crank , John P. 1995 . Understanding Police Culture . Paper presented at the Annual Meeting of the Academy of Criminal Justice Sciences , Boston , March 7–11.

Dillman , Don A. 1983 . Mail and Other Self-Administered Questionnaires . In Handbook of Survey Research , edited by Peter Rossi , James D. Wright , and Andy B. Anderson , 359 – 77 . New York : Academic Press .

Dowler , Kenneth . 2005 . Job Satisfaction, Burnout, and Perception of Unfair Treatment: The Relationship between Race and Police Work . Police Quarterly 8 ( 4 ): 476 – 89 .

Drummond , Douglas . 1976 . Police Culture . Beverly Hills, CA : Sage Publications . Eterno , John A. 2001 . Zero Tolerance Policing in Democracies: The Dilemma of

Controlling Crime without Increasing Police Abuse of Authority . Police Practice and Research: An International Journal 2 ( 3 ): 189 – 217 .

———. 2003 . Policing within the Law : A Case Study of the New York City Police Department . Westport, CT : Praeger .

———. 2006 . Gender and Policing: Do Women Accept Legal Restrictions More than Their Male Counterparts? Women and Criminal Justice 18 ( 1–2 ): 49 – 78 .

Eterno , John A. , and Eli B. Silverman . 2012 . The Crime Numbers Game : Management by Manipulation . Boca Raton, FL : CRC Press .

Eterno John A. , Arvind Verma , and Eli B. Silverman . 2016 . Police Manipulations of Crime Reporting: Insiders’ Revelations . Justice Quarterly 33 ( 5 ): 811 – 35 .

Evans , Douglas N. , and Cynthia-Lee Williams . 2015 . Stop, Question, and Frisk in New York City: A Study of Public Opinions . Criminal Justice Policy Review . Published electronically on October 8. doi: 10.1177/0887403415610166 .

Fagan , Jeffrey , and Garth Davies . 2000 . Street Stops and Broken Windows: Terry, Race and Disorder in New York City . Fordham Law Journal 28 ( 2 ): 457 – 504 .

Fallon , Kaitlyn . 2013 . Stop and Frisk City: How the NYPD Can Police Itself and Improve a Troubled Policy . Brooklyn Law Review 79 ( 1 ): 321 – 45 .

Ferrandino , Joseph . 2013 . The Efficiency of Frisks in the NYPD, 2004–2010 . Criminal Justice Review 38 ( 2 ): 149 – 68 .

Fine , Michelle , Nick Freudenberg , Yasser Payne , Tiffany Perkins , Kersha Smith , and Katya Wanzer . 2003 . Anything Can Happen with the Police Around: Urban Youth Evaluated Strategies of Surveillance in Public Places . Journal of Social Issues 59 ( 1 ): 141 – 58 .

Fratello , Jennifer , Andres F. Rengifo , and Jennifer Trone . 2013 . Coming of Age with Stop and Frisk: Experiences, Self-Perceptions, and Public Safety Implications . New York : Vera Institute of Justice, Center on Youth Justice .

Gau , Jacinta , and Rod K. Brunson . 2010 . Procedural Justice and Order Maintenance Policing: A Study of Inner-City Young Men ’ s Perceptions of Police Legitimacy . Justice Quarterly 27 ( 2 ): 255 – 79 .

Gerrish , Ed . 2016 . The Impact of Performance Management on Performance in Public Organizations: A Meta-Analysis . Public Administration Review 76 ( 1 ): 48 – 66 .

Glaser , Barney , and Anselm L. Strauss . 1967 . The Discovery of Grounded Theory . Hawthorne, NY : Aldine .

Goldstein , Herman . 1977 . Policing in a Free Society . Cambridge, MA : Ballinger . ———. 1990 . Problem-Oriented Policing . New York : McGraw-Hill . Greene , Judith A. 1999 . A Case Study of Policies and Practices in New York City .

Crime & Delinquency 45 ( 2 ): 171 – 87 . Harcourt , Bernard F. 2001 . Illusion of Order : The False Promise of Broken Windows

Policing . Cambridge, MA : Harvard University Press . Hinds , Lyn . 2007 . Building Police–Youth Relationships: The Importance of

Procedural Justice . Youth Justice 7 ( 3 ): 195 – 209 . Johnson , Bruce , Andrew Golub , Eloise Dunlap , and Stephen J. Sifneck . 2008 .

An Analysis of Alternatives to New York City ’ s Current Marijuana Arrest and Detention Policy . Policing 31 ( 2 ): 226 – 50 .

Jones , Jeffrey M. 2015 . In U.S., Confidence in Police Lowest in 22 Years. Gallup, June 19 . http://www.gallup.com/poll/183704/confidence-police-lowest-years. aspx [accessed September 27, 2016].

Jones-Brown , Delores . 2000 . Debunking the Myth of Officer Friendly: How African American Males Experience Community Policing . Journal of Contemporary Criminal Justice 16 ( 2 ): 209 – 29 .

Jones-Brown , Delores , Brett Stoudt , Brian Johnston , and Kevin Moran . 2013 . Stop, Question and Frisk Policing Practices in New York City : A Primer (Revised) . New York : Center on Race, Crime and Justice, John Jay College of Criminal Justice, City University of New York .

Manning , Peter K. 1989 . The Police Occupational Culture in Anglo-American Societies . In The Encyclopedia of Police Science , edited by William G. Bailey , 472 – 75 . New York : Garland .

Margeson , William A. 2014 . Bringing Down the Gavel on Stop and Frisks: The Equitable Regulation of Police Power . American Criminal Law Review 51 : 739 – 72 .

Mastroski , Stephen D. , Michael D. Reisig , and John D. McCluskey . 2002 . Police Disrespect toward the Public: An Encounter-Based Analysis . Criminology 40 ( 3 ): 515 – 51 .

Neuman , Lawrence . 2000 . Social Research Methods : Qualitative and Quantitative Approaches , 4th ed. Boston : Allyn & Bacon .

New York City Office of the Public Advocate . 2013 . Breaking through Bureaucracy: Evaluating Government Responsiveness to Information Requests in New York City . New York : Office of the Public Advocate .

New York Civil Liberties Union . 2016 . Stop-and-Frisk Data. http://www.nyclu.org/ content/stop-and-frisk-data [accessed September 29, 2016].

Packer , Herbert . 1966 . The Courts, the Police, and the Rest of Us . Journal of Criminal Law and Criminology 57 : 238 – 40 .

Paddock , Barry . 2013 . Kelly: Stop and Frisk Saves Lives by Getting Guns Off the Streets . New York Daily News , April 4 . http://www.nydailynews.com/new-york/ kelly-stop-and-frisk-saves-lives-article-1.1306828 [accessed September 27, 2016].

Penner , Erika K , Jody L. Viljoen , Kevin S. Douglas , and Ronald Roesch . 2014 . Procedural Justice versus Risk Factors for Offending: Predicting Recidivism in Youth . Law and Human Behavior 38 ( 3 ): 225 – 37 .

Radin , Beryl A. 2015 . Baltimore: When Good Intentions Bring Negative Consequences . Public Administration Review 75 ( 4 ): 511 – 12 .

Sarre , Rick , and Colette Langos . 2013 . Policing Young People: Can the Notion of Police Legitimacy Play a Role? Journal of the Institute of Justice & International Studies 13 : 1 – 10 .

Silverman , Eli B. 2001 . NYPD Battles Crime : Innovative Strategies in Policing . Boston : Northeastern .

Stewart , Walter F. , James A. Tonascia , and Genevieve M. Matanoski . 1987 . The Validity of Questionnaire-Reported Work History in Live Respondents . Journal of Occupational Medicine 29 ( 10 ): 795 – 800 .

Stoudt , Brett G. , Michelle Fine , and Madeline Fox . 2011 . Growing Up Policed in the Age of Aggressive Policing Policies . New York Law School Law Review 56 : 1331 – 70 .

Strokes , Larry D. , and James F. Scott . 1996 . Affirmative Action and Selected Minority Groups in Law Enforcement . Journal of Criminal Justice 24 ( 1 ): 29 – 38 .

Tyler , Tom R. , Jeffrey Fagan , and Amanda Geller . 2014 . Street Stops and Police Legitimacy: Teachable Moments in Young Urban Men ’ s Legal Socialization . Journal of Empirical Legal Studies 11 ( 4 ): 751 – 85 .

U.S. Department of Justice, Civil Rights Division . 2016 . Investigation of the Baltimore City Police Department . https://www.justice.gov/opa/file/883366/ download [accessed September 27, 2016].

Weisburd , David ., Mastrofski , Stephen D. , Ann Marie McNally , Rosann Greenspan , and James J. Willis . 2001 . CompStat and Organizational Change: Findings from a National Survey . Report submitted to the National Institute of Justice by the Police Foundation. Washington, DC : Department of Justice .

Zhao , Jihong (Solomon), Ni He, and Nicholas Lovrich. 2002 . Predicting Five Dimensions of Police Officer Stress: Looking More Deeply into Organizational Settings for Sources of Police Stress . Police Quarterly 5 ( 1 ): 43 – 62 .

Copyright of Public Administration Review is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.