Program Profile: Community-Oriented Policing in New Haven (Conn.)

This is a brief policing intervention for improving public attitudes toward police, through positive nonenforcement police contact. The program is rated Promising. Compared with control residents, intervention household residents reported statistically significant improvements in attitudes toward police and more positive perceptions of police performance and legitimacy and greater willingness to cooperate with police. But there was no statistically significant effect on compliance with police.

A Promising rating implies that implementing the program may result in the intended outcome(s).

This program's rating is based on evidence that includes at least one high-quality randomized controlled trial.

Program Goal/Target Sites

Community-oriented policing is a policing strategy designed to encourage cooperative police–public relationships through a variety of positive nonenforcement interactions, such as community meetings and neighborhood watch initiatives. In 2019, the city of New Haven, Conn., implemented a community-oriented policing intervention to increase public trust and improve perceptions of police legitimacy (Peyton et al. 2019a).

Specifically, this intervention focused on the importance of 1) establishing common connections with residents, 2) emphasizing cooperation toward common goals (such as police needing residents’ help to make the community safe), and 3) forming personalized interactions, as an approach to building public trust and enhancing police legitimacy. Overall, the goal of the intervention was to examine whether a single positive, nonenforcement interaction with a uniformed police officer caused meaningful changes in New Haven residents’ general views about police legitimacy.

Program Activities

The community-oriented policing strategy in New Haven consisted of a single unannounced community home visit conducted by uniformed patrol officers from the New Haven Police Department. Community home visits involved a conversation (about 10 minutes long) between a resident and a uniformed patrol officer from their district. Contact was initiated by patrol officers knocking on residents’ doors between the hours of 6 p.m. and 8 p.m. during weeknights and before 8 p.m. on weekends. These time slots were selected to maximize the number of conversations between patrol officers and residents.

Specifically, patrol officers conducted community home visits with registered voters (with addresses listed in their beat), who participated in the “New Haven Opinion Survey 2018.” During these community home visits, patrol officers performed the following steps:

• Disarmed Anxiety. Patrol officers sought to defuse any potential anxiety within the first 20 seconds of contact with the resident by communicating respect (such as extending a handshake) and emphasizing that the visit was an “equal status engagement” intended to improve the community. Officers made it clear to residents that they were not there in a law enforcement capacity. For example, after introducing themselves, patrol officers would say: “I am one of the officers in this area. Everything’s OK. No one is in trouble, and everyone is safe. I’m here because I care about our community, and today I’m personally introducing myself to our community members.”
• Identify Resident (or Residents). Patrol officers would ask to speak to the registered voter who was on their list. After matching the individual (or individuals) with voter file records, the patrol officer would mark them as contacted in a mobile application installed in their department-issued cell phone. For households with multiple residents, only those with whom the officer made direct contact were marked as contacted.
• Providing Voice. Patrol officers sought feedback from residents and emphasized their common connection toward pursuing a superordinate goal (such as working together to improve the community). Common connections were made by identifying and using signal words (such as “we: and “our community”). For example, patrol officers would ask residents, “How can we work together to improve our community?” or they would say, “Today I’m here to ask about how you’re doing and learn about your views on how we can work together to keep our community safe.”
• Address Concerns. Patrol officers allowed residents to express their concerns or complaints without judgment, followed by the patrol officer acknowledging their perspective in a response that was resident-centered. For example, patrol officers would communicate that both they individually, and the New Haven Police Department more generally, were committed to community policing. Officers sought to demonstrate that they care about community members and their well-being, thus reinforcing their trustworthy motives.
• Make It Personal. Patrol officers personalized interactions by shaking hands with residents and providing their personal business card (with their department-issued cell phone number handwritten on the back in a designated area), containing the name of the patrol officer and additional contact information for the police department. Before leaving, the patrol officer reiterated their commitment to building a cooperative relationship with the resident and the importance of police and residents working together to keep the community safe.

Given the unannounced visits, patrol officers were trained to anticipate nervous or suspicious reactions from residents and not to interpret this as suspicious behavior during conversations.

Key Personnel

Senior officers (e.g., the Assistant Chief of the Patrol Division and Police District Managers) at the New Haven Police Department selected 20 patrol officers from the B-squad (i.e., patrol officers with shifts between 3 p.m. and 11 p.m. or 4 p.m. and 12 a.m.) to participate in the program. The median age of participating patrol officers was 32 (minimum of 25 and maximum of 49), and the median years of service on patrol was 3 (minimum of 2 and maximum of 19). Patrol officers attended a 1-hour training session delivered by the study researchers and a police lieutenant at police headquarters (additional information on the training can be found under Implementation Information).

Program Theory

The New Haven community-oriented policing intervention is based on a social psychological theory which suggests that positive interpersonal contact, under appropriate conditions, can improve intergroup relations (Peyton et al. 2019a). Specifically, the program operates on the idea that interpersonal contact between the public and government officials (such as police officers) is a fundamental part of democratic political socialization, with negative experiences undermining trust and political efficacy (Soss, 1999). These types of positive interpersonal contact are especially salient in the forging positive relationships between police and the public. Even in the case of a single brief interaction, positive interpersonal contact can have a strong effect on public attitudes toward police (Broockman and Kalla 2016).

Study 1

Peyton and colleagues (2019a) conducted a randomized clustered controlled trial to evaluate the effects of a community-oriented policing strategy on individual-level attitudes toward police in New Haven, Conn. The New Haven Police Department (NHPD) is an urban police department with 10 police districts. Study participants were identified and recruited from Aug. 15 to Sept. 6, 2018 (i.e., the recruitment period), using voter registration records that included mailing addresses for 49,757 residents.

During the recruitment period, the study authors mailed invitation letters to the addresses listed in residents’ voter registration records. Mailed letters asked residents to participate in an online survey (called the “New Haven Opinion Survey 2018”). Of the 49,757 potential residents, 2,013 residents in 1,852 households were included in the study sample based on their completion of the online survey and agreement to participate in follow-up surveys. Because residents (n = 2,013) were nested in households (n = 1,852), blocked random assignment was used to match households into quartets (i.e., groups of four houses) across 463 blocks. Household quartets were matched based on blocking variables (such as the proportion of residents above the median income level of $50,000–59,999, average household age, and proportion of residents indicating any prior arrest by the police department) and randomized to the intervention (n = 926) and control (n = 926) conditions. There were 1,007 individuals in intervention households and 1,006 individuals in the control households.

During the intervention period (Sept. 11 to Oct. 17, 2018), patrol officers attempted to contact all 1,007 individuals in intervention households. Individuals in control households did not receive any type of community-oriented policing intervention. Among the 1,007 persons assigned to the intervention, 412 were successfully reached at the door and received the community-oriented policing intervention, which included a brief (about 10-minute) conversation with a uniformed patrol officer from their district.

Of the individuals in intervention households, 62 percent were female, 57 percent were White, 19 percent were Black, 13 percent were Hispanic, and 11 percent were “Other” (Asian, Native American, Middle Eastern, multiracial, or another race or ethnicity). In the last 12 months (before the intervention), 39 percent of individuals had prior face-to-face contact with an officer from NHPD, 9 percent had previously been arrested by NHPD, and 14 percent reported prior unfair treatment by NHPD. The average age of individuals was 44 years, and the average household size was 1. Individuals in the control households were 62 percent female, 57 percent White, 20 percent Black, 12 percent Hispanic, and 11 percent “Other.” Of these individuals, 40 percent had face-to-face police contact with an officer in the last 12 months, 9 percent had previously been arrested, and 15 percent reported prior unfair treatment by NHPD. The average age of individuals was 44 years, and the average household size was 1. There were no statistically significant differences at baseline on background characteristics between individuals in intervention and control households.

In addition to data collected at baseline, two follow-up surveys (emailed online) were conducted 3 days (T1) and 21 days (T2) after the intervention. Outcomes of interests included overall attitudes toward police, willingness to cooperate and comply with police, perceptions of police performance and police legitimacy, confidence in police, perceived warmth toward police, and negative beliefs about police. All the outcomes, except for perceived warmth toward police, were measured by 19 questions derived from a larger, widely used 26-question survey (see Sunshine and Tyler 2003; Tyler et al. 2014; Tyler and Jackson 2014), using a seven-point scale (which ranged from Extremely Unlikely to Extremely Likely or Strongly Disagree to Strongly Agree, with a neutral midpoint). Perceived warmth toward police was measured using a 100-point “feeling thermometer” intended to tap into people’s feelings about the police. Negative beliefs about police were measured using a five-item scale of negative stereotypes about police as a group, with each item scored on a seven-point scale (e.g., “compassionate” (1) to “cold hearted” (7), with a neutral midpoint).

A linear regression analysis was used to estimate the intent-to-treat (ITT) effect to determine the differences between individuals in intervention and control group households, at the 21-day follow-up. Subgroup analyses were conducted to examine the impact of race on overall program effects.

A partnership between the New Haven Police Department and the research team was developed to implement the community-oriented policing intervention in New Haven, Conn. (Peyton et al. 2019a). Although the police department had implemented a similar strategy before 2017 (which included having officers walk beats in each of the 10 police districts to regularly introduce themselves and interact with residents), patrol officer shortages made it difficult to maintain regular walking beats.

Training sessions for patrol officers who participated in the intervention focused on the meaning of positive intergroup contact and the importance of procedural justice in police–civilian interactions. Some of the elements of positive intergroup included:

• Equal status between groups (i.e., police and public) in the situation
• Cooperation toward a common goal
• Support of authorities, law or custom
• Personal Interactions

Meanwhile, efforts to promote procedural justice in police–civilian interactions focused on 1) encouraging citizen participation (i.e., providing voice), 2) promoting dignity and respect, and 3) communicating trustworthy motives (Peyton et al. 2019a). Patrol officers typically began their weekday shift at 3 p.m. or 4 p.m. For the first 1 to 2 hours of each new shift, the incoming patrol officers normally would “clear the board” of any outstanding calls for service that the previous squad was unable to finish before conducting the brief, 10-minute community home visits.

Subgroup Analysis

Peyton and colleagues (2019a) conducted subgroup analyses to examine the impact of race on overall program effects. At the 21-day follow-up, White residents in intervention households who received the community-oriented policing intervention reported more-positive overall attitudes toward police, compared with White residents in control households. Hispanic residents in intervention households also reported more-positive overall attitudes toward police, compared with Hispanic residents in control households. These differences were statistically significant. However, there was no statistically significant difference between Black residents in intervention households and control households in overall attitudes toward police at the 21-day follow-up. There was also no statistically significant difference between individuals of other racial backgrounds in intervention and control households in overall attitudes toward police at the 21-day follow-up.

These sources were used in the development of the program profile:

These sources were used in the development of the program profile: