Introduction

Within the past three decades, attitudinal research in policing has received tremendous scholarly attention from scholars who are genuinely interested in policing matters. Majority of these research efforts have focused on enhancing relationships between the police and various segments of the population. In doing so, these studies have identified numerous factors that influence the way citizens view the police and have argued that improvement in the factors will yield favorable outcomes for the police. For instance, previous studies have observed the effects of neighborhood characteristics, demographic factors, victimization, and police-citizen contact (Boateng, 2016; Boateng, 2018; Nofziger & Williams, 2005; Ren, Cao, Lovrich, & Gaffney, 2005; Wu, Sun, & Triplett, 2009).

Despite the impressive research efforts in this area of policing, one particular question remains unanswered: to what extent does residential stability influences residents’ opinion, behavior, and attitudes toward the police? This important question has not been answered primarily due to the lack of research examining the relationship between residential stability and attitudes toward the police. Residential stability has been used in previous studies to evaluate its relationship with crime, with several studies finding an association between residential stability and crime levels as well has citizen perception of political inequities (Hajnal & Trounstine, 2014; Morenoff & Sampson, 1997; Sozer & Merlo, 2013; Stark, 1987). However, the few studies in policing need some acknowledgments. While some of these studies have concluded that residents who own a home or resided at their residence for an extended period had an impact on citizen’s perception of the police (Bradford, 2011; Kochel & Weisburd, 2017; Marschall & Shah, 2007; Serra, 1995; Taylor & Lawton, 2012), others have failed to identify a significant association between these variables, or returned results opposite of common assumptions (Gau, Corsaro, Stewart, & Brunson, 2012; Holmes, Painter II, & Smith, 2017; Pattillo, 1998; Rosenbaum, Lawrence, Hartnett, McDevitt, & Posick, 2015; Sharp & Johnson, 2009; Taylor & Lawton, 2012; Warner & Pierce, 1993; Warner & Rountree, 1997). Both the lack of adequate research examining this relationship and the inconsistencies among the few existing studies require further examination of the extent to which residential stability variables can impact citizens’ assessments of the police.

The primary purpose of this study is therefore to evaluate the significance of residential stability and its impact on public’s trust in the police. This study attempts to advance our understanding of public trust in the police in two ways. First, findings obtained from this study will provide further insights into how local police departments can improve relationships with various communities by paying attention to specific residential needs.

Literature Review

Residential Stability and Institutional Trustworthiness

The significant variable being investigated is residential stability. Residential stability is often characterized by several elements. Traditionally the length of stay within a neighborhood is considered for the determination of residential stability. Previous studies have found that homeowners tend to remain in the neighborhood for longer periods of time and become economically and socially connected to the area (Bolan, 1997; Blum & Kingston, 1984; Campbell & Lee, 1992; Frank, Smith, & Novak, 2005; Lee, Oropesa, & Kanan, 1994; Oh, 2004; Speare, Goldstein, & Frey, 1975). Prior literature have also found that neighborhoods that portray characeristics such as poverty, heterogeneous racial and ethnic residents, and residential mobiity produce lower social cohesion and distruct among residents in which has an impact on residents trust and attitude towards the police (Gau et al., 2012; Nix, Wolfe, Rojek, & Kaminski, 2015; Lee, Kim, Woo, & Reyns, B., 2017; Sampson & Jeglum-Bartusch, 1998; Zhao & Ren, 2015). The social context within a neighborhood such as social cohesion, informal social control, level of collective efficacy has an impact on crime and disorder (Sampson, Raudenbush, & Earls, 1997). Neighborhoods or communities that display high levels of social disorganization such as poverty, high mobility among residents, and low social cohesions will reduce the overall collective efficacy within the community.

The concept of residential stability has been conceptualized in a variety of ways. For instance, the concept has been operationalized as citizens who consider themselves a homeowner or a renter (Hajnal & Trounstine, 2014; Rosenbaum et al., 2015; Sharp & Johnson, 2009; Taylor & Lawton, 2012). Other studies have operationalized residential instability as residents who lived in the same house for a period of time, whether the house was owner-occupied, the percentage of residents who did not live at the same house in the previous five years and the percentage of renters who occupied the units (Gau et al., 2012; Isom, 2016).

Studies examining the effects of residential stability have considered the importance of homeownership and years of residence on citizen’s perception of the government and the police. Marschall and Shah (2007) observed that homeownership was a significant variable in the determination of trust in local police. In relation to this, residents who considered themselves renter’s was found to have an impact on citizen’s perception of procedural fairness of the criminal justice system and citizen’s satisfaction with local services (Bradford, 2011; Serra, 1995).

Prior research has found significant correlations between resident’s who own their homes to their perception of the police (Kochel & Weisburd, 2017; Taylor & Lawton, 2012). Kochel and Wiesburd (2017) looked at how the implementation of hot spot policing had an impact on citizen’s perception of procedural justice, police abuse and the perception of legitimacy. In the study, years at address and homeownership were variables included in the demographic characteristics. They found that years at address had an impact on procedural justice, police abuse, and legitimacy. However, they found that residents who owned their home were not a significant variable. In contrast, other studies have found that there were no significant correlations between residents who own their home on citizen’s perception of the government and the police (Gau et al., 2012; Holmes et al., 2017; Isom, 2016; Rosenbaum et al., 2015; Sharp & Johnson, 2009). To address this issue of mixed results, this study further evaluates the impact of whether homeownership is correlated to trust in the police. By evaluating homeowner trust in the police compared to renters, it will provide an appropriate relationship for residential stability and trust in the police. Based on the reviews above, we hypothesize that residential stability measures will influence residents’ trust in the local police. Specifically, we expect that:

  1. 1.

    Residents who own homes will report higher trust in the police than those who rent.

  2. 2.

    Residents who have stayed longer in the community will have higher trust in the police than those who are newcomers.

  3. 3.

    Levels of trust in the police will vary based on type of residence.

Determinants of Citizens’ Attitudes toward the Police

Past attitudinal research has found the effects of several attitudinal variables on trust and confidence in the police. For instance, studies have found that public perceptions toward governmental institutions influence citizen’s level of confidence in the police (Boateng, 2018; Boateng, Lee et al., 2017; Boda, 2015; Kwak, San Miguel, & Carreon, 2012; Stack & Cao, 1998; Van Craen, 2013). Kwak et al. (2012) examined factors that influence the level of confidence in the police among Mexican citizens. According to the article, citizens who supported the regime institution and local governments had a positive effect on confidence in the police. Similarly, Cao and Zhao (2005), examined and found support that trust in the government had an impact on the level of confidence in the police. Trust in judges is a small component of public institutions). Confidence in the police can be a part of a larger political complex and trust in local judges can be a small component of public institutions (Cao & Zhao, 2005; Stack & Cao, 1998). Generally, citizens who have higher confidence in institutions such as the courts tend to have higher confidence in the police (Boateng, Lee, & Abess, 2016). Likewise, people who reported higher trust in their government also expressed having more confidence in the police (Cao & Zhao, 2005). Previous studies also have found that satisfaction with the police has an impact on public’s overall perception of the police. Citizens who report low levels of satisfaction with police service tend to show less favorable perception of the police (Smith & Hawkins, 1973). In a similar vein, if citizens are more satisfied with the police, they will have a positive perception of the police.

The effects of demographic characteristics have also been examined by prior research. Studies examining the relationship between race and trust have argued that minority citizens have less trust in the police than their white counterparts (Callanan & Rosenberger, 2011; Engel, 2005; Garofolo, 1977; Hindelang, 1974; Huang & Vaughn, 1996; Lai & Zhao, 2010; Schuck, 2013; Schuman, Steeh, Bobo, & Krysan, 1997; Van Craen, 2013; Wu, 2014). Peck (2015) conducted a meta-analysis of articles pertaining to minorities and their perceptions of the police. According to the systematic analysis, the researcher found that a majority of the studies found that minorities were more likely to hold negative attitudes and perception of the police compared to white citizens. Additionally, there have been studies conducted to determine trusting differences between men and women, and the results of such analyses have been mixed. While some argue that women have a more positive view of the police than men (Callanan & Rosenberger, 2011; Cao, Frank, & Cullen, 1996; Cheurprakobkit, 2000; Ivkovic, 2008; Reisig & Correia, 1997; Reisig & Giocomazzi, 1998) because of the favorable treatment women received relative to that of men, others have argued otherwise, believing that despite men having greater propensity to be more aggressive toward an officer (than women), they still have higher trust in the police than women do. Yet, there are some studies that have found no significant relationship between gender and police trustworthiness (Correia, Reisig, & Lovrich, 1996; Frank, Brandl, Cullen, & Stichman, 1996; Lai & Zhao, 2010; Sampson & Jeglum-Bartusch, 1998; Ren et al., 2005). Prior research has suggested that political ideology may have an impact on citizens’ perception about the police (Benson, 1981; Cao & Hou, 2001; Cao & Zhao, 2005; Ivkovic, 2008; Lai, Cao, & Zhao, 2010; Stack & Cao, 1998). Cao and Zhao (2005), analyzing data obtained from 17 Latin American countries to determine the link between confidence in the police and trust in the political system found that trust in the political system is an important explanatory variable for citizen’s support for the police.

Contacts with the police have been found to significantly influence citizens’ perceptions of the police. The quality of contacts between police officers and citizens have an impact on citizens’ confidence in the police (Alberton & Gorey, 2018; Cao, 2011; Frank et al., 2005; Gau, 2011; Lim, 2015; Ren et al., 2005; Skogan, 2005; Wells, 2007). Citizens who have positive contact with the police are more likely to have higher trust in the police. In addition, the frequency of contact between citizens and the police is another determinant of trust in the police. Prior research has found that a higher frequency of police-citizen contacts leads to negative attitudes toward the police (Correia et al., 1996; Warren, 2010). This suggests the more frequently citizens are in contact with the police, the more negative their views become.

Methods

The Office of Community Oriented Policing Services (COPS) is the federal office responsible for advancing community-oriented policing and it promotes community policing via various initiatives. The Western Regional Institute for Community Oriented Public Safety (WRICOPS) is one of those programs. WRICOPS represents a partnership among Washington State University, the Washington Association of Sheriffs and Police Chiefs, and the Criminal Justice Training Commission/Peace Officer Standards and Training organization of its member states.

Data for the current study was from a five-state community policing survey that was administered in 2001. The five states included in the study were Idaho, Montana, South Dakota, Washington, and Wyoming. Prior to the data collection period, the selected sites were previously assessed within 12 to 18 months. The questionnaire featured items regarding citizen’s trust in the police, residency within a community, attitudes towards governmental institutions, political ideology, and contacts with the police. Overall, 923 responses were received and the current analysis focuses on these responses. The study sites were several medium-sized cities in the five-state region which are amid the agricultural surroundings. The descriptive statistics of the respondents are provided in Table 1. The Table shows that about 59% of the respondents were males and 99% were employed. Majority of the respondents were whites (93%) and had received high school education at the time of the survey (49%).

Table 1 Descriptive Statistics of Study Variables (N = 923)
Full size table

In terms of political ideology, almost half (50%) of the respondents were independents, 37% were conservatives and only about 15% self-reported being liberals. Moreover, 84% of the respondents reported owning a home and 16 were renters. Majority of them (80%) reside in single family homes, with few (4%) living in duplex and condos.

Measures

Dependent and Independent Variables

The only dependent variable in this study was an ordinal measure of trust in the police. Survey respondents were asked to describe their levels of trust in their local police departments. Response categories were 1) not at any time, 2) hardly ever, 3) just about always and 4) always.

The effects of three measures of residential stability were assessed on residents’ levels of trust in their local police. The first was Home ownership which measured whether a respondent was a homeowner or a renter (0 = renter and 1 = homeowner). Second, residential longevity, a continuous variable which gauged the number of years one has lived in the community/neighborhood. Third, residence type was categorically measured by asking respondents to indicate the type of residence they resided in (1 = apartment, 2 = single family home, 3 = duplex, mobile home, condominium and 4 = other). The responses of residents who reside in other types of residence were used to compare to responses of all other respondents who participated in the survey.

Control Variables

Considering that residents’ evaluations of the police can be influenced by several other variables apart from the specific characteristics of their residency, we addressed potential spuriousness in the relationship between the residential stability variables and perception of police trustworthiness by including variables that past research has observed to be important. For instance, the evidence about demographic characteristics is substantial and critical to the understanding of and development of a trusting relationship between citizens and the police. As a result, the effects of five demographic variables were accounted for in the model: gender, employment, ethnicity, education, and political ideology. Three of these variables -gender, employment, and ethnicity - were dichotomously measured while two were categorically assessed. Specifically, gender was coded as 0 = female and 1 = male, employment was coded as 0 = unemployed and 1 = employed, and ethnicity was coded as 0 = nonwhite and 1 = white. Respondents’ levels of education was measured as an ordered variable with 1 = less than high school, 2 = high school, and 3 = post-high school. The views of respondents who had attained post-high school were compared to those with less than high school or high school education. Political ideology was nominally measured as 1 = Liberal, 2 = independent and 3 = conservative. Respondents who self-identified as liberals were compared to those who identified themselves as independents and conservatives.

Previous attitudinal researchers in policing have discussed the effects of attitudinal and contact variables on public trust in the police (Boateng, 2018; Cao & Zhao, 2005; Gau, 2011). For this overwhelming evidence, we controlled for two attitudinal variables and two contact variables in the regression model. For the attitudinal variables, we included trust in judges (0 = no trust and 1 = trust) and satisfaction with police, measured on a 7-point scale from 1 = strongly disagree to 7 = strongly agree. This was reversed from its original coding. The two contact variables were: quality of contact (0 = poor and 1 = good) and frequency of contact (0 = one to two contacts and 1 = more than two contacts).

Plan of Analysis

To estimate the effects of residential stability variables on trust in the police controlling for other effects, three major statistical analyses were conducted. First, we run descriptive statistics to gauge the distribution of respondents’ responses across the variables we used in the model. Second, we conducted a collinearity diagnostic analysis to assess whether there are issues with multicollinearity in our data. Third, given the nature of the dependent variable, we conducted ordinal logistic regression to estimate the effects of our predicting variables on the outcome.

Results

To determine the influence of the three residential stability measures on residents’ perceptions of police trustworthiness, we analyzed our data using ordinal logistic regression technique. The results of this analysis are presented in Table 2 below. The overall model fit was significant (X2 = 162.79***, p < .001) and with an R-square of 0.42, the model is a better fit than the null model. The table shows that the relationship between residential longevity and trust is negative and statistically significant (Wald = 3.95, p < .05), controlling for the effects of other variables in the model. This result indicates that the longer an individual stay in the community, the less trust the person has in the police. Another residential stability variable that had a significant relationship with trust was duplex and condo (Wald = 4.02, p < .05). The effect was positive, suggesting that people who stay in duplex or condos have higher trust in the police than those who have other living arrangements such as a mobile home. However, there was no significant variation in residents’ level of trust for those who live apartments, single-family homes, and other living arrangements.

Table 2 Ordinal Regression Estimating the Effects of Residential Stability on Trust in police (N = 923)
Full size table

In addition to the effects of the residential stability variables, five control variables were found to predict trust in the police. Residents’ education, political ideology, trust in judges, and the two contact variables were all found to influence citizens’ levels of trust in the police. While there was no statistically significant difference between people who completed less than high school and those who had a post-high school education, citizens who have high school education reported higher trust in the police than those who have more than high school education. In essence, education is an important predictor of one’s trust in the police. Comparatively, citizens who self-identified as liberals reported lower levels of trust than those who identified themselves as either independents or conservatives.

In terms of the influence of attitudinal variables on police trustworthiness, the results demonstrated a significant correlation between people who expressed having greater trust in judges and their levels of trust in the police. This suggests that trust in the local police is linked to individual’s assessment of courtroom workers such as judges. Finally, while residents who perceived their contact with the police as good reported higher trust in the police than those who considered otherwise, people who reported having more than two contacts with the police expressed having lower trust than those with fewer contacts.

Discussion and Conclusion

In several areas of study, residential in/stability has been observed to influence several behavioral outcomes such as mental health, child abuse, adolescent substance uses and crime/delinquency (Aneshensel & Sucoff, 1996; Coulton, Korbin, Su, & Chow, 1995; Ennett, Flewelling, Lindrooth, & Norton, 1997; Shaw & McKay, 1969). Despite its record of predicting behavior, residential stability has barely been explored to explain citizens’ behavior and attitudes toward their local police departments. This lack deeply affects the extent to which we can formulate policies to strengthen police and community relationship. The purpose of the present study was to explore the predictive effects of three dimensions of residential stability on residents’ perception of police trustworthiness after accounting for the effects of individual, attitudinal, and contact variables. The overall observation gleaned from the ordinal regression analysis was that residential stability indeed matters in explaining citizens’ behavior toward the police. However, specific observations made deserve attention.

In this study, we expected that trust levels will vary based on the type of residence that people have in the community. This expectation was met with the finding that residents who live in duplex or condos expressed higher levels of trust in the police than those who live in other types of residence such as mobile homes. This observation aligns with prior studies that have examined the relationship between residence type and attitudes toward the police (Bradford, 2011; Marschall & Shah, 2007). At the individual level, experiences that people may have may differ by the type of residence they live in, and as research has shown, such experiences are important indicators of trust and confidence in the police. For instance, people living in condos may have an entirely different experience from those living in mobile homes or similar living option. For one thing, condos come with their own built-in community and generally have community events, get-togethers, and clubs that help bring people together. Shared spaces also afford residents the ability to know their neighbors, helping create strong bonds and friendships that may not be possible in other types of residence. As suggested by the social capital theory, this strong support network and cohesion among members will undoubtedly shape their thinking about the police in a positive way.

Moreover, our analysis observed an effect on the length of stay in the community on residents’ assessment of the police. This general observation is consistent with prior research (see Kochel & Weisburd, 2017). However, the exact direction of the effect was not as expected. It has widely been noted that people who stay in a particular neighborhood for a long time develop strong social support network and bonds that insulate them from engaging in certain negative behaviors in the community (see Coulton et al., 1995; Ennett et al., 1997). Therefore, drawing a link between the length of stay and criminal offending. Looking beyond this, because of the social bonds and networking, people who have stayed in the neighborhood longer tend to trust their neighbors and again, gleaning from the social support theory, such trust will translate into trusting local institutions, especially the police that work closing with these communities. Another possible explanation could be that the “old timers” in the neighborhood may have become familiarized with members of the police department, and depending on whether the neighborhood is small or large and or open or close, some members due to their longevity, may have several of the officers as friends. Such people will, therefore, tend to be more sympathetic to the police than those who are new comers, have no friendship ties, and see the police from a distance. Despite these compelling explanations, a conclusion drawn from our analysis indicates that “old-timers” in the neighborhood rather view the police to be less trustworthy. There are some reasons one can speculate to explain this behavior. First, old timers may have stayed in the neighborhood long enough to know and understand the inner workings of their local police and can as well determine whether the police have been meeting the needs and expectations of the community. In essence, communities where there are frequent criminal activities – whether minor/serious and violent/nonviolent, community members will be resentful toward the police for either nonperformance or ineffectiveness. The level of resentment will depend on the length of time one has stayed and one’s level of knowledge in issues facing the community.

Given the abundance of evidence supporting the claim that home ownership significantly predict people’s attitudes toward the police (Taylor & Lawton, 2012), this study did not find any such evidence. Rather, it was observed was that homeowners and renters do not significantly differ regarding their views about the police. While this observation contradicts arguments made by some previous scholars and does not meet one of the study’s expectation, it offers credence to those who believed that homeownership does not matter in explaining variations in public trust in the police (see Holmes et al., 2017; Isom, 2016). This finding suggests a further examination to address possible misconceptions about the relationship between the two variables.

In addition to the effects of the two residential stability variables, the study also observed the importance of specific individual attributes, attitudes, and experiences in explaining opinions about police trustworthiness. For example, it was observed that high school graduates perceived the police to be more trustworthy than their post-high school folks, supporting the work of previous authors. For instance, Boateng et al. (2016) found that the less educated (less than high school and high school) reported greater confidence in the police in South Korea and the US respectively. Also, individuals who self-identified as liberals were less optimistic about the police being trustworthy than those who identified as either independent or conservative. Police institutions are part of the larger criminal justice system that interacts with other subsystems to be effective for the administration of justice. Because of this interaction and the level of dependency that exist among the components, people tend to blame the failure of one on others. It is therefore not surprising that views expressed about one institution are linked to views about others. Consistent with prior research (Boateng et al., 2016; Boateng & Buckner, 2017), we found that positive views about judges translate into positive views about the police and the reason is that both institutions are all part of the system administering justice. Finally, it was observed that residents who had quality contact with the police expressed positive views than people who perceived their contact to be bad. However, people who had more contacts with the police tend to hold negative opinions about the police than those with fewer contacts. These two observations echo the importance of contact in undermining public and police relationship.

There are two limitations that we want to acknowledge and caution readers against further interpretation of our findings. First, our analysis was based on a sample that was predominantly white, employed, and individuals that owned homes. There is obviously lack of variations in the responses that we analyzed and the implication is that the voices of individuals that were minorities, unemployed and do not own homes were not adequately considered and that could impact our results. Second, the data analyzed were fairly old and as a result, our analysis failed to account for recent changes and events that could affect how residents currently view the police. These changes include the use of body-worn cameras by officers and video recording of police-citizens’ interactions by the public. Each of these is important in shaping behavior and attitudes toward the police. Although the lack of consideration of these current events does not negate the importance of our findings, we recommend that future study be conducted to explore the relationship between residential stability variables and citizens’ attitudes toward the police by analyzing current information and considering current events.

In spite of these shortcomings, the findings discussed above add to the complex relationship between the police and residents, and supplement previous efforts in understanding the extent to which key residential stability variables can shape people’s views about the local police. Theoretically, this study adds to the limited research in this area by untangling this complex relationship. Practically, the findings provide important insights for police reform, both attitudinal and structural. While the police will not wish for high turnover in the community – because as shown in this study, new comers have more trust in the police – they should focus on increasing satisfaction with police work in the community. Addressing issues such as crime and disorder will definitely go a long way to help.