Skip to main content
SearchLoginLogin or Signup

Community policing, social capital, and residents’ feelings of safety in Taiwan

Police Practice and Research (2022)

Published onMay 02, 2022
Community policing, social capital, and residents’ feelings of safety in Taiwan
·

Abstract:

The history of community policing has now reached a half-century mark, but evidence regarding its effectiveness in reducing fear of crime has been inconsistent. A closer examination of existing research suggests that there must be an underlying process linking community policing and community residents’ fear of crime. Thus, this study draws on Putnam’s theoretical framework to test a conceptual mechanism of social capital underlying the association between elements of community policing and residents’ feelings of safety. Using survey data from a sample of Taiwanese residents, this study applies structural equation modeling to assess the potential mediating effect of social capital. The results suggest community policing is positively associated with residents’ social capital, residents’ social capital is positively associated with their feelings of safety, and community policing has a significant indirect effect on residents’ feelings of safety through residents’ social capital. Based on the findings of this study, suggestions for future research and criminal justice policy are provided.

Keywords:

Fear of crime, community policing, social capital, social disorder, feelings of safety

Citation:

Lo, T.-Y., Wolff, K. T., Liu, Y.-H., & Tsai, H.-E. (2022). Community policing, social capital, and residents’ feelings of safety in Taiwan. Police Practice and Research. https://doi.org/10.1080/15614263.2022.2067155

Introduction

Community policing has gone through half-century of development since the community crime prevention movement emerged in the 1970s (Skogan, 2008). Community policing has been widely adopted both nationwide in the U.S. and worldwide across western and eastern countries, although there is also considerable variation in the way these tactics have been implemented. Nevertheless, there are several elements of community policing that are generally accepted, among which building cooperative relationships between the police and the community has always been the core tenet of community policing philosophy. Moreover, community policing approaches stressed the active role citizens have to play. As Skogan (2008) pointed out, the police could not rely on their limited capacity to get all their job done on their own. Residents were encouraged to shift their mindsets away from simple demand for protection by the police toward shared responsibility for public safety (Scott and Goldstein, 2005).

Community policing was developed to address a broad range of objectives (Gill et al., 2014), such as increasing the quality of service, improving public satisfaction with the police, and reducing people’s fear of crime. In order to better understand how community policing strategies may generate positive outcomes, the present study applies the concept of social capital to explore the process by which community policing may be associated with greater feelings of safety. Social capital refers to the social relationships and connections that an individual or group possesses and that require reciprocity and trustworthiness to sustain (Putnam, 2000) and has been shown to be associated with greater feelings of safety. In this vein, strategies aimed at increasing social capital among community residents are likely to produce positive outcomes for both law enforcement and the communities which they serve. Applying the concept of social capital to the relationship between community policing and residents’ feelings of safety, this study seeks to address three research questions. First, is social capital associated with individuals’ feelings of safety? Secondly, is community policing associated with higher levels of social capital? And ultimately, does community policing have an impact on residents’ feelings of safety through its effect on residents’ social capital? To our knowledge, no studies on fear of crime have examined the effects of community policing on perception of safety through social capital in the Asian context. Accordingly, the current study which is focused on these associations within Taiwanese culture, contributes to our knowledge on the potential linkages between community policing, social capital, and fear of crime.

We begin the current study by providing a succinct review of the literature surrounding the fear of crime. We go on to discuss the concept of social capital, research surrounding its association with community policing and feelings of safety, as well as the potential role of social capital in the relationship between community policing and feelings of safety. To test the proposed associations, we rely on data drawn from a sample of Taiwanese residents to evaluate whether the proposed relationships hold true in the context of Eastern culture, something that is notably missing from the existing body of research. Following the presentation of our results, we provide some suggestions for future research and criminal justice policy.

Literature Review

Residents’ Feelings of Safety

The idea of measuring and tackling fear of crime was believed to have originated in the United States (Hough, 2017), where the President’s Commission on Law Enforcement and Administration of Justice (1967) recognized the existence of fear of crime and the potential harm caused by it. Since then, over fifty years of research in this area has yielded rich findings both in the United States as well as internationally. Previous research suggests several demographic characteristics such as gender, age, income, education, and race are significant predictors of residents’ feelings of safety1. Women and the elderly were more likely to report having lower levels of feelings of safety than their younger and male counterparts (Baumer, 1978; Hale, 1996). People with higher incomes and higher education, on the other hand, tended to report feeling safer (Hale, 1996). In addition, living in urban cities, as opposed to suburbs or rural areas, was found to be associated with higher levels of fear of crime (Belyea and Zingraff, 1988). In terms of personal and household crime victimization experience, while some research indicated that personal victimization has only a small, if not negligible, impact on fear of crime (Garofalo, 1979), other research found property victimization relative to violent victimization is a significant predictor of higher levels of fear of crime (Smith and Hill, 1991).

Beyond demographics, research has also shown that proximal social environmental factors appear to have an immediate and direct impact on people’s feelings of safety. A variety of structural factors have been identified, including high rates of crime and disorder, racial and socioeconomic segregation, high population density, residential instability, and low social cohesion (Rountree and Land, 1996; Brunton-Smith and Sturgis, 2011). Among this list, social disorder has emerged as an important predictor of individuals perceptions of safety. For example, Skogan and Maxfield (1981) found that people’s perception of disorder were closely associated with their fear of crime. This finding has since been replicated in various research settings, including recent studies using survey data from residents in the U.S. such as Seattle (Helfgott et al., 2020), Houston, (Ren et al., 2019), and California (Hinkle, 2013) as well as from countries outside the U.S. such as Belgium (Hardyns, Pauwels, and Heylen, 2018), Croatia (Butorac et al., 2018), and London, U.K. (Brunton-Smith et al., 2014).

Notably, there is also a sizeable body of research which has examined correlates of fear of crime within Asian countries (Abdullah et al., 2015; Jing et al., 2021; Liu et al., 2009; Roh, Kwak, and Kim 2013; Zhang, et al., 2009), including demographic characteristics and social environmental factors such as perceived disorder, collective efficacy, and policing tactics. However, to our best knowledge, studies on the fear of crime have not explored the potential effects of community policing on perceptions of safety through social capital among Asian samples. Taiwanese society has some unique characteristics which may affect the hypothesized relationships, including unique neighborhood characteristics, stricter neighborhood policing practices, and different collective norms surrounding individual behavior. To further understand the potential links between community policing, social capital and fear of crime under different cultural and social contexts, it is meaningful to explore this association in the Taiwanese context.

Social Capital—Concept, Consequences and Sources

Over the past three decades, there has been a growing consensus among researchers from several different fields that an individual’s social environment plays an important role in their well-being (Coleman, 1988; Putnam, 1995; 2000). Among various factors explored, social capital, or the resources available to individuals or groups through their social relationships has become one of the most popular concepts in social science research (Kawachi and Berkman, 2001; Portes, 1998). More specifically, past research has linked social capital to a number of important outcomes, such as health (Elgar et al., 2011) and employment (Nakhaie and Kazemipur, 2013), as well as criminological outcomes such as homicide (Messner, Rosenfeld, and Baumer, 2004), criminal victimization (Takagi, Ikeda, and Kawachi, 2012), youth delinquency (Özbay, 2008), and perceptions of safety (Scarborough et al., 2010).

Putnam’s (2000) concept of social capital refers to “connections among individuals” (p. 16). Specifically, it consists of generalized reciprocity and generalized social trust among members of a community. According to Putnam, generalized reciprocity is an informal norm that people are willing to follow to help each other because they believe it is beneficial for themselves as well as others. Instead of requiring the knowledge of the exact amount of benefit they will get, people simply have the confidence that their help to others will one day be repaid. This type of generalized reciprocity involves mutual trustworthiness, which is the degree to which people feel others are good and can be trusted. Taken as a whole, generalized trust and the norm of generalized reciprocity are inextricably linked. Here Putnam argued that generalized “thin” trust (i.e., trust in the anonymous others, p. 144) among members of a community is more useful than strong trust among long, intimate personal relations because the former can function in a wider range of one’s daily activities. By and large, it is believed that higher levels of social capital can produce positive consequences at the individual, community, and even national levels, including outcomes related to education and child development, neighborhood safety, economic prosperity, democracy, and general health and happiness.

In terms of neighborhood safety, Putnam (2000) referred to Robert Sampson and his colleagues’ research on collective efficacy to illustrate the positive effects of neighborhood cohesion, trust, and altruism on crime rates. Their concept of collective efficacy, which was defined as the combination of residents’ confidence in mutual support (i.e., social cohesion and trust) and residents’ shared expectations of others’ willingness to intervene for the common good (i.e., informal social control), was found to be inversely related to violence at the neighborhood level (Sampson et al., 1997). Likewise, Rosenfeld, Messner, and Baumer (2001) used social trust and civic engagement as the indicators of social capital to examine its impact on homicide rates and found a significant direct effect of social capital on lower homicide rates, even after taking into account the reciprocal effect of homicide rates on social capital.

While Putnam (2000) did not discuss the relationship between social capital and fear of crime explicitly, research that subsequently applied Sampson and his colleagues’ concept of collective efficacy (which, in Putnam’s point of view, coincides with social capital) has found that higher levels of collective efficacy were associated with lower levels of fear of crime (Abdullah et al., 2015; Brunton-Smith et al., 2014; Gibson et al., 2002; Kochel and Nouri, 2021; Yuan and McNeeley, 2017). Similarly, strong social networks and a sense of belonging were found to be the salient social factors that were associated with less fear of crime (Lorenc et al., 2013).

Given its demonstrated association with perceptions of safety, it becomes important to identify potential sources of social capital as well as the factors that potentially undermine its formation. Existing research in this area suggests that structural characteristics, such as higher concentrated affluence, higher residential stability, and lower population density, for example, are associated with higher levels of collective efficacy (Sampson et al., 1999). On the other hand, other neighborhood characteristics, such as physical and social disorder, were found to reduce levels of collective efficacy (Hipp, 2016) and to increase mistrust (Ross and Jang, 2000).

Importantly, in addition to social-structural and contextual factors, formal institutions and organizations may also play an important role in the development of social capital (Nahapiet and Ghoshal, 1998). In fact, some researchers deemed the resources from public sectors and private sectors so crucial to shape people’s social lives that they could be classified as a “linking” component of social capital (Grootaert et al., 2003, p. 4; World Bank, 2000). One such institution examined in prior research is the police (Grootaert et al., 2003). Below we discuss the relationship between the police and social capital in further detail, highlighting the need for additional research to elucidate the connections between community policing tactics, social capital and feelings of safety.

Community Policing and Feelings of Safety: The Role of Social Capital

As others have argued, assessing the relationship between community policing and fear of crime without considering the public’s role can leave out a significant piece of the puzzle. Given the popularity of social capital within social science research, it is not surprising that the concept has gained considerable attention in criminology, as researchers have long been interested in social determinants of individuals’ fear of crime. A closer look at the literature highlights potential linkages between community policing activities, the generation of social capital among community residents and resulting perceptions of safety. For example, Gill et al. (2014) argued that the effect of community policing on citizens’ feelings of safety may take time to manifest as police must first establish a trusting relationship with community residents, one which fosters social capital, before residents’ fear of crime can be reduced. In other words, an underlying mechanism linking community policing to feelings of safety, such as residents’ social capital, must exist. This is in line with Putnam’s (2000) suggestions that a potentially effective way for police to address crime-related issues includes, “… the creation and activation of local social capital” (p. 344). Furthermore, these relationships are especially important to advocates of community policing who argue that in addition to reducing rates of crime, this style of policing may be capable of producing stronger and more viable communities, as social capital has been linked to numerous positive outcomes. Below we review the existing literature linking each of the central concepts, highlighting the need for additional research in this area.

Social Capital and Perceptions of Safety

Scholars have relied on social disorganization theory to explain the relationship between social capital and fear of crime, emphasizing the roles of social integration and cohesion in facilitating informal social control and ensuring neighborhood safety. At the individual level, higher levels of social capital are likely associated with lower levels of fear, as regular social interactions among other people facilitate diffusion of knowledge and information (Coleman, 1990). Regular social interactions are also likely to contribute to lower levels of fear because of the impact on individuals’ trust and reciprocity toward others (Putnam, 2000).

Existing research provides some evidence to support these assertions. Early research found that social ties, based on the proportion of respondents knowing any neighbor by face or name, reduce fear of crime among residents in Baltimore (Taylor, Gottfredson, and Brower, 1984). Similarly, greater social cohesion was found to be associated with lower levels of perceived risk among residents in Seattle (Rountree and Land, 1996), and social integration was found to be negatively associated with residents’ fear of crime in Los Angeles County (Adams & Serpe, 2000). Furthermore, Ferguson and Mindel’s (2007) study in Dallas, Scarborough et al.’s (2010) study in Kansas City, and Zhao, Lawton, and Longmire’s (2015) study in Houston also found the negative association between social capital and fear of crime.

Findings, however, are not unanimous. For example, Kanan and Pruitt (2002) found that once respondents’ perceptions of disorder were controlled for, there was little evidence that neighborhood integration contributed to perceptions of safety. Similarly, Thomas (2007) found that collective efficacy contributed to the transmission of rumors related to criminal victimization following during the aftermath of Hurricane Katrina and was therefore related to increases in fear through its effect on rumors. Scholars have also suggested that the relationship between social ties and perceptions of safety is likely to be reciprocal in nature, meaning that diminished social ties are both a cause and a consequence of elevated levels of fear (Baumer 1985; Liska et al. 1988).

Outside the United States, findings from a cross-national study of eleven countries found strong support for the association between community cohesion and lower levels of perceived risk of criminal victimization (Lee and Earnest, 2003). However, Villarreal and Silva (2006) found a positive association between levels of social cohesion and the perceived risk of crime among their sample of respondents drawn from a large city in Brazil. Like Thomas (2007), authors suggested that greater social cohesion may actually facilitate more communication related to criminal activity in one’s neighborhood, thus contributing to greater fear of victimization. Most recently, Han (2021) found that a measure of generalized trust was negatively associated with fear of crime among residents in Seoul, Korea. The limited and inconsistent findings among studies focused outside the U.S. highlight the need for additional research utilizing international sample to determine the generalizability of these findings to different cultures.

Community Policing and Social Capital

Scholars have identified three potential pathways by which policing may impact social capital/collective efficacy. First, researchers have asserted that trust and confidence in the police is likely to foster collective efficacy among community residents. When residents view the police as an effective and capable authority, they may be more likely to take collective action in order to address problems facing their community knowing the police will aid them in this effort (Yesberg, Brunton-Smith, and Bradford, 2021). Second, previous research has suggested that police may enhance collective efficacy if they are seen as legitimate authorities. When police are seen as legitimate, shared norms and values are likely to be reinforced and people become more likely to take pro-social action to protect their community (Kochel, 2012).

Finally, and most relevant to the current study, scholars have suggested that certain police strategies, including those commonly associated with community policing, are likely to increase social capital among community residents (Gill et al., 2014). Community policing advocates argue that these strategies are expected to increase collective efficacy by creating additional opportunities for residents to interact with one another (Pino, 2001), as well as increase access to police resources and by stimulating community-based forms of informal social control (Scott, 2002). In a similar vein, Kochel and Gau’s (2019) study using data from a panel community survey found that perceived police engagement was a significant predictor of social cohesion.

However, existing results are not entirely supportive of the link between community policing and the generation of social capital/collective efficacy. For example, among Scott’s (2002) alternative measures of community policing, the frequency of police involvement in neighborhood events was unrelated to levels of social capital. Additionally, Kerley and Benson’s (2000) study found limited evidence that community policing tactics (represented by door-to-door interviews with residents) were associated with community outcomes such as cohesion. Similarly, Renauer’s (2007) study found little support for the idea that community policing tactics are likely to foster informal social control among community residents with low levels of trust in the police.

Limitations of Existing Research and Current Study

Although a wide range of empirical evidence on the relationship between community policing, social capital and perceptions of safety among community residents exists, it is also apparent that a number of important research questions remain unanswered. Primarily, scholars have argued that by focusing on a limited number of outcomes, such as fear of crime, existing research has largely ignored the broader community processes that are presumed to contribute to strong, healthy communities (Bennett, 1998; Kerley & Benson, 2000). While the relationship between social capital and perceptions of safety has been more-or-less established, less is known about the role of community policing in the generation of social capital itself. Findings from the small body of existing research are inconclusive, especially when it comes to examining these relationships outside of the U.S. Critically, there are very few existing studies that explore the potential for social capital to mediate the relationship between community policing efforts on perceptions of safety, which is the primary goal of the current analysis.

Drawing on prior literature, this study seeks to test a conceptual mechanism of social capital underlying the effect of community policing on residents’ feelings of safety. The current study extends the exploration of this hypothetical link beyond major Western countries, which dominate research in this area, to determine whether there is evidence that a similar mechanism exists in Asian settings. Using structural equation modeling to analyze survey data collected from residents of Changhua County in Taiwan, we assess the mediating effect of social capital on the relationship between community policing activities and residents' feelings of safety. The proposed model tested in the current study is shown in Figure 1.

Data and Method

Sample

This study used data from a questionnaire survey of 1,000 residents conducted in Changhua County, Taiwan in 2017. Changhua County is located in central Taiwan. It is the most populous county in Taiwan with around 1.27 million people residing in and its population is only less than six major cities (i.e., Taipei, New Taipei, Taoyuan, Taichung, Tainan, and Kaohsiung) in Taiwan. The strategy to select the participants of this study was by quota sampling (Maxfield and Babbie, 2012), where 12 out of total 26 townships were randomly selected, and the decision of the sample size in each selected township was based on the proportion of the population of each selected township to the total population of the 12 selected townships, where the county government’s open data were the source of population statistics2. Residents of the 12 townships aged 20 to 80 years were eligible to participate. The researcher first reached out to village chiefs in each selected township. Village chiefs then helped recruit their residents through neighborhood meetings at their village activity centers to participate in the study. The researcher met with the participants, distributed the survey questionnaires in person with a paper-and-pencil self-administered format, and collected the answered questionnaires on the day of the survey at each site.

Measures

Dependent Variable—Residents’ Feelings of Safety

Consistent with the general question (e.g., Do you feel safe walking alone in your neighborhood?) that has been used in past research, in this study the survey participants were asked “Do you feel safe to walk alone on the streets near your house during the daytime?” “Do you feel safe to walk alone on the streets near your house during the nighttime?” and “Overall, how safe is your neighborhood?” A five-point Likert scale ranging from 1 (very unsafe) to 5 (very safe) was used for the respondents to answer these questions. Each of these items loaded sufficiently (factor loadings >.8) onto a single latent construct when we call feelings of safety.

Primary Explanatory Variable—Awareness of Community Policing Activities

Although there is no universal definition and operational strategy of community policing, the measure of community policing in this study included three main activities—Community Development Association, Community Safety Conference, and crime prevention information dissemination—that are widely adopted across Changhua County as well as all other cities and counties in Taiwan. In fact, all the police precincts in Changhua County were required to implement the three community policing activities under the supervision of county police department. Community Development Associations specified in the survey are resident-based units that are ran by citizen representatives and supervised by county government. They hold open, regular meetings and often invite local politicians, village chiefs, and staff from local government agencies such as firefighters and police officers to attend. Through the meetings police officers were able to provide the audience with some educational lessons about anti scam and crime prevention measures. The meetings also served as a platform for the local government to present awards toward those citizen volunteers and units that contributed to community safety and service. Community Safety Conferences, on the other hand, are the meetings that are held by police precincts and are thus more focused on community safety issues. Police officers such as branch chiefs and detectives were required to attend these Community Safety Conferences. Representatives from district schools, banks, companies, NGOs were also invited to come. In the conferences private sector representatives and individual citizens could express their concerns and needs for public safety issues. The police could also communicate with private sectors and residents and share useful information regarding crime prevention through these meetings. Crime prevention information posters are the projects that the police and community members identify local crime and disorder problems, and the police disseminate crime prevention information to all residents. The forms of dissemination included posters in neighborhood bulletins, flyers sent to household, and posts via websites and social media platforms.

Instead of measuring the residents’ direct engagement in these activities, this study cast a wider net by measuring their awareness of the three community policing activities. The residents were asked whether or not they know such activities implemented in their community, regardless they had ever actively participated in any. Each answer of the three binary questions was coded 1 for yes and 0 for no. The three items were then combined to form a latent variable representing residents’ awareness of community policing activities. The standardized factor loadings for the three dichotomous indicators were .697, .921, and .842, respectively.

Mediator Variable—Residents’ Social Capital

Following Nahapiet and Ghoshal's (1998) definition of social capital in the relational dimension, this study used five items to include the components of generalized trust and norms of reciprocity. For generalized trust, the question was “Generally speaking, to what extent do you trust your neighbors?” As for norms of reciprocity, the questions were “Are you willing to help your neighbors?” “If you do your neighbor a favor, do you believe he/she will do something good for you in return one day?” and “Do you believe your neighbors will help you watch your house?” As Nahapiet and Ghoshal (1998) proposed, identification, or the sense of belonging individuals have towards a group of people, influences whether or not group values and norms are taken, an additional question asking “Do you think of yourself as a part of the neighborhood?” was also included in this measure. The five questions were asked with a five-point Likert-type scale ranging from 1 (definitely not) to 5 (very probably). These five responses were combined into the latent variable of social capital, and the standardized factor loadings ranged from .594 to .761.

Control Variables

In view of prior research’s findings that social disorder has negative effects on both social capital (Gibson et al., 2002; Hipp, 2016; Ross and Jang, 2000) and feelings of safety (Gibson et al., 2002; Helfgott et al., 2020; Ren et al., 2019), we included the measure of residents’ perceived social disorder as the major covariate in our model. The construct of perceived social disorder in this study was the combination of five disorder situations (i.e., juvenile auto racing on the roads, drunk or homeless people wandering on the streets, rowdy public places such as arcade, violation of traffic regulations or drunk driving, and drug dealing/use). The five items were derived from the survey questions that asked the respondents “Does any of the following circumstances apply to your neighborhood?” with a five-point Likert-type scale ranging from “extremely” to “not at all.” The five indicators represented the latent variable of perceived social disorder and their factor loadings were between .803 and .892.

Some sociodemographic characteristics were also included in the model considering personal background factors could possibly impact feelings of safety as well as social capital. The demographic factors that were considered in the present study were gender, age, education, marital status, employment, income, the length of residence of years, and residential area. Marital status, employment, house type, and urbanicity were coded as binary variables, with value 1 indicating, female, married person, worker, house, and urban area, respectively. For the measure of urbanicity, the decision to classify each township as urban or rural was made based on its population density above or under 1,000 people per square kilometer. Seven of the twelve townships were designated as urban areas, while the remaining five were designated as rural areas. Age was coded as an ordinal variable of five levels ranging from “20-29” (assigned value 1) to “60 and above” (assigned value 5). Education was also coded as an ordinal variable of five levels ranging from “less than a high school diploma” (assigned value 1) to “master’s/doctoral degrees” (assigned value 5). Monthly household income was also an ordinal variable of seven levels ranging from “less than NT$20,000” (assigned value 1) to “more than NT$120,000” (assigned value 7).

In addition, this study took into account residents’ personal and intimate family’s experience of crime victimization and residents’ experience of direct contact with the police. The respondents were asked whether or not they and their family members have experienced crime victimization either at home or in the neighborhood over the past year. The question of the police-citizen encounter was “Have you had any direct contact with the police over the past year, including call for service, reporting a crime, random checkpoints, asking for direction on the road, etc.” Each of the above two questions was coded as a dichotomous variable with value 1 indicating yes and value 0 indicating no.

The descriptive statistics for all of the measures are shown in Table 1. Compared to the demographic characteristics of the overall population of Taiwan, the survey participants were slightly overrepresented by people aged 40 to 59 years and people who were married at the time. The mean household income they reported in the survey was below the national average, while the proportion of gender and educational level in the sample approximated it in the overall population of Taiwan. In terms of the measure of the awareness of community policing activities, nearly half of respondents answered yes while the other half of respondents answered no. As for the measure of social capital, the average level was between “not disagree and not agree (scored 3)” and “agree (scored 4).” For the measure of feelings of safety, the average level was between “not unsafe and not safe (scored 3)” and “safe (scored 4),” suggesting that the average respondent felt relatively safe in their neighborhood.

Table 1. Descriptive Statistics

Mean / Frequency (%)

SD

Min

Max

Dependent Variable

Feelings of safety

Feel safe to walk alone in neighborhood during the daytime

3.92

.73

1

5

Feel safe to walk alone in neighborhood at night

3.48

.99

1

5

Overall perceived safety of neighborhood

3.95

.76

1

5

Explanatory Variables

Awareness of community policing

Community development association

1 = Aware

0 = Not aware

463 (54%)

398 (46%)

- -

0

1

Community safety conference

1 = Aware

0 = Not aware

335 (39%)

526 (61%)

- -

0

1

Crime prevention information posters

1 = Aware

0 = Not aware

516 (60%)

345 (40%)

- -

0

1

Social capital

Think of yourself as a part of the neighborhood

3.94

.71

1

5

Willing to help your neighbors

4.06

.63

1

5

Trust your neighbors in general

3.86

.81

1

5

Believe your neighbors will do something good for you in return

3.75

.77

1

5

Believe your neighbors will help you watch your house

3.59

.98

1

5

Control Variables

Perceived social disorder

Juvenile auto racing on the roads

2.19

.97

1

5

Drunk or homeless people wander on the streets

2.11

.95

1

5

Violation of traffic regulations or drunk driving

2.26

1.01

1

5

Drug dealing or drug use

2.26

.96

1

5

Rowdy public places such as arcade

2.08

.95

1

5

Gender

1 = Female

0 = Male

472 (55%)

389 (45%)

- -

0

1

Age (20-29; 30-39; 40-49; 50-59; 60 and above)

2.90

1.28

1

5

Education

2.78

1.13

1

5

Marriage

1 = Married

0 = Not married

577 (67%)

284 (33%)

- -

0

1

Work

1 = Employed

0 = Not Employed

642 (75%)

219 (25%)

- -

0

1

Length of residence of years

22.16

15.36

1

80

Income in NT dollars

3.33

1.65

1

7

House type

1 = House

0 = Apartment

742 (86%)

119 (14%)

- -

0

1

Victimization experience in the past year

1 = Yes

0 = No

82 (10%)

779 (90%)

- -

0

1

Residential area

1 = Urban area

0 = Rural area

609 (71%)

252 (29%)

- -

0

1

Contact with police in the past year

1 = Yes

0 = No

306 (36%)

555 (64%)

- -

0

1

N

861

Analysis

The unit of analysis of this study was the individual. A structural equation modeling (SEM) approach was used to analyze the hypothesized relationships among variables of interest in the mediation model, as the SEM technique is ideal for constructing latent variables, is practical for simultaneous estimation of multiple equations (Kline, 2016), and is more effective to control for measurement errors when including mediation testing with latent variables (Cheung and Lau, 2008). Given the cross-sectional nature of the survey data with all variables measured concurrently, we specified the directional mediation model based on previous research. In order to reduce the probability of spurious results we include a number of covariates shown in past research to relate to the key constructs in the current study.

As the endogenous variables in the model were constructed from the indicators that were five-point Likert-scale items, the response distribution was asymmetrical. Therefore, instead of using maximum likelihood estimation that assumes multivariate normality, robust weighted least squares (WLS) estimation was used as it makes no distributional assumptions (Kline 2016, p. 323), given that our sample size is also large enough for this method to perform adequately. All analyses were done using Mplus software version 8 with the mean- and variance- adjusted weighted least squares (WLSMV) estimator.3

After listwise deletion of the cases that contained missing values, our final sample size is 861.4 We first specified the key factors in our measurement models with separate sets of indicators correspondent to each factor. Our measurement models analyzed in the confirmatory factor analysis (CFA) were found to have positive and moderate to high inter-correlations of effect indicators of the same constructs, which implied convergent validity of all constructs. The factor loadings in CFA for the latent variables are reported in Table 2. We then specified the structural model to test the direct and indirect effects hypothesized in this study with all considered covariates being included. For the assessment of model fit, we considered several measures of model fit rather than the chi-square test since the chi-square test is sensitive to sample size such that a large sample size usually ends up with statistically significant chi-square. The Root Mean Square Error of Approximation (RMSEA) for the model was .043 (90% CI .039, .047). The Standardized Root Mean Square Residual (SRMR) was .077, the Comparative Fit Index (CFI) was .972, and the Tucker–Lewis Index (TLI) was .967. The above values of fit indexes all suggested the goodness-of-fit test for the full model was acceptable.

Table 2. Confirmatory Factor Analysis (CFA) for Latent Variables

Factor

Indicator

Estimate

Standard Error

Z-Score

Standardized Estimate

Feelings of Safety

Feel safe to walk alone during the daytime

1.000

(constrained)

.815

Feel safe to walk alone at night

.987

.042

23.758***

.805

Overall perceived safety of neighborhood

1.013

.039

25.719***

.825

Community Policing

Community Development Association

1.000

(constrained)

.697

Community Safety Conference

1.320

.094

14.079***

.921

Crime Prevention Information Posters

1.207

.084

14.419***

.842

Social Capital

Think of yourself as a part of the neighborhood

1.000

(constrained)

.748

Willing to help your neighbors

.854

.049

17.473***

.645

Trust your neighbors

1.019

.049

20.767***

.761

Believe your neighbors will do something good for you in return

.784

.047

16.791***

.594

Believe your neighbors will help you watch your house

.855

.045

18.880***

.645

Social Disorder

Juvenile auto racing on the roads

1.000

(constrained)

.803

Drunk or homeless people wander on the streets

1.111

.020

56.150***

.892

Violation of traffic regulations or drunk driving

1.049

.019

54.226***

.842

Drug dealing or drug use

1.025

.020

52.120***

.823

Rowdy public places such as arcade

1.008

.020

50.020***

.809

***p < .001

Results

As shown in Figure 2, our structural equation model indicates that the strongest predictor of residents’ feelings of safety was their social capital, and the standardized coefficient for this predictor was .436 (95% CI = .346, .525). Respondents who reported higher levels of social capital were more likely to report higher levels of feelings of safety. Perceived social disorder accounted for some of the differences in feelings of safety with a standardized regression coefficient of –.284 (95% CI = –.350, –.219), which was slightly smaller than the magnitude of social capital. Respondents who perceived the social environment more disorderly tended to have lower levels of feelings of safety, but the negative effect of perceived social disorder was not as strong as the positive effect of social capital. The direct relationship between residents’ awareness of community policing activities and feelings of safety, however, did not reach statistical significance once the pathway through social capital was considered.

It is noteworthy that what stood out the most in this model was the strong significant relationship between residents’ awareness of community policing activities and residents’ social capital. The standardized effect of community policing activities on social capital was .458 (95% CI = .384, .533). That is, respondents who were aware of community policing activities, compared to those who were not aware, tended to have higher levels of social capital. At the same time, perceived social disorder was negatively associated with residents’ social capital (β =–.188, 95% CI = –.255, –.122). Respondents who perceived their neighborhood more disorderly were more likely to report lower levels of social capital, but the negative effect of perceived social disorder on social capital was relatively small compared to the positive effect of community policing on social capital.

As for the other control variables that were simultaneously accounted for in our full model (see Table 3), the results were mostly consistent with prior research. Gender was found to be the strongest predictor of feelings of safety (β = –.130, 95% CI = –.208, –.052) among all sociodemographic factors. Females, compared to males, reported lower feelings of safety. Married people, compared to those people who were not in a marital status, tended to report lower levels of feelings of safety (β = –.093, 95% CI = –.177, –.009). Besides, respondents who reported higher monthly household income also reported higher levels of feelings of safety (β = .115, 95% CI = .042, .189). As may be expected, people who lived in houses seemed to feel slightly safer than those who lived in the apartments (β = .081, 95% CI = .008, .154). Instead, age, education levels, workers compared to non-workers, the length of residence, personal and household victimization experience over the past year, living in urban versus rural areas, and direct contact with the police over the past year all have no explanatory power on variations of feelings of safety among residents.

On the other hand, age, living in urban versus rural areas, and direct contact with the police were found to be significant predictors of residents’ social capital. Older respondents were more likely to have higher levels of social capital (β = .175, CI = .066, .284). Urban residents tended to report lower levels of social capital than rural residents (β = –.137, CI = –.215, –.060). Lastly, residents who had experience of police-citizen encounter over the past year were more likely to report higher levels of social capital (β = .085, CI = .008, .163).

Table 3. The standardized effects of control variables on mediator variable and dependent variable (N = 861)

Social Capital

Feelings of Safety

Gender

-.062

-.130**

Age

.175**

.079

Education

.025

.004

Marriage

.029

-.093*

Worker

-.006

-.009

Length of residence

.116*

-.029

Income

.072

.115**

House type

.053

.081*

Victimization experience

-.072

-.059

Residential area

-.137**

-.057

Contact with police

.085*

.000

*p < .05; **p < .01; ***p < .001

The mediation analysis is based on Baron and Kenny’s (1986, p. 1176) three-step approach. We also followed MacKinnon, Warsi, and Dwyer’s (1995) explanation of effect decompositions to assess the magnitude of the mediated effect. First, the total effect is the coefficient relating the predictor to the outcome when the mediator is not considered. Second, the direct effect means the coefficient relating the predictor to the outcome when the mediator is considered. Third, the mediated effect, or the indirect effect, is the product of the coefficient relating the predictor to the mediator and the coefficient relating the mediator to the outcome, and the indirect or mediated effect is equal to the value of the total effect minus the direct effect. MacKinnon, Warsi, and Dwyer suggest that by looking at the proportion of the total effect being mediated as well as the ratio of the mediated effect to the residual direct effect could we obtain the information on the relative magnitude of the mediation effect (MacKinnon et al., 1995, p. 43).

The test of mediated effects in our structural equation model supported full mediation of the effect of community policing activities on feelings of safety through social capital. As shown in the upper half of Figure 3, a-path regression coefficient, b-path regression coefficient, and c-path regression coefficient are all significant. The total effect of community policing on feelings of safety was statistically significant at the .001 level when the mediator variable was not considered. However, when the mediating variable was included, the residual effect of community policing on feelings of safety was no longer statistically significant, which suggests the relationship between community policing and residents’ feelings of safety is mediated by community policing’s effect on social capital. As for the relative magnitude of the mediated effect, the standardized total effect was .218, while the standardized indirect effect of community policing on feelings of safety was .200 (i.e., the product of a-path and b-path). The proportion of the indirect effect in total effect was 92%, meaning that 92% of the effect of community policing on feelings of safety operated through social capital. Besides, the residual direct effect (c') is .018, and the ratio of the indirect effect to the direct effect is 11.11 (i.e., the product of a-path and b-path divided by c'-path), meaning that the mediated effect is 11 time larger than the direct effect of community policing on feelings of safety5.

We also adopted Shrout and Bolger’s (2002) recommendation to use the bootstrap method to further assess the significance of the mediated effect. We used the bootstrap method in Mplus with 5,000 draws to estimate standard errors for the mediation effects within 99% confidence intervals. Results showed that the indirect effect of community policing on feelings of safety through social capital was significantly different from zero as the 99-percent confidence intervals did not contain an estimate of zero. These ancillary findings provide additional support for the results presented above, the implications of which we turn to now.

Discussion

This study has several major findings that support the importance of both community policing activities and residents’ social capital when it comes to residents’ feelings of safety. Specifically, results confirm that a conceptual mechanism of social capital (Gill et al., 2014; Putnam, 2000) may give a better understanding of whether and how community policing may contribute to feelings of safety. In line with prior research that indicated the link between collective efficacy and fear of crime (Gibson et al., 2002; Kochel and Nouri, 2021; Yuan and McNeeley, 2017), results from our analysis indicate that social capital has a strong positive impact on feelings of safety even after adjusting for perceived social disorder and other sociodemographic factors. Although residents’ awareness of community policing activities does not have a significant direct effect on feelings of safety, which is consistent with Gill et al.’s (2014) systematic review of the effectiveness of community policing on fear of crime, community policing does have a significant indirect effect on the outcome through social capital. The relationship between community policing and feelings of safety is fully mediated by social capital. In other words, the mediated effect is such strong that it indicates only through promoting social capital can community policing activities enhance individual resident’s feeling of safety. The finding implies that evaluation research of the effectiveness of community policing on reducing fear of crime may also measure residents’ social capital at each time point to capture this underlying mechanism. Put it differently, the demand for community policing to achieve the goal of reducing fear of crime may not be manageable if it missed the link to developing residents’ social capital. Because residents’ feelings of safety appear to be affected more directly by their social capital than by the community policing activities known to the individual. It is generalized social trust and generalized reciprocity held by an individual that more probably determines his or her own feelings of safety. Therefore, a more viable way is to utilize the community policing activities as a platform, a source of knowledge, and the resources for the residents to grow their social capital (Sampson, 2004), which will eventually enhance their feelings of safety in the long run.

It is equally important to note that perceived social disorder is a big risk factor that can undermine the mechanism. Our results suggest that perceived social disorder is a significant covariate in the model. Consistent with prior findings, perceived social disorder has significant, negative effects on both the mediator variable (i.e., social capital) (Gibson et al., 2002; Hipp, 2016; Ross and Jang, 2000) and the outcome variable (i.e., feelings of safety) (Gibson et al., 2002; Helfgott et al., 2020; Ren et al., 2019). However, after accounting for the effect of social disorder, the relationships between community policing, social capital, and feelings of safety remain substantively and statistically significant. In fact, the magnitude of the relationship between the awareness of community policing activities and social capital is much stronger than the magnitude of the relationship between perceived social disorder and social capital. Also, the magnitude of the relationship between social capital and feelings of safety is stronger than between perceived social disorder and feelings of safety.

However, if interpreted from a different angle, residents’ perception of social disorder does matter because it significantly accounts for variation in both social capital and feelings of safety among individuals. Moreover, the negative effect of perceived social disorder on feelings of safety is direct and is hardly mediated by social capital. Put it another way, residents’ perception of social disorder may pose a significant threat to their perception of neighborhood safety, and the role of social capital as a mediator to break the link is quite limited. Even worse, the scenario could be that perceived social disorder has prevented individual social capital from fostering in the first place, and fear of crime is inevitably the result of not only the prevalence of social disorder but also the lack of social capital. In this sense, findings of this study lend some support to broken windows theory (Wilson and Kelling, 1982) and suggest the problem of disorderly environments should be improved. However, there must be thorough consideration regarding how to tackle these issues. Research has shown how excessive formal social control could possibly undermine the community dynamic and cause turmoil (Rose and Clear, 1998). Indeed, effective community policing does not often require arrest to resolve issues of social disorder (Bureau of Justice Assistance, 1994). Instead, community policing philosophy encourages the police to solve community problems with flexible strategies. As emphasized in the final report of the President’s Task Force on 21st Century Policing (2015, p. 20), policies for law enforcement agency training should enhance de-escalation strategies. It could be, for example, stopping a fight without making any official record or giving the person in need a referral to other suitable social or health agencies. Every time a police officer successfully helps residents resolve a neighborhood issue, the residents witness, feel, and learn. The trust, expectation and obligation to help others are nourished within the residents. A successful community policing program will not result in a distrusting and disturbing community. Rather, the success of community policing is accompanied by the increase of residents’ social capital that glue a community together to act for the common good.

Despite the strengths of the current study, there are a number of limitations which should be mentioned. First and foremost, the cross-sectional survey data nature did not give us the advantage of examining the causal relationship. The variables of interest were derived at a single time point. Accordingly the directionality of the association between community policing, social capital, and feelings of safety remained unclear, and any potential reciprocal effects between the variables could not be explored. Second, the nonprobability sample used in this study yield results with unclear generalizability. Because the respondents were recruited by the village chiefs at their village activity centers, residents who participated might have significantly higher social capital or higher feelings of safety than the general population in these areas. We were unable to correct for this selection bias, therefore, the issue of external validity remains, and the interpretation of the results can only be applied to a very limited context. Third, the measure of perceived community policing only revealed whether or not residents were aware of those police activities. It did not specify whether the residents directly participated in the activities, and it provided no information about levels of involvement. Future studies may want to tap into more details of the extent to which residents participate in community policing to examine if levels of engagement make a difference. Lastly, although not examined in the current study, the relationship between community policing, social disorder, and feelings of safety may be an interesting focus for future studies.

To conclude, the current study adds to the body of evidence that community policing can be used to foster a generalized trust among police and residents in the community. Our findings show that community policing is associated with residents’ feelings of safety through the process of promoting residents’ social capital. Moreover, the full mediation effect found in this study implies that the impact of community policing on feelings of safety can be easily dismissed if the association between community policing and residents’ social capital is overlooked. Community policing may not directly reduce fear of crime, but it could result benefits such as residents’ social capital which has important implications for residents’ feelings of safety.

References

Abdullah, A., Marzbali, M. H., Bahauddin, A., & Tilaki, M. J. M. (2015). Broken windows and collective efficacy: Do they affect fear of crime? SAGE Open, 5(1), 1–11.

Adams, R. E., & Serpe, R. T. (2000). Social integration, fear of crime, and life satisfaction. Sociological perspectives, 43(4), 605-629.

Baron, R. M. & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173–1182.

Baumer, T. L. (1978). Research on fear of crime in the United States. Victimology, 3, 254–264.

Baumer, T. L. (1985). Testing a general model of fear of crime: Data from a national sample. Journal of Research in Crime and Delinquency, 22(3), 239-255.

Belyea, M. J. & Zingraff, M. T. (1988). Fear of crime and residential location. Rural Sociology, 53(4), 473–486.

Bennett, L. (1998). The uncivil culture: Communication, identity, and the rise of lifestyle politics. PS: Political Sciences and Politics, 31(4), 740-761.

Brunton-Smith, I., & Sturgis, P. (2011). Do neighborhoods generate fear of crime? An empirical test using the British Crime Survey. Criminology, 49(2), 331–369.

Brunton-Smith, I., Jackson, J., & Sutherland, A. (2014). Bridging structure and perception: On the neighbourhood ecology of beliefs and worries about violent crime. The British Journal of Criminology, 54(4), 503–526.

Bureau of Justice Assistance. (1994). Understanding community policing: A framework for action. Washington, DC: Department of Justice.

Butorac, K., Mraović, I. C., & Žebec, M. S. (2018). Fear of victimisation in relation to the key community policing components. Public Security and Public Order, 20, 43–54.

Cheung, G. W. & Lau, R. S. (2008). Testing mediation and suppression effects of latent variables: Bootstrapping with structural equation models. Organizational Research Methods, 11(2), 296–325.

Coleman, J. S. (1988). Social capital in the creation of human capital. The American Journal of Sociology, 94(1), 95–120. https://doi.org/10.1086/228943

Coleman, J. S. (1990). Foundations of social theory. Cambridge, MA: Harvard University Press.

Elgar, F. J., Davis, C. G., Wohl, M. J., Trites, S. J., Zelenski, J. M., & Martin, M. S. (2011). Social capital, health and life satisfaction in 50 countries. Health & Place, 17(5), 1044-1053.

Ferguson, K. M., & Mindel, C. H. (2007). Modeling fear of crime in Dallas neighborhoods: A test of social capital theory. Crime & Delinquency, 53(2), 322-349.

Garofalo, J. (1979). Victimization and the fear of crime. Journal of Research in Crime and Delinquency, 16(1), 80–97.

Gibson, C. L., Zhao, J., Lovrich, N. P., & Gaffney, M. J. (2002). Social integration, individual perceptions of collective efficacy, and fear of crime in three cities. Justice Quarterly, 19(3), 537–564.

Gill, C., Weisburd, D., Telep, C. W., Vitter, Z., & Bennett, T. (2014). Community-oriented policing to reduce crime, disorder and fear and increase satisfaction and legitimacy among citizens: a systematic review. Journal of Experimental Criminology, 10(4), 399–428.

Grootaert, C., Narayan, D., Jones, V. N., & Woolcock, M. (2003). Measuring social capital: an integrated questionnaire. Washington, DC: World Bank Publications.

Hale, C. (1996). Fear of crime: a review of the literature. International Review of Victimology, 4(2), 79–150.

Han, S. (2021). Compositional and contextual associations of social capital and fear of crime. Deviant Behavior, 42(6), 718-732.

Hardyns, W., Pauwels, L. J. R., & Heylen, B. (2018). Within-individual change in social support, perceived collective efficacy, perceived disorder and fear of crime: results from a two-wave panel study. The British Journal of Criminology, 58(5), 1254–1270.

Helfgott, J. B., Parkin, W. S., Fisher, C., & Diaz, A. (2020). Misdemeanor arrests and community perceptions of fear of crime in Seattle. Journal of Criminal Justice, 69, 1–19.

Hinkle, J. C. (2013). The relationship between disorder, perceived risk, and collective efficacy: A look into the indirect pathways of the broken windows thesis. Criminal Justice Studies, 26(4), 408–432.

Hipp, J. R. (2016). Collective efficacy: How is it conceptualized, how is it measured, and does it really matter for understanding perceived neighborhood crime and disorder? Journal of Criminal Justice, 46, 32–44.

Hough, M. (2017). The discovery of fear of crime in the U.K. In: M. Lee, and G. Mythen, eds. The Routledge international handbook on fear of crime. Routledge, 35–46.

Jing, F., Liu, L., Zhou, S., & Feng, J. (2021). How does crime-specific victimization impact fear of crime in urban China? The role of neighborhood characteristics. International Journal of Offender Therapy and Comparative Criminology.

Kanan, J. W., & Pruitt, M. V. (2002). Modeling fear of crime and perceived victimization risk: The (in) significance of neighborhood integration. Sociological inquiry, 72(4), 527-548.

Kawachi, I. & Berkman, L. F. (2001). Social ties and mental health. Journal of Urban Health, 78(3), 458–467.

Kerley, K. R.,& Benson, M. L. (2000). Does community-oriented policing help build stronger communities? Police Quarterly, 3, 46-69.

Kline, R. B. (2016). Principles and practice of structural equation modeling. 4th ed. Guilford Press.

Kochel, T. R. & Gau, J. M. (2019). Examining police presence, tactics, and engagement as facilitators of informal social control in high-crime areas. Justice Quarterly, 38(2): 301–321.

Kochel, T. R. & Nouri, S. (2021). Drivers of perceived safety: Do they differ in contexts where violence and police saturation feel ‘normal’? Journal of Crime and Justice, 44(5), 515-534.

Kochel, T. R. (2012). Can police legitimacy promote collective efficacy? Justice Quarterly, 29(3), 384-419.

Lee, M. R., & Earnest, T. L. (2003). Perceived community cohesion and perceived risk of victimization: A cross-national analysis. Justice Quarterly, 20(1), 131-157.

Li, C.-H. (2016). Confirmatory factor analysis with ordinal data: comparing robust maximum likelihood and diagonally weighted least squares. Behavior Research Methods, 48(3), 936–949.

Liska, A. E., Sanchirico, A., & Reed, M. D. (1988). Fear of crime and constrained behavior specifying and estimating a reciprocal effects model. Social Forces, 66(3), 827-837.

Liu, J., Messner, S. F., Zhang, L., & Zhuo, Y. (2009). Socio-demographic correlates of fear of crime and the social context of contemporary urban China. American Journal of Community Psychology, 44(1), 93-108.

Lorenc, T., Petticrew, M., Whitehead, M., Neary, D., Clayton, S., Wright, K., Thomson, H., Cummins, S., Sowden, A., & Renton, A. (2013). Fear of crime and the environment: Systematic review of UK qualitative evidence. BMC Public Health, 13(1), 496.

MacKinnon, D. P., Warsi, G., & Dwyer, J. H. (1995). A simulation study of mediated effect measures. Multivariate Behavioral Research, 30(1), 41–62.

Maxfield, M. G. & Babbie, E. (2012). Basics of research methods for criminal justice and criminology, 4th ed. Cengage.

Messner, S. F., Rosenfeld, R., & Baumer, E. P. (2004). Dimensions of social capital and rates of criminal homicide. American Sociological Review, 69(6), 882–903.

Nahapiet, J. & Ghoshal, S. (1998). Social capital, intellectual capital, and the organizational advantage. Academy of Management Review, 23(2), 242–266.

Nakhaie, M. R., & Kazemipur, A. (2013). Social capital, employment and occupational status of the new immigrants in Canada. Journal of International Migration and Integration, 14(3), 419-437.

Özbay, Ö. (2008). Does social capital deter youth from cheating, alcohol use, and violence in Turkey?: Bringing torpil in. Journal of criminal justice, 36(5), 403-415.

Pino, N. W. (2001). Community policing and social capital. Policing: An International Journal of Police Strategies & Management, 24, 200-215.

Portes, A. (1998). Social capital: its origins and applications in modern sociology. Annual Review of Sociology, 24, 1–24.

President’s Commission. (1967). The challenge of crime in a free society: a report of the president’s commission on law enforcement and administration of justice. Washington, DC: US Government Printing Office.

President’s Task Force on 21st Century Policing (2015). Final report of the President’s Task Force on 21st Century Policing. Washington, DC: Office of Community Oriented Policing Services.

Putnam, R. (1995). Bowling alone: America's declining social capital. Journal of Democracy. 6: 65–78.

Putnam, R. (2000). Bowling alone: The collapse and revival of American community. New York: Simon and Schuster.

Ren, L., Zhao, J. S. & He, N. P. (2019). Broken windows theory and citizen engagement in crime prevention. Justice Quarterly, 36(1), 1–30.

Renauer, B. C. (2007). Is neighborhood policing related to informal social control? Policing: An International Journal of Police Strategies & Management, 30(1), 61–81.

Roh, S., Kwak, D. H., & Kim, E. (2013). Community policing and fear of crime in Seoul: A test of competing models. Policing: An International Journal of Police Strategies & Management, 36(1), 199-222.

Rose, D. R. & Clear, T. R. (1998). Incarceration, social capital, and crime: implications for social disorganization theory. Criminology, 36(3), 441–480.

Rosenfeld, R., Messner, S. F., & Baumer, E. P. (2001). Social capital and homicide. Social Forces, 80(1), 283–310.

Ross, C. E., & Jang, S. J. (2000). Neighborhood disorder, fear, and mistrust: The buffering role of social ties with neighbors. American Journal of Community Psychology, 28(4), 401–420.

Rountree, P. W., & Land, K. C. (1996). Perceived risk versus fear of crime: Empirical evidence of conceptually distinct reactions in survey data. Social forces, 74(4), 1353-1376.

Sampson, R. (2004). Neighbourhood and community. New Economy, 11(2), 106–113.

Sampson, R. J., Morenoff, J. D., & Earls, F. (1999). Beyond social capital: spatial dynamics of collective efficacy for children. American Sociological Review, 64(5), 633–660.

Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997). Neighborhoods and violent crime: a multilevel study of collective efficacy. Science, 277(5328), 918–924.

Scarborough, B. K., Like-Haislip, T. Z., Novak, K. J., Lucas, W. L., & Alarid, L. F. (2010). Assessing the relationship between individual characteristics, neighborhood context, and fear of crime. Journal of Criminal Justice, 38(4), 819–826.

Scott, J. D. (2002). Assessing the relationship between police-community coproduction and neighborhood-level social capital. Journal of Contemporary Criminal Justice, 18(2): 147–166.

Scott, M. & Goldstein, H. (2005). Shifting and Sharing Responsibility for Public Safety Problems. Washington, DC: Office of Community Oriented Policing Services, US Department of Justice.

Shrout, P. E. & Bolger, N. (2002). Mediation in experimental and nonexperimental studies: new procedures and recommendations. Psychological Methods, 7(4), 422–445.

Skogan, W. G. & Maxfield, M. G. (1981). Coping with crime: individual and neighbourhood reactions. Beverly Hills, CA: SAGE.

Skogan, W. G. (2008). Community policing: origins, concepts and implementation. In T. Williamson (Ed.), The handbook of knowledge-based policing: Current conceptions and future directions (pp. 43–57). John Wiley & Sons.

Smith, L. N. & Hill, G. D. (1991). Victimization and fear of crime. Criminal Justice and Behavior, 18(2), 217–239.

Takagi, D., Ikeda, K. I., & Kawachi, I. (2012). Neighborhood social capital and crime victimization: Comparison of spatial regression analysis and hierarchical regression analysis. Social science & medicine, 75(10), 1895-1902.

Taylor, R. B., Gottfredson, S. D., & Brower, S. (1984). Block crime and fear: Defensible space, local social ties, and territorial functioning. Journal of Research in Crime and Delinquency, 21(4), 303–331.

Thomas, S. A. (2007). Lies, damn lies, and rumors: An analysis of collective efficacy, rumors, and fear in the wake of Katrina. Sociological Spectrum, 27:6, 679-703.

Villarreal, A. & Silva, B.F.A. (2006). Social cohesion, criminal victimization and perceived risk of crime in Brazilian neighborhoods. Social Forces, 84(3), 1725-1753.

Wilson, J. Q. & Kelling, G. L. (1982). Broken windows: The police and neighborhood safety. The Atlantic Monthly, 3, 249(3), 29–38.

World Bank. (2000). World development report 2000/2001: Attacking poverty. New York: Oxford University Press.

Yesberg, J., Brunton-Smith, I., & Bradford, B. (2021). Police visibility, trust in police fairness, and collective efficacy: A multilevel structural equation model. European Journal of Criminology.

Yuan, Y. & McNeeley, S. (2017). Social ties, collective efficacy, and crime-specific fear in Seattle neighborhoods. Victims & Offenders, 12(1), 90–112.

Zhang, L., Messner, S. F., Liu, J., & Zhuo, Y. A. (2009). Guanxi and fear of crime in contemporary urban China. The British Journal of Criminology, 49(4), 472-490.

Zhao, J. S., Lawton, B., & Longmire, D. (2015). An examination of the micro-level crime–fear of crime link. Crime & Delinquency, 61(1):19–44.

Comments
0
comment
No comments here
Why not start the discussion?