|
|
GUANXI AND FEAR OF CRIME IN CONTEMPORARY URBAN CHINA(1) |
|
GUANXI AND FEAR OF CRIME IN CONTEMPORARY URBAN CHINA(1)
Lening Zhang [FNa1], Steven F. Messner, Jianhong Liu, Yue Angela Zhuo
British Journal of Criminolog , July, 2009
Copyright © 2009 by Centre for Crime & Justice Studies (formerly ISTD); Lening Zhang, Steven F. Messner, Jianhong Liu, Yue Angela Zhuo
Western research has investigated individual correlates of fear of crime with a primary focus on people's vulnerability. This vulnerability model examines the possible effects on fear of indicators of people's physical vulnerability (e.g. age and gender) and social vulnerability (e.g. income and education). As is well documented in the research on China, guanxi is a unique aspect of social capital in Chinese society. The present study argues that guanxi in the immediate neighbourhood is an important indicator of the social vulnerability of individuals in urban China. We accordingly hypothesize that residents who have strong neighbourhood guanxi are less likely to be fearful of crime. This hypothesis is assessed with data collected from a recent survey in the city of Tianjin, China. The results of multilevel analysis show that guanxi in the neighbourhood is a significant predictor of fear of crime in contemporary urban China when other important factors are controlled.
Introduction
People's perceptions of public safety and security have been important topics in the criminological literature for several decades. This literature has examined the key social determinants of fear of crime at both the individual level and the neighbourhood level (see Hale 1996 and Pain 2000 for reviews of the literature). With respect to individual-level factors, a major line of inquiry has been guided by a 'vulnerability' model that focuses on the correlations between socio-demographic characteristics and fear of crime (Gates and Rohe 1987; Hindelang et al. 1978; Liska et al. 1988; Skogan and Maxfield 1981). Demographic characteristics presumed to be indicative of physical and social vulnerability, such as being female and having low income, are expected to increase levels of fear of crime (Braungart et al. 1980; Clemente and Kleiman 1977; Taylor and Hale 1986; Will and McGrath 1995). Most studies along this line have examined gender, age, education, income and poverty as important indicators of physical and social vulnerability that are associated with fear of crime (Killias and Clerici 2000; Pantazis 2000; Smith and Torstensson 1997).
Drawing upon the general logic of the vulnerability model, the present study examines a unique indicator of individual vulnerability--guanxi--in explaining fear of crime in contemporary urban China. Guanxi is a form of social capital in Chinese society. It is commonly defined as a personal network that involves instrumental and affective bonds of individuals. The Chinese rely on these networks for virtually every aspect of their life. The present study focuses on a particular aspect of guanxi--close relationships with others in the immediate neighbourhood--and argues that guanxi in the neighbourhood may be an important indicator of social vulnerability of individuals, one that is likely to be significantly associated with fear of crime. Moreover, we expect that the effect of guanxi is likely to emerge despite controls for other predictors of fear of crime that have been proven to be significant in Western research. These hypotheses are assessed using data collected from a recent survey in the city of Tianjin.
The Contemporary Chinese Context
China offers a particularly instructive context for assessing the determinants of fear of crime. Pre-reform China had enjoyed very low crime rates and had earned a reputation of being a virtually 'crime free' society (Fairbank 1987; Rojek 1996). People's fear of crime was not a prominent social issue. However, economic reform in China has resulted in rising crime rates and growing public concern about public safety and security (Liu 2005; 2006). China's official statistics show that there were substantial increases in serious and violent offences from 1978 to 2006. During the 27-year period, the homicide rate was more than doubled; assault increased 7.9 times; robbery grew by 4.7 times; and larceny rose 8.7 times (China Law Yearbooks 1989-2006). As crime rates are surging, Chinese fear of crime and worries about personal safety have also risen, especially in urban areas.
The increasing concern with public safety and security has received a fair degree of attention from the Chinese government and researchers. Since 2001, China's Census Bureau has conducted annual surveys of people's concern with their safety and security using a national sample of 100,000. The surveys report that about 81 per cent of the respondents in 2001 and 84 per cent in 2002 felt safe and fairly safe (Wang and Li 2005). These data suggest that most citizens are not particularly fearful, but official reports are open to question. Some Chinese polling firms have also conducted surveys of Chinese concern about safety and security using national samples (Yuan and Fan 2001; Zen and Yuan 2005). For example, the Ling Dian company conducted a poll of people's opinions on the quality of their life using a national sample of 3,859 in 2004. The safety and security issue was part of the survey. The poll used a Likert - type scale with five response categories ranging from 1 = very unsafe to 5 = very safe and computed the mean score of the responses. The results indicate a mean score of 3.51 (Zen and Yuan 2005). These results also indicate that the general tendency is for Chinese citizens to feel fairly safe (the midpoint of the scale is '3'), but they also suggest that fear of crime has crept into public consciousness to a much greater extent than was the case in the pre-reform days of a 'crime-free society'.
In addition to public opinion polls, Chinese scholars have devoted some attention to the topic of perceptions of safety and security (e.g. Luo and Wang 2004; Shen 2006; Sun and Shi 2004). However, the Chinese studies on this topic that have appeared in professional journals usually involve general discussions or speculations without a guiding theoretical framework and empirical analysis, which reflects the Chinese tradition in social research (Zhou and Cong 2001). Consequently, people's fear of crime and its determinants in China have been understudied. Indeed, the only prior studies of fear of crime in a Chinese city that apply the techniques of Western research are the works by Liu et al. (2008; 2009). These studies (discussed below) offer some support for the vulnerability framework but also indicate the need to adapt the framework to accommodate distinctive features of Chinese society. We extend this line of inquiry by incorporating the important dimension of guanxi into the analysis of the social determinants of fear of crime in contemporary urban China.
Prior Research
The Western research on fear of crime that has been guided by the vulnerability model has focused mainly on the correlations between demographic characteristics and fear of crime. Measures of gender and age have commonly been used as indicators of individual physical vulnerability. The notion is that women and the elderly perceive themselves as less able to protect themselves, 'either because they cannot run fast, or lack the physical prowess to ward off attackers' (Hale 1996: 95). Western studies have indicated that women and the elderly tend to report higher levels of fear of crime compared with men and the younger population. This pattern has been regarded as supportive of the hypothesis that physical vulnerability is associated with greater fear of crime (Braungart et al. 1980; Clemente and Kleiman 1977; Smith and Torstensson 1997; Vanderveen 2002; Will and McGrath 1995; Wittebrood 2002).
Along with examining the hypothesis of physical vulnerability, researchers have also relied on the notion of social vulnerability to inform the search for the social patterning of fear of crime. Disadvantages associated with income, education, race and ethnicity are often considered as indicative of social vulnerability. Racial-ethnic minorities or people who have lower socio-economic statuses are likely to lack material and social resources to protect themselves or their property. 'This may increase the sense of lack of control and, potentially, fear of crime' (Hale 1996: 103). Some studies have examined the effects of income and education (Taylor and Hale 1986; Will and McGrath 1995); others have assessed race and ethnicity (Braungart et al. 1980; Clemente and Kleiman 1977; Covington and Taylor 1991; Liska et al. 1988; Skogan and Maxfield 1981). The findings on social vulnerability have not been entirely consistent, but there is some support for the hypothesis that members of minority groups and persons with low income and less education are more socially vulnerable and thus more likely to report higher levels of fear (Gibson et al. 2002).
A few Western studies have examined the possible effect of local ties on residents' perceptions of their safety or fear of crime (Ferguson and Mindel 2007; Gibson et al. 2002; Sacco 1993; Taylor et al. 1984; Thompson and Krause 1998). Local ties in these studies were usually measured by asking respondents how many neighbours they knew by faces or names, how often they talked with their neighbours or how often they asked for help/assistance from neighbours or were asked for such help/assistance. The studies have yielded mixed findings. Taylor et al. (1984) found that local ties at the block level significantly predicted block-level fear. Gibson et al. (2002) conceptualize these ties as a kind of social integration at the individual level and hypothesize that residents who are more securely integrated into their neighbourhoods should be less likely to be fearful. Using data collected from three US cities, Gibson et al. assess the effect of social integration along with a measure of 'perceived collective efficacy'. They find that the effects of a standard measure of social integration on fear of crime are indirect, operating through perceptions of the degree of collective efficacy in the neighbourhood.
Using similar indicators, Ferguson and Mindel (2007) analyse fear of crime by conceptualizing individual ties in the community as social support networks in a model that also includes other predictors such as perceived neighbourhood incivility and police presence. Their structural equation modelling indicates that social support networks have no significant effect on fear of crime. We propose that guanxi personal networks in Chinese society are distinct from those individual or local ties conceptualized in Western studies (discussed more fully below) and that, as a result, guanxi is likely to emerge as an important determinant of fear of crime in China.
In addition to the vulnerability model and the research on local ties, the victimization model and the disorder model have received considerable attention in the literature on fear of crime (Ferguson and Mindel 2007; Skogan 1990; Skogan and Maxfield 1981; Stafford and Galle 1984; Wilson and Kelling 1982). The victimization model is based on the premise that victims of crime will be more likely to be fearful than non-victims. Despite the intuitive appeal of this model, the evidence has actually been rather mixed (Gibson et al. 2002). The disorder model shifts attention away from individual characteristics to features of the neighbourhood, and more specifically, to various forms of physical and social 'incivilities' (e.g. abandoned buildings, litter, graffiti, drunkenness, prostitution solicitation, rowdy youths, etc.) (Taylor 2001). Research has offered a good deal of support for the proposition that disorder, both as perceived by residents and as reflected in neighbourhood-level indicators of incivilities, increases fear of crime (Gibson et al. 2002: 541).
The literature on fear of crime also includes a number of cross-cultural studies (Murck 1997; Pantazis 2000; Quann and Hung 2002; Vanderveen 2002; Wittebrood 2002). Wittebrood (2002) has conducted multilevel analyses of fear of crime across 16 Western industrialized countries using data collected from the International Crime Victim Surveys (ICVS). Although the study indicates no relationship between nations' demographic characteristics such as proportions of young men, divorces and poverty and fear of crime, it shows a social patterning of fear of crime across all the investigated nations. Women, the elderly and those with low socio-economic statuses are more likely to feel unsafe, consistent with much of the research in the West based on data for a single nation. The study also indicates a strong linkage between victimization experience and fear of crime. Quann and Hung (2002) report similar findings using data collected from ICVS in 17 industrialized countries.
Finally, two prior studies based on survey data for the city of Tianjin have applied the prominent Western frameworks to the study of fear of crime in the Chinese context. Liu et al. (2009) find that females tend to report higher levels of fear, and that those with a high degree of perceived physical strength and self-defence capabilities report lower levels of fear. These patterns are consistent with the logic of the vulnerability framework and with findings in the West. In contrast, they discover that younger respondents and more educated respondents express higher levels of fear than their counterparts. These findings are opposite to those commonly observed in Western research. Liu et al. propose that in the Chinese context, young people and the highly educated are more likely to be exposed to non-traditional media not controlled by the state. As a result, the young and highly educated perceive more disorder in their neighbourhoods and in society and are more fearful of crime. Consistent with this interpretation, they find in their multilevel models that both individual perceptions of disorder and disorder measured at the neighbourhood level are associated with greater fear of crime. In further analyses of the Tianjin data, Liu et al. (2008) report that perceptions of disorder in the neighbourhood mediate the influence on fear of crime of some of the demographic variables and of victimization experiences.
The studies by Liu et al. demonstrate the partial applicability of the Western analytic frameworks to the understanding of fear of crime in China, but they also underscore the importance of adapting these frameworks to the Chinese context. We build upon the prior work of Liu et al. by conceptualizing and operationalizing a variable that reflects a distinctive feature of Chinese society--guanxi in the neighbourhood--and by distinguishing the effect of this variable in multilevel regression models from those of other predictors considered in Western studies of fear of crime.
Guanxi as an Indicator of Social Vulnerability
Both Chinese and Western researchers have observed a relational culture of guanxi that permeates every aspect of Chinese life (Liang [1949] 1986; Fei [1949] 1992; King 1985). Guanxi is commonly defined as a personal network that involves individual bonding, empathy, reciprocity and trust (Geddie et al. 2005). It dictates and facilitates Chinese people's daily life. As the notable Chinese scholar, Liang Shuming, has argued, Chinese society is neither individual-based nor society-based, but relation-based instead (cited from Gold et al. 2002: 10). This cultural characteristic represents a sharp contrast with typical Western individualistic orientations, as reflected in weak personal ties, infrequent interaction and low intimacy. Although Westerners may also have social networks that involve friends, colleagues, former classmates or other relationships, guanxi as a personal network has distinctive features. It is 'based implicitly (rather than explicitly) on mutual interest and benefit. Once guanxi is recognized between two people, each can ask a favor of the other with the expectation that the debt incurred will be repaid sometime in the future' (Yang 1994: 1-2). This reciprocity grounded in obligation and indebtedness makes individuals in the guanxi circle feel that they can rely on others for support in a wide range of aspects of their lives.
Another distinctive feature is the component of affection in guanxi (Bell 2000; Gold et al. 2002; Hillman 2005; Kipnis 1997). Guanxi is not only instrumental for achieving individual interests and benefits. It also involves mutual affection or sentiment (ganqing) that is personal and is derived from extensive exposure of individuals' private lives to each other in the guanxi circle. The subtle combination of instrumental and affective components in guanxi creates a basis of trust in the relationship (Smart 1993). Therefore, individuals may benefit from their guanxi not only instrumentally, but also affectively.
Guanxi is rooted in blood relationships and then extends outward to other connections/relationships (Bell 2000; Zhang and Zhang 2006; Yang 1994). A prominent Chinese sociologist, Fei Xiaotong [1949] (1992), uses the image of ripples formed from a stone thrown into a lake to illustrate the development of guanxi. Family members and relatives are located at the core of the guanxi, while friends, classmates, neighbours, co-workers or other acquaintances are located at more peripheral positions in terms of the centrality of the relationships and the degrees of trust (Zhang and Zhang 2006).
The impact of the profound changes associated with the introduction of market reforms in China on the role of guanxi is a controversial matter. Some scholars have suggested that as a market-oriented economy expands, the role of guanxi networks in people's lives is actually becoming stronger because Chinese life becomes more uncertain during the transformation (Bian 1994; Bian and Zhang 2001). In contrast, others maintain that the significance of guanxi in Chinese life is declining along with the institutionalization of the market economy and the development of formal law (Guthrie 2002; Hanser 2002; Situ and Liu 1996). Still others adopt a more intermediate stance between these two views, proposing that the role of guanxi is being transformed but not eliminated (Potter 2002; Yang 2002). Its role may be declining in some areas, but it continues to flourish in new areas and is manifested in new forms. Although total consensus has yet to be reached among scholars, there is general agreement that g uanxi networks still serve as a significant form of individuals' social capital in Chinese society at present (Gold et al. 2002). It thus seems reasonable to propose that guanxi represents a key social resource that is likely to indicate social vulnerability (or relative social invulnerability), and that it is likely to be associated, in turn, with the likelihood of being fearful of crime.
Analytical Framework
The major hypothesis under examination in this study is that residents who have strong guanxi in the neighbourhood are less likely to exhibit fear of crime. The hypothesis is assessed when other important individual and contextual variables are controlled in a multilevel regression framework. In selecting the control variables, we have drawn upon the previous research and theorizing on fear of crime in the West, as well as the results of the prior Tianjin study.
The regression models include the standard socio-demographic characteristics that have been used as indicators of physical and social vulnerability: gender, age, education and family income. [FN1] We also control for the respondents' ratings of physical strength and self-defence/alertness, which can similarly be viewed as reflecting vulnerability. Two measures represent the victimization model of fear of crime: having been the victim of violent or personal property crime. Our control variable at the individual level drawn from the disorder perspective is respondents' perceptions of disorder in the neighbourhood.
The incorporation of the measures of local ties used in Western studies into our analysis of guanxi entails certain complexities. The concept of guanxi in the neighbourhood overlaps with broad notions of local ties in the West in that it involves social connections with neighbours. However, our underlying premise is that guanxi has distinctive qualities in the Chinese context. To substantiate this premise empirically, we include a measure of local ties that is based on the degree to which neighbours provide various forms of assistance to one another. We also include a measure of 'perceived collective efficacy' following Gibson et al. (2002), as well as Sampson et al. (1997), which is based on a combination of informal social control and social cohesion among neighbours. Before estimating the multivariate regression models, we assess the extent to which our operational measure of guanxi is in fact empirically distinct from the measures of local ties and social cohesion commonly used in Western studies (see Taylor 2002).
In addition to the control variables reflecting individual-level properties, our dataset also permits multilevel analyses. It is thus possible to examine the effect of residents' guanxi in the neighbourhood, net of selected contextual variables. We include five such control variables measured at the neighbourhood level. The first four are analogous to individual-level variables: guanxi, local assistance ties, collective efficacy and social disorder, but these variables refer to aggregate measures rather than individual perceptions. Finally, the regression models also include a measure of neighbourhood disadvantage as a control variable to separate the effect of family income from the possible contextual effect of economic disadvantage.
Data and Methods
Data collection
The data were collected from a multi-stage survey of residents in the city of Tianjin, China, in 2004. [FN2] Tianjin is the third largest city in China, surpassed by only Shanghai and Beijing. [FN3] The city is located to the southeast of Beijing. Tianjin is a major industrial centre and is the most important port of Northern China. The city covers a geographic area of approximately 4,365 square miles.
The survey implemented a multi-stage cluster sampling design that drew approximately 2,500 respondents who were 18 and over. Tianjin has 15 administrative districts and three counties. The sample was drawn from the six traditional districts located in the central urban area of the municipality. These districts include the Heping, Nankai, Hongxiao, Hexi, Hebei and Hedong districts. Each district has approximately six to ten City-Street Offices, which are the grass-roots organizations of the Tianjin government. At the first stage, the survey randomly selected two City-Street Offices from each of the selected districts, yielding a total of 12 City-Street Offices.
At the second stage, the survey drew two large offices that included a relatively large number of neighbourhood committees from the 12 selected City-Street Offices. Five neighbourhood committees were then randomly selected from each of these two large City-Street Offices, while four neighbourhood committees were randomly drawn from each of the remaining 10 City-Street Offices. A total of 50 neighbourhood committees were thus obtained through a combination of purposive and random selection. Members of the research team met the supervisor in each of the selected neighbourhood committees to explain the purpose and importance of the survey, the financial sources of the survey and compensation for costs associated with administration. Upon securing agreements for assistance, the research team requested a complete list of households in that neighbourhood.
Fifty-one households were selected from each of the 50 selected neighbourhoods in hopes of reaching the target of 2,500 households. Using the household roster provided by the neighbourhood committee in each selected neighbourhood, the research team conducted systematic sampling. A starting point was randomly determined and every eighth household from each neighbourhood was selected until the specified number of households was obtained. The research team defined a criterion date for selecting a specific respondent from a selected household with more than one member 18 years old or older. The individual with a birthday closest to the criterion date was chosen to be the respondent.
Data were collected through anonymous, self-administered questionnaires at convenient sites within the neighbourhood (e.g. recreational areas). With the assistance of the neighbourhood committees, the research team contacted the respondents to schedule the administration of the questionnaire. A total of 2,474 valid questionnaires were obtained, yielding an extremely high response rate--97 per cent. Missing data on some variables result in a sample size of 2,455 for analyses of individual characteristics across the 50 neighbourhoods.
[FNa1]. Direct all correspondence to Lening Zhang, Department of Sociology and Criminal Justice, Saint Francis University, Loretto, PA 15940. Voice: (814) 472-3052; Fax: (814) 472-2787; lzhang@francis.edu.
[FN1]. The population in Tianjin is racially homogeneous, and thus no control for race is required.
[FN2]. The survey was conducted in collaboration with the Tianjin Academy of Social Sciences.
[FN3]. The following description of Tianjin is taken from the online version of Encyclopaedia Britannica (www.britannica.com/eb/article-9108664/Tientsin).
|
|
|
|
|
|
