Victimization,Offending,and Perceived Risk for Early Violent Death

Abstract

Research indicates that adolescents tend to overestimate their risk for early death, and those who anticipate early deaths are at risk for numerous negative consequences. Little is known, however, about the factors that influence early violent death expectations. The present study develops and tests hypotheses about the influence of victimization and offending on perceived risk for early violent death using longitudinal data collected from a nationally representative sample of American adolescents. Logistic regression analysis was used to estimate the effects of prior violent victimization, witnessing serious violence, and various types of delinquency on adolescents’ perceptions that there is a 50% or greater chance they will be killed by the age of 21. Implications for theory, policy, and future research are discussed.

Keywords

perceived risk violence early death expectations victimization

Prior research suggests that adolescents who anticipate dying young are at risk for a range of negative outcomes, including unsafe sexual activity, HIV, arrest, and regular substance use (Borowsky, Ireland, & Resnick, 2009; Fischhoff et al., 2000; Jamieson & Romer, 2008; Nguyen, Hussey, et al., 2012; Nguyen, Villaveces, et al., 2012). Some scholars have suggested that those who are uncertain about their own survival may be more likely to engage in risky behavior because they minimize the future consequences of their behavior (see Brezina, Tekin, & Topalli, 2009; Hill, Ross, & Low, 1997; Nagin & Pogarsky, 2004). For example, some have speculated that adolescents who live in disadvantaged communities and expect an early death have little incentive to avoid crime and violence (e.g., Wilson & Daly, 1997). Recent research by Brezina et al. (2009) suggested that fatalism might facilitate or even encourage subsequent criminal behavior, as youths who discount the future and view death as inevitable are not deterred by the risks associated with a criminal lifestyle. Although there is a growing literature that examines the consequences of early death expectations, little is known about the factors that influence perceived risk for early violent death. Given the range of serious harmful outcomes associated with anticipated early death, researchers have argued that additional research is needed to understand the origins of early death expectations (Brezina et al., 2009; Jamieson & Romer, 2008). The present study contributes to this endeavor by examining the factors that influence perceived risk for early violent death.

The present study draws from two separate literatures to explore possible correlates of perceived risk for early violent death. First, research on perceived risk for victimization emphasizes the role of prior victimization and risky lifestyles/routine activities in shaping perceptions of risk for violent victimization in general (e.g., Melde, 2009; Tillyer, Fisher, & Wilcox, 2011), although there is little research on perceived risk for homicide victimization, specifically.¹ Second, qualitative research conducted in disadvantaged urban communities also provides some insights regarding the factors that might shape one’s perceptions of risk for an early violent death (Anderson, 1999; Brezina et al., 2009). Similar to the literature on perceived risk for violent victimization, this research suggests that both exposure to violence (via personal and vicarious violent victimization experiences) and one’s own delinquent behavior may shape perceived risk for early violent death, although these studies do not systematically examine the relative contributions of various factors to early death expectations. The present study, which uses two waves of data from a nationally representative sample of adolescents, builds on these existing lines of inquiry to examine the influence of various factors—including violent victimization, witnessing violence, and delinquent activities—on perceived risk for early violent death.

Literature Review

Research on the perceived risk for victimization, which is part of the broader fear of crime literature, suggests some possible sources of perceived risk for early violent death. Studies focusing on fear of crime have operationalized the construct in various ways, with measures ranging from feelings of worry, to perceptions of risk, to general statements about feeling safe in one’s neighborhood (LaGrange & Ferraro, 1989). In recent decades, scholars have argued that fear and perceived risk of criminal victimization are two related, yet distinct, concepts. The former represents an emotional state of worry, whereas the latter is a cognitive assessment regarding the perceived likelihood of victimization (see Ferraro & LaGrange, 1987; LaGrange, Ferraro, & Supancic, 1992; Warr, 1984, 2000; Warr & Stafford, 1983). Empirical investigations have confirmed the distinct nature of fear and perceived risk (e.g., LaGrange & Ferraro, 1989; Wilcox Rountree & Land, 1996a). Chiricos, McEntire, and Gertz (2001) argued, “Perceived risk of victimization is arguably a more accessible and unambiguous measure of criminal threat than fear,” as some people (e.g., young adult males) may be more willing and able to assess their chances of victimization than to admit they are fearful (p. 327). As such, the present study focuses on the perceived risk of early violent death, rather than on an emotional state of worry or feelings of safety.

Scholars have drawn from the lifestyle/routine activities literature (Cohen & Felson, 1979; Cohen, Kluegel, & Land, 1981; Hindelang, Gottfredson, & Garofalo, 1978; Miethe, Stafford, & Long, 1987) when studying perceived risk for victimization, arguing that risky lifestyles and routine activities shape not only real opportunities for victimization but also one’s perceptions of risk (e.g., Wilcox Rountree & Land, 1996b). As Wilcox Rountree and Land (1996a) stated, “Risk perception . . . is presumed to be rationally conceived on the part of an individual based on upon such things as vulnerability, perceived seriousness, actual incidence of victimization, or neighborhood/city conditions” (p. 151). Indeed, several studies have identified prior victimization as a correlate of perceived risk for violent victimization (Melde, 2009; Melde & Esbensen, 2009; Otis, 2007; Tillyer et al., 2011; Wilcox, May, & Roberts, 2006). In other words, it appears that individuals use past experiences to make cognitive assessments about the likelihood of future events occurring, suggesting that past violent victimization might increase one’s perceived risk for early violent death.

That being said, a central theme emerging from the literature is the crime-type specific nature of both fear and perceived risk of victimization; prior robbery victimization, for example, is hypothesized to influence perceptions of future risk for robbery victimization (Ferraro & LaGrange, 1987; Fisher & Sloan, 2003; LaGrange & Ferraro, 1989; Wilcox Rountree, 1998). It is unknown, however, whether individuals who experience violent victimization in the form of assaults or robberies will interpret that experience as an indicator of their risk for early violent death.

Researchers report mixed findings for measures of delinquent or criminal behavior thought to tap risky lifestyles or routine activities that increase the perceived risk of violent victimization. For example, Tillyer et al. (2011) found that self-reported criminal behavior was significantly related to perceived risk for violent victimization. In contrast, Melde (2009) found no direct significant relationship between a delinquent lifestyle and perceived risk. Neither of these studies, however, focused on specific forms of crime or examined perceived risk of homicide victimization. Indeed, most empirical studies that examine the sources of perceived risk for victimization focus on property crime (e.g., Wilcox Rountree & Land, 1996a, 1996b) or sexual victimization (e.g., Orchowski, Creech, Reddy, Capezza, & Ratcliff, 2012; Rinehart & Yeater, 2012; Walsh, DiLillo, & Messman-Moore, 2012). Those who do examine perceived risk for violent victimization either do not include homicide (e.g., Melde, 2009; Tillyer et al., 2011), include it as part of a composite measure that captures many forms of violent victimization (e.g., LaGrange et al., 1992), or tap more general perceptions of neighborhood safety rather than personal risk (e.g., Gomme, 1986; Wilcox, Quisenberry, & Jones, 2003). Given the serious and often personal nature of homicide (Katz, 1988),² it is unclear whether the factors that shape its perceived risk differ from those that might influence perceived risk for other forms of victimization.

Consistent with the implications from the perceived risk of victimization literature, qualitative data collected in inner cities also suggest that exposure to violence—both through direct violent victimization and witnessing the victimization of others—may influence early death expectations. Although the primary focus of the aforementioned Brezina et al. (2009) study was to examine how anticipated early death influences youth crime, the authors’ in-depth interviews with active street offenders in Atlanta also revealed potential sources of risk perception for early violent death. Brezina et al. report that young offenders’ fatalism “emanated from their day-to-day exposure to violence” (p. 1113), stating that “predictably, our interviewees described their own lives and neighborhoods as plagued by the persistent threat of violence” (p. 1114). This threat was evidenced by the participants’ own reports of victimization and the victimizations of their friends and family members. Similarly, Anderson’s (1999) ethnographic research in inner city Philadelphia suggests that exposure to violence leads many youths to expect an early death. As he describes, “Those residing in some of the most troubled areas typically have witnessed much street violence that has at times resulted in maiming or death. All of this contributes to the posture that dying ‘ain’t no big deal’” (p. 135).

Anderson’s (1999) work also reveals somewhat contradictory narratives about the perceived role of one’s own behavior as contributing to the risk for early violent death. On one hand, fate plays a central role in how people make sense of their lives:

The belief that whatever one does or says was meant to be allows one to take chances that are not perceived as chances, risks that are not seen as risks, because what will be will be. Hence the person is able to walk the streets almost fearlessly, knowing that “when my time is here, it is here, and there’s nothing I can do about it.” (p. 137)

On the other hand, there seems to be an implicit acknowledgment that one’s own lifestyle and behavior contribute to the risk for early violent death: “There is a strong belief in fate and the notion that a person has a time to be on the planet, but that people can ‘rush’ their time by ‘living fast’ or ‘running in the fast lane’” (Anderson, 1999, pp. 137-138). Although this perceived risk of early death is distinct from perceptions about being apprehended and sanctioned for one’s crimes, it is plausible that the former makes the latter less influential in decisions about criminal behavior (Brezina et al., 2009).

Hoffman’s (2004) qualitative study on young people in Boston and Los Angeles who had been victims and/or perpetrators of youth acquaintance violence suggests that for some, there is a clear association between their own criminal offending and getting shot; many view it as inevitable and even as a rite of passage. Hoffman reports that

the young people in this study accepted the likelihood of death as normal or expected . . . Some might ascribe this acceptance to sub-cultural value system. However, the participants also expressed a realistic assessment of and perhaps necessary accommodation to the dangers and possible costs involved of living a certain lifestyle: the widespread use of guns, the pervasiveness of trauma, the strong likelihood of hospitalization for injury, and the ancillary risks of employment in the drug trade. (p. 62)

The perception that one’s own criminal lifestyle may contribute to the risk of being murdered is consistent with research on the actual correlates of homicide victimization risk. Research suggests, however, that not all types of criminal behavior are risky. For example, Ezell and Tanner-Smith (2009) analyzed longitudinal arrest and mortality data from individuals released from the California Youth Authority to examine how lifestyle and criminal history variables influence homicide victimization risk. Their findings reveal that gang membership and violent arrest history increased homicide victimization risk, whereas other lifestyle and criminal history variables (such as alcohol and drug abuse and total arrests) were not significantly related to homicide victimization risk. In short, it may be that reality affects perceptions: Perceived risk for early violent death may be shaped by the specific criminal behaviors that influence actual risk.

Method

The present study draws on the perceived risk of victimization literature and qualitative studies on the violent street life to develop and test hypotheses about the origins of early violent death expectations. Existing survey research suggests that prior victimization and risky lifestyles/routine activities may influence perceived risk for violent victimization, but no study to date has examined perceived risk for early violent death, specifically. Qualitative research conducted in disadvantaged urban communities describes the potential role of exposure to violence—in the form of direct victimization and/or witnessing violence perpetrated against others—coupled with one’s own delinquent behavior in shaping perceived risk for early violent death. However, no study to date has systematically examined the relative contributions of these factors to perceived risk for early violent death. The present study contributes to these areas of research by testing the following hypotheses:

Hypothesis 1: Adolescents who experienced or witnessed a violent victimization in the past year are more likely to estimate that there is a 50% or greater chance that they will be killed by the age of 21.

Hypothesis 2: Adolescents who report delinquent behavior—including violence, gang membership, property crime, drug dealing, and drug use—in the past year are more likely to estimate that there is a 50% or greater chance that they will be killed by the age of 21.

Data

The present study tests the hypotheses using data collected from a nationally representative sample of students enrolled in American middle and high schools as part of the National Longitudinal Study of Adolescent Health (Add Health; Harris et al., 2009). The first two waves of the public-use Add Health data were used. A sample of 80 American high schools and 52 American middle schools was selected using systematic sampling and stratification techniques (Harris et al., 2009). Students in Grades 7 through 12 completed the in-school Wave 1 self-report survey during the 1994-1995 academic year. A sample was drawn from these students using stratified random sampling, and additional data were collected during in-home interviews. Wave 2 data were collected during in-home interviews conducted during April through August of 1996. In addition, the Wave 2 public-use data include community contextual variables from 1990 Census data that are created from block group levels based on the Wave 2 addresses (Billy, Wenzlow, & Grady, 1998). Wave 1 of the public-use Add Health data contains data about 6,504 adolescents (Harris et al., 2009). Of these, 4,834 participants were re-interviewed during Wave 2. Listwise deletion of cases based on missing data on Wave 1 and Wave 2 variables of interest resulted in 4,349 cases for analysis.³

Measures

The dependent variable of interest, perceived risk for early violent death, was created based on participants’ responses to the following question at Wave 2⁴: “What do you think are the chances that each of the following things will happen to you? You will be killed by age 21.” Possible responses included “almost no chance,” “some chance, but probably not,” “a 50-50 chance,” “a good chance,” and “almost certain.” Similar to other studies on early death expectations (e.g., Brezina et al., 2009; Nguyen, Hussey, et al., 2012; Nguyen, Villaveces, et al., 2012), responses were dichotomized so that the final measure reflects whether or not the respondent reports that there is a 50% or greater chance that he or she will be killed by age 21 (0 = no, 1 = yes). A similar measure of perceived risk for early violent death at Wave 1 was included as a control variable so that the effects of events occurring between Wave 1 and Wave 2 on perceived risk for early violent death at Wave 2 interviews could be isolated.⁵ Approximately 15% of the sample reported in Wave 2 that there is a 50% or greater chance that they will be killed before the age of 21. This is a slight increase from the previous wave (14%). The first four columns of Table 1 report the descriptive statistics for all study variables.

Table 1:

Descriptive Statistics and Bivariate Logistic Regressions of Perceived Risk of Being Killed by Age 21 (W2)

	Minimum	Maximum	M	SD	B	SE	Exp(B)
Individual characteristics
Perceived risk of being killed, W2	0.00	1.00	0.15	0.36
Perceived risk of being killed, W1	0.00	1.00	0.14	0.35	1.70**	0.10	5.50
Violent victimization, W2	0.00	1.00	0.15	0.36	0.65**	0.11	1.91
Witnessed violence, W2	0.00	1.00	0.08	0.27	0.53**	0.14	1.70
Violent offending, W2	0.00	1.00	0.12	0.32	1.06**	0.11	2.88
Gang membership, W2	0.00	1.00	0.04	0.20	1.05**	0.16	2.86
Property crime, W2	0.00	1.00	0.22	0.42	0.05	0.10	—
Drug dealing, W2	0.00	1.00	0.08	0.27	0.64**	0.14	1.89
Recent drug use, W2	0.00	1.00	0.16	0.37	0.23*	0.11	1.25
Male	0.00	1.00	0.50	0.50	0.04	0.09	—
Black	0.00	1.00	0.13	0.34	0.56**	0.11	1.75
Hispanic	0.00	1.00	0.12	0.33	0.22	0.12	—
Age, W2	11.00	19.00	15.87	1.58	0.13**	0.03	1.14
Welfare, W2	0.00	1.00	0.08	0.28	0.53**	0.13	1.70
Deceased parent, W2	0.00	1.00	0.04	0.20	0.72**	0.18	2.05
Depression, W1	0.00	2.63	0.56	0.38	0.79**	0.10	2.19
Impulsivity, W1	1.00	5.00	2.22	0.64	0.04	0.07	—
Block group characteristics
High poverty, W2	0.00	1.00	0.19	0.39	0.40**	0.10	1.49
Low education, W2	0.00	1.00	0.14	0.34	0.40**	0.11	1.49
Modal race Black, W2	0.00	1.00	0.10	0.30	0.53**	0.12	1.71
High proportion Hispanic, W2	0.00	1.00	0.04	0.19	0.05	0.22	—

p < .05. **p < .01.

Two variables were created to measure the respondents’ direct and indirect violent victimization experiences in the past year. Violent victimization is a dichotomous variable that measures whether respondents reported at Wave 2 that any of the following happened in the past 12 months: They had a knife or gun pulled on them, someone shot them, someone stabbed or cut them, and/or they were jumped. Responses were coded so that a 0 reflects no violent victimization in the past 12 months and a 1 reflects one or more violent victimization experiences in the past 12 months. Witnessed violence is a dichotomous variable that measures whether respondents report at Wave 2 that they saw someone shoot or stab another person during the past 12 months (0 = no, 1 = yes).

Several variables were created to measure the respondents’ violent and delinquent activities. Violent offending is a dichotomous variable that measures whether respondents reported at Wave 2 that they had done any of the following in the past 12 months: pulled a knife or gun on someone, shot or stabbed someone, hurt someone badly enough to need bandages or care from a doctor or nurse, and/or used or threatened to use a weapon to get something from someone. Responses were coded so that a 0 reflects no violent offending in the past 12 months and a 1 reflects one or more violent offenses in the past 12 months. Gang membership is a dichotomous variable that measures whether respondents reported at the Wave 2 interview that they have been initiated into a named gang (0 = no, 1 = yes). Property crime is a dichotomous variable that measures whether respondents reported at Wave 2 that they had done any of the following in the past 12 months: taken something from a store without paying for it, stolen something worth less than US$50, stolen something worth more than US$50, and/or went into a house or building to steal something. Responses were coded 0 (no) and 1 (yes). Drug dealing is a dichotomous variable that measures whether the respondents reported at the Wave 2 interview that they had sold marijuana or other drugs in the past 12 months. Recent drug use is a dichotomous variable that measures whether the respondents reported at Wave 2 that they used any illegal drugs (including marijuana, cocaine, inhalants, LSD, PCP, ecstasy, mushrooms, speed, ice, heroin, or pills without a doctor’s prescription) in the past 30 days (0 = no, 1 = yes).

Prior studies suggest that socio-demographic characteristics are associated with fear and/or perceived risk of criminal victimization (see, for example, Ferraro, 1995; LaGrange & Ferraro, 1989; Wilcox Rountree, 1998; Wilcox Rountree & Land, 1996a, 1996b). Therefore, several additional variables were included to control for respondents’ socio-demographic characteristics, including male, Black, and Hispanic (coded 0 = no, 1 = yes). Age measures the age of the respondent, in years, at the Wave 2 interview. Welfare is a dichotomous variable that measures whether the respondent reported at Wave 2 that his or her resident mother or resident father receives public assistance (0 = no, 1 = yes).

Research also suggests that young adults who lose a parent prematurely estimate shorter life spans for themselves relative to others their age (e.g., Denes-Raj & Ehrlichman, 1991). Therefore, the analysis also includes a control for deceased parent, a dichotomous variable that measures whether the respondent reported at Wave 2 that one or both of his or her biological parents were deceased (0 = no, 1 = yes). The analysis also controls for depressive symptoms and impulsivity measured during the Wave 1 interview. Consistent with prior research using the Add Health data (e.g., Bobadilla, Vaske, & Asberg, 2013), depression is measured by taking the mean of 19 items that asked respondents how often each of the following things was true during the past week (Cronbach’s α = .86). Items included “You felt that you could not shake off the blues, even with help from your family and your friends”; “You felt depressed”; “You felt lonely”; “You felt sad”; and so forth. Possible responses ranged from 0 (never or rarely) to 3 (most of the time or all of the time), with higher levels indicating higher levels of depressive symptoms. Similar to prior studies using the Add Health data (e.g., Vazsonyi, Cleveland, & Wiebe, 2006), the present study measures impulsivity by taking the mean of 4 items (Cronbach’s α = .75). Items included “When you have a problem to solve, one of the first things you do is get as many facts about the problem as possible,” “When you are attempting to find a solution to a problem, you usually try to think of as many different ways to approach the problem as possible,” and so forth. Possible responses ranged from 1 (strongly agree) to 5 (strongly disagree), with higher values indicating higher levels of impulsivity.

Finally, prior research has documented the influence of contextual factors on perceived risk for victimization (e.g., Wilcox et al., 2003), as well on the actual risk for homicide (Dobrin, Lee, & Price, 2005). Four variables were included to control for the characteristics of the respondent’s block group at Wave 2.⁶ The Add Health data include a variable that measures the proportion of persons below poverty level, with the pre-constructed categories “low,” “medium,” and “high.” Block groups falling into the “High” category were among the highest 25% in poverty and had a proportion of 23.9% or greater living below the poverty level. High poverty is a dichotomous variable that measures whether the respondent lived in a high poverty block group at Wave 2 (0 = no, 1 = yes). The Add Health data also include a variable that measures the modal educational attainment of individuals 25 years and older, with the pre-constructed categories of “No high school degree or equivalency,” “High school degree, no college degree,” and “College degree or more.” Low education is a dichotomous variable that measures whether the respondent lived in a block group in which “no high school degree or equivalency” was the modal educational attainment (0 = no, 1 = yes). The Add Health includes a measure of block group modal race, with the pre-constructed categories of “White,” “Black,” and “Other.” Modal race Black is a dichotomous variable that measures whether the respondent lived in a block group with a modal race of Black during Wave 2. Finally, the Add Health data include a proportion Hispanic measure, with pre-constructed categories: low = less than 25% of the population was Hispanic, medium = 25% to 49%, high = 50% to 74%, and very high = 75% or more. High proportion Hispanic is a dichotomous variable that measures whether the respondent lived in a block group with a population that was 50% or more Hispanic. Multicollinearity diagnostics did not raise any concerns, with all Variance Inflation Factors less than or equal to 1.99 (please see the appendix for the bivariate correlations among all study variables).

Results

The final three columns of Table 1 present the bivariate logistic regressions of perceived risk of being killed by age 21. As expected, perceived likelihood of early violent death at Wave 1 is significantly and positively associated with perceived likelihood of early violent death at Wave 2. In addition, respondents who experienced violent victimization, witnessed violence, perpetrated violence, had been initiated into a gang, sold drugs, and used illegal drugs recently were significantly more likely to report a 50% or greater chance of being killed by the age of 21 at Wave 2. Black race, age, welfare, deceased parent, and depression were all positively and significantly related to perceived risk for early death at the bivariate level. Finally, respondents who lived in block groups marked by high levels of poverty, low levels of educational attainment, and a high proportion of Black residents were significantly more likely to report a 50% or greater chance of being killed by the age of 21 at Wave 2.

Table 2 reports the results of the multivariate logistic regression analyses. Model 1, which serves as a base model, indicates that age, welfare, having a deceased parent, and depression were all positively and significantly related to perceived risk of being killed by the age of 21. Gender, race, ethnicity, impulsivity, and the block group characteristics were unrelated to the dependent variable in the multivariate analysis.

Table 2:

Multivariate Logistic Regressions of Perceived Risk of Being Killed by Age 21 (W2)

	Model 1			Model 2			Model 3			Model 4
	B	SE	Exp(B)	B	SE	Exp(B)	B	SE	Exp(B)	B	SE	Exp(B)
Individual characteristics
Perceived risk of being killed, W1	1.57**	0.10	4.80	1.55**	0.10	4.73	1.52**	0.10	4.57	1.53**	0.10	4.60
Violent victimization, W2				0.39**	0.13	1.47				0.11	0.14	—
Witnessed violence, W2				0.00	0.16	—				−0.28	0.18	—
Violent offending, W2							0.78**	0.13	2.17	0.79**	0.14	2.21
Gang membership, W2							0.44*	0.20	1.55	0.48*	0.21	1.61
Property crime, W2							−0.21	0.12	—	−0.21	0.12	—
Drug dealing, W2							0.20	0.18	—	0.23	0.18	—
Recent drug use, W2							−0.19	0.13	—	−0.18	0.13	—
Male	0.09	0.09	—	0.02	0.09	—	−0.06	0.10	—	−0.07	0.10	—
Black	0.29	0.16	—	0.29	0.16	—	0.26	0.16	—	0.27	0.16	—
Hispanic	0.20	0.15	—	0.17	0.15	—	0.11	0.16	—	0.12	0.16	—
Age, W2	0.09**	0.03	1.09	0.09**	0.03	1.09	0.09**	0.03	1.09	0.09**	0.03	1.09
Welfare	0.37*	0.15	1.45	0.36*	0.15	1.43	0.33*	0.15	1.39	0.33*	0.15	1.38
Deceased parent, W2	0.56**	0.19	1.76	0.52**	0.19	1.69	0.54**	0.19	1.71	0.52**	0.20	1.69
Depression, W1	0.39**	0.11	1.47	0.34**	0.11	1.41	0.32**	0.12	1.37	0.31**	0.12	1.36
Impulsivity, W1	0.00	0.07	—	0.01	0.07	—	0.02	0.07	—	0.02	0.07	—
Block group characteristics
High poverty, W2	0.00	0.14	—	−0.03	0.14	—	−0.09	0.14	—	−0.09	0.14	—
Low education, W2	0.14	0.16	—	0.15	0.16	—	0.13	0.16	—	0.14	0.16	—
Modal race Black, W2	0.05	0.19	—	0.03	0.19	—	0.07	0.19	—	0.08	0.19	—
High proportion Hispanic, W2	−0.27	0.28	—	−0.25	0.28	—	−0.23	0.29	—	−0.21	0.29	—
Constant	−3.90**	0.51	0.02	−3.90**	0.51	0.02	−3.89	0.52	—	−3.90**	0.52	0.02

p < .05. **p < .01.

Recall that Hypothesis 1 predicted a significant and positive relationship between violent victimization occurring between Waves 1 and 2 and the perceived risk for early violent death reported by the respondent at the Wave 2 interview. It also predicted that witnessing violence would significantly increase perceived risk for early violent death. The results from the multivariate analysis presented in Model 2 reveal partial support for Hypothesis 1. Violent victimization exerted a positive and significant effect, whereas witnessing violence was not significantly related to perceived risk of being killed by the age of 21. The odds of anticipating early death were 1.47 times greater for those who reported violent victimization in the prior year controlling for the other variables in the model.

The results presented in Model 3 indicate mixed support for Hypothesis 2, which predicted a significant positive relationship between delinquent behavior and perceived risk for early violent death. Specifically, violent offending and gang membership were significantly and positively related to perceived risk for early violent death. Controlling for the other variables in the model, the odds of anticipating early death were 2.17 times greater for those who reported violent offending in the previous year, and 1.55 times greater for those who reported gang membership. Property crime, drug dealing, and recent drug use, however, were non-significant in the multivariate analysis.

The full model, including both the victimization and offending variables, is displayed in Model 4. Violent victimization was rendered non-significant, suggesting that the effect of violent victimization on perceived risk for early violent death might be mediated by violent offending and/or gang membership. The effects of violent offending and gang membership remained significant and slightly increased; the odds of anticipating early death were 2.21 times greater for those who reported violent offending, and 1.61 times greater for those who reported gang membership.

Discussion

Prior research on the consequences of early death expectations indicates that those who anticipate dying young are at risk for a range of negative outcomes, leading scholars to call for research on the origins of early death expectations (Brezina et al., 2009; Jamieson & Romer, 2008). The present study contributes to this area of research by examining the sources of perceived risk for early violent death specifically, drawing from the perceived risk of victimization literature and qualitative research conducted in inner cities to develop hypotheses. Data from a nationally representative sample of adolescents were used to examine the influence of prior victimization, witnessing violence, and delinquency on the perceived risk of being killed before the age of 21.

Studies on perceived risk for non-fatal violent victimization have generally demonstrated a significant relationship between prior victimization and perceived risk (e.g., Melde, 2009; Melde & Esbensen, 2009; Otis, 2007; Tillyer et al., 2011; Wilcox et al., 2006). Findings from the present study, however, indicate that the effect of violent victimization on the perceived risk for being killed by age 21 was rendered non-significant once the delinquency variables were introduced into the model. The lack of a direct effect may be due to the unique nature of homicide and the assumption that there are distinct motives underlying various types of violent victimization (Katz, 1988). For example, personal property may be viewed as the actual target in a robbery, whereas an individual is the target of a homicide; adolescents, therefore, may not view past robbery victimization as an indicator of risk for homicide victimization.

These findings suggest that recent prior victimization may indirectly influence perceived risk via recent violent offending and gang membership. Perhaps surviving violent victimization leads to feelings of invincibility, anger, and/or recklessness, thus emboldening individuals to engage in delinquency, which in turn influences perceptions of risk. That being said, the well-documented relationship between victimization and offending is likely more complex than that which can be gleaned from the present study (for reviews, see Jennings, Piquero, & Reingle, 2012; Lauritsen & Laub, 2007). There are three broad explanations that may contribute to the victim–offender overlap: (a) Individuals respond to victimization by engaging in offending, (b) a delinquent lifestyle exposes one of opportunities for victimization, and (c) common traits and/or routine activities influence both offending and victimization (e.g., Cohen et al., 1981; Gottfredson & Hirschi, 1990; Jensen & Brownfield, 1986; Lauritsen, Sampson, & Laub, 1991; Mustaine & Tewksbury, 2000; Schreck, 1999). Unpacking the sources of the victim–offender overlap is beyond the scope of the present study, which was designed to explore the effects of recent victimization and offending on perceived risk for early violent death. Future research should examine the long-term effects of victimization and offending to see how such experiences might influence one another and shape perceptions of risk over time.

There was no significant relationship between witnessing serious violence and the perceived risk for early violent death. Again, perceptions of homicide risk may be individualized. Perhaps witnessing the stabbing or shooting of another person does not influence one’s perceptions of personal risk because individuals assess risk based on specific motives and threats related to interpersonal dynamics, rather than a general state of danger assumed to uniformly influence risk for all potential victims.

In addition, although this research was informed by the perceived risk for victimization literature, the purpose of the present study was to examine the influences of perceived risk for early violent death given its influence on subsequent criminal behavior (Brezina et al., 2009). The wording of the dependent variable (“What do you think are the chances that each of the following things will happen to you? You will be killed by age 21”) does not explicitly describe homicide victimization. It is possible that experiencing or witnessing violent victimization increases perceived risk for homicide victimization, specifically. Additional studies using alternative data sources are needed to determine whether the sources for perceived risk of homicide victimization are similar to those for perceived risk for early violent death.

The present study did find support for the influence of delinquent behavior on perceived risk of early violent death, although the results indicate that not all forms of delinquency are associated with perceived risk. Those who reported violent behavior and gang membership were significantly more likely to anticipate an early violent death, whereas property crime, drug dealing, and recent drug use were not significantly related to perceived risk. These findings are consistent with research by Ezell and Tanner-Smith (2009) on the actual correlates of homicide victimization risk, suggesting that adolescents’ perceptions of risk factors may be fairly accurate.

It is unclear, however, whether adolescents consciously consider their violent behavior and gang membership when assessing risk for violent death, or whether the association is less intentional. On one hand, adolescents may accurately recognize the risks associated with violent behavior and gangs, a recognition that may come from observing the early violent deaths of others who have lived similar lifestyles (see Hoffman, 2004). On the other hand, adolescents might not willfully acknowledge the specific lethal risk of a violent lifestyle, but it may expose them to people and/or situations that shape general perceptions of danger. As reviewed above, Anderson’s (1999) work revealed a set of contradictory narratives about the perceived role of one’s own risky behavior in the risk of early violent death. Although fate is perceived to determine the time of one’s death, it is believed that individuals can “rush” their time. The data analyzed herein were not able to detect the extent to which these narratives contribute to individual assessments of early violent death risk. Future research should examine such possibilities.

The findings from the present study, coupled with prior research on the negative consequences of early death expectations, suggest that serious delinquency is positively associated with subsequent perceived risk for early violent death, which in turn is related to a range of negative outcomes in adulthood, including suicide, substance use, unsafe sexual activity, HIV, dropping out of school, and arrest (Borowsky et al., 2009; Fischhoff et al., 2000; Jamieson & Romer, 2008; Nguyen, Hussey, et al., 2012; Nguyen, Villaveces, et al., 2012). This is consistent with the ideas of cumulative continuity and state dependence described by Laub and Sampson (1993), who argued,

Delinquency incrementally mortgages the future by generating negative consequences for the life chances of stigmatized and institutionalized youths. . . . The cumulative continuity of disadvantage is thus not only a result of stable individual differences in criminal propensity, but a dynamic process whereby childhood antisocial behavior and adolescent delinquency foster adult crime through the severance of adult social bonds. (p. 306)

This cumulative continuity is often traced to social reactions to the delinquency, by which labeling processes stigmatize the individual (e.g., Sampson & Laub, 1992), thus weakening social bonds. The findings from the present study suggest that there might be an additional way in which serious delinquency severs social bonds, thus explaining observed continuity across the life course. Perhaps early death expectations that are the consequence of serious delinquency lead individuals to discount their future, leaving them little incentive to invest in social bonds (Brezina et al., 2009). In other words, social bond formation might be undermined by both the reaction of society to the delinquency and the individual’s own calculation about his or her prospects for a future. Future research should explore this possibility by examining whether perceived risk for early violent death influences social bonds and mediates the relationship between adolescent and adult criminal behavior.

If perceived risk for early violent death is both a consequence and cause of serious offending, this has implications for strategies aimed at controlling crime and reducing recidivism. Deterrence scholars, for example, have begun to focus on the varying effects of sanctions across individuals and situations (see Jacobs, 2010; Piquero, Paternoster, Pogarsky, & Loughran, 2011). As Brezina et al. (2009) have argued, deterrence-based strategies that only aim to enhance the costs of criminal behavior are unlikely to have a significant impact on adolescents who anticipate an early death and as a result, discount the future. Rather, interventions that address fatalistic thinking will be needed to disrupt the reciprocal relationship between serious offending and perceived risk for early violent death that may help account for continuity in offending across the life course. For example, Duke, Skay, Pettingell, and Borowsky (2011) reported that nurturing positive connections with prosocial adults can change youths’ early death perceptions, as can strategies that enhance self-worth and confidence. Researchers have argued that schools and health care providers should screen adolescents for perceived risk for early death and refer them to the appropriate services (Borowsky et al., 2009; Jamieson & Romer, 2008); the present study’s findings suggest that those working in correctional settings should also perform this function. In addition, Topalli, Brezina, and Bernhardt (2013) argued that faith-based programs need to be aware of the unexpected ways religion may be used by some offenders. Their interviews with active street offenders revealed that rather than deterring offenders, religious beliefs were used to neutralize fears about anticipated early death and facilitate ongoing criminal behavior based on assumptions that all crimes would be forgiven in the afterlife.

Understanding the role perceived risk for early violent death plays in maintaining a violent lifestyle may also prove important for understanding and enhancing the effects of deterrence-based initiatives that go beyond increasing sanction severity. For example, focused deterrence strategies (i.e., “pulling levers”) are typically comprised of law enforcement, social service, and community partners who create a clear message of violence rejection and directly communicate it to individuals associated with violent groups or gangs; the message includes predictable and meaningful consequences for engaging in violence, as well as various social services for those who want them (e.g., Engel, Tillyer, & Corsaro, 2013). A recent meta-analysis indicates that focused deterrence strategies are significantly related to reductions in violence (Braga & Weisburd, 2012). Scholars have noted that these strategies implicate several mechanisms in addition to deterrence, including collective efficacy, correctional rehabilitation, procedural justice and legitimacy, situational discouragement, and informal social control (e.g., Braga & Weisburd, 2012; Engel et al., 2013; Tillyer, Engel, & Lovins, 2012). Researchers have struggled to measure and isolate the effects of various elements of the program; in particular, those elements falling under the broad umbrella of “community” (see Engel et al., 2013). As Engel et al. described (p. 407), such activities aim to “create a ‘moral voice’ of the community by delivering a clear message of non-violence and rejecting the norms and narratives of the street which promote violence,” including, in some cases, the hopelessness and the inevitability of an early death described by Anderson (1999). Most commentary surrounding these initiatives emphasizes the deterrence potential of law enforcement and, to a lesser degree, the rehabilitative potential of various social services. Challenging fatalistic norms, however, might be necessary for either deterrence or rehabilitation to work; both presumably require an expectation of one’s future survival. Future focused deterrence research should examine how directly addressing fatalistic beliefs influences program outcomes.

Although not the focus of the present study, the results indicate there was no relationship between block group characteristics and perceived risk for early violent death. There are several possible explanations for this. As the aforementioned findings suggest, it might be that individuals perceive homicide victimization to be personal, and thus, block group characteristics do little to influence assessments of individual risk. It is also possible that the measures included in this study do not adequately tap the real and/or perceived neighborhood contexts that shape perceptions of risk. Prior studies on other crime types suggest that perceptions of community conditions, such as social and physical disorder, are important for shaping both perceived risk and fear of crime (e.g., Chiricos et al., 2001; Hinkle & Weisburd, 2008; LaGrange et al., 1992; Lee & Earnest, 2003; Melde, 2009; Randa & Wilcox, 2010; Scarborough, Like-Haislip, Novak, Lucas, & Alarid, 2010; Swatt, Varano, Uchida, & Solomon, 2013). However, most of these studies do not include measures of respondent delinquency (see Melde, 2009, for an exception), and none examine perceived risk for early violent death, specifically. It is unclear whether the effects of perceived neighborhood conditions would be observed once various forms of delinquency are controlled for, and whether they are related to perceived risk of early violent death, specifically. Unfortunately, the present study was unable to examine the influence of perceived neighborhood conditions because the Add Health data do not include such measures. Future research should examine the relative contributions of perceived neighborhood conditions, individual delinquency, and prior victimization in shaping perceptions of risk.

In addition, the findings from the present study highlight areas for future research. For example, the present study found a relationship between gang membership and perceived risk, suggesting that delinquent peers may expose one to fatalistic beliefs. However, a recent study by Haynie, Soller, and Williams (2014) indicated that although the fatalism of one’s friends is significantly associated with violent delinquency, the relationship is not mediated by one’s own fatalism. In short, more research is needed to determine how delinquent groups shape behaviors and beliefs. In addition, the present study found that depression and parental death were significantly associated with perceived risk for early violent death; future studies should more fully explore how mental illness and loss contribute to perceived risk for early violent death.

Conclusion

The present study demonstrates that violence and gang membership influence perceived risk for early violent death, whereas prior violent victimization appeared to be indirectly related. Witnessing violence and other forms of delinquency were not significantly related to early violent death expectations, suggesting that risk may be perceived to be highly individualized. Future research should examine whether anticipated early death weakens social bonds, thus helping to explain continuity in antisocial behavior across the life course. In addition, research is needed to determine whether the effects of certain crime control strategies can be enhanced by directly addressing perceptions of early death risk.

Footnotes

Appendix

Authors’ note:

This research uses data from Add Health, a program project directed by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill, and funded by Grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 23 other federal agencies and foundations. Special acknowledgment is due to Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design. Information on how to obtain the Add Health data files is available on the Add Health website (). No direct support was received from Grant P01-HD31921 for this analysis.

Notes

Marie Skubak Tillyer is an associate professor in the Department of Criminal Justice at the University of Texas at San Antonio. Her research interests include violence, victimization, and crime prevention. Her recent research has appeared in Crime and Delinquency, Criminal Justice and Behavior, Justice Quarterly, and Journal of Research in Crime and Delinquency.

References

Anderson

(1999). Code of the street: Decency, violence, and the moral life of the inner city. New York, NY: W.W. Norton.

Billy

J. O. G.

Wenzlow

A. T.

Grady

W. R.

(1998). National Longitudinal Study of Adolescent Health: Public Use Contextual Database. Carolina Population Center, University of North Carolina at Chapel Hill.

Bobadilla

Vaske

Asberg

(2013). Dopamine receptor (D4) polymorphism is related to comorbidity between marijuana abuse and depression. Addictive Behaviors, 38, 2555-2562.

Borowsky

I. W.

Ireland

Resnick

M. D.

(2009). Health status and behavioral outcomes for youth who anticipate a high likelihood of early death. Pediatrics, 124, 81-88.

Braga

A. A.

Weisburd

D. L.

(2012). The effects of focused deterrence strategies on crime: A systematic review and meta-analysis of the empirical evidence. Journal of Research in Crime & Delinquency, 49, 323-358.

Brezina

Tekin

Topalli

(2009). “Might not be a tomorrow”: A multi-methods approach to anticipated early death and youth crime. Criminology, 47, 1091-1129.

Chiricos

McEntire

Gertz

(2001). Perceived racial and ethnic composition of neighborhood and perceived risk of crime. Social Problems, 48, 322-340.

Cohen

L. E.

Felson

(1979). Social change and crime rate trends: A routine activities approach. American Sociological Review, 44, 588-608.

Cohen

L. E.

Kluegel

J. R.

Land

K. C.

(1981). Social inequality and predatory criminal victimization: An exposition and test of a formal theory. American Sociological Review, 46, 505-524.

10.

Denes-Raj

Ehrlichman

(1991). Effects of premature parental death on subjective life expectancy, death anxiety, and health behavior. Journal of Death and Dying, 23, 309-321.

11.

Dobrin

Lee

Price

(2005). Neighborhood structure differences between homicide victims and non-victims. Journal of Criminal Justice, 33, 137-143.

12.

Duke

N. N.

Skay

C. L.

Pettingell

S. L.

Borowsky

I. W.

(2011). Early death perception in adolescence: Identifying factors associated with change from pessimism to optimism about life expectancy. Clinical Pediatrics, 50, 21-28.

13.

Engel

R. S.

Tillyer

M. S.

Corsaro

(2013). Reducing gang violence using focused deterrence: Evaluating the Cincinnati Initiative to Reduce Violence (CIRV). Justice Quarterly, 3, 403-439.

14.

Ezell

M. E.

Tanner-Smith

E. E.

(2009). Examining the role of lifestyle and criminal history variables on the risk of homicide victimization. Homicide Studies, 13, 144-173.

15.

Ferraro

K. F.

(1995). Fear of crime: Interpreting victimization risk. Albany: State University of New York Press.

16.

Ferraro

K. F.

LaGrange

(1987). The measurement of fear of crime. Sociological Inquiry, 57, 70-101.

17.

Fischhoff

Parker

A. M.

Bruin

W. B.

Downs

Palmgren

Dawes

Manski

C. F.

(2000). Teen expectations for significant life events. The Public Opinion Quarterly, 64, 189-205.

18.

Fisher

B. S.

Sloan

J. J.

(2003). Unraveling the fear of victimization among college women: Is the “shadow of sexual assault hypothesis” supported? Justice Quarterly, 20, 633-659.

19.

Gomme

I. M.

(1986). Fear of crime among Canadians: A multi-variate analysis. Journal of Criminal Justice, 14, 249-258.

20.

Gottfredson

Hirschi

(1990). A general theory of crime. Stanford, CA: Stanford University Press.

21.

Harris

K. M.

Halpern

C. T.

Whitsel

Hussey

Tabor

Entzel

Udry

J. R.

(2009). The National Longitudinal Study of Adolescent Health: Research design. Retrieved from http://www.cpc.unc.edu/projects/addhealth/design

22.

Haynie

D. L.

Soller

Williams

(2014). Anticipating early fatality: Friends’, schoolmates’ and individual perceptions of fatality on adolescent risk behaviors. Journal of Youth and Adolescence, 43, 175-192.

23.

Hill

E. M.

Ross

L. T.

Low

B. S.

(1997). The role of future unpredictability in human risk-taking. Human Nature, 8, 287-325.

24.

Hindelang

M. J.

Gottfredson

M. R.

Garofalo

(1978). Victims of personal crime: An empirical foundation for a theory of personal victimization. Cambridge, MA: Ballinger Publishing.

25.

Hinkle

J. C.

Weisburd

(2008). The irony of broke windows policing: A micro-place study of the relationship between disorder, focused police crackdowns and fear of crime. Journal of Criminal Justice, 36, 503-512.

26.

Hoffman

J. S.

(2004). Youth violence, resilience, and rehabilitation. New York, NY: LFB Scholarly Publishing LLC.

27.

Jacobs

B. A.

(2010). Deterrence and deterrability. Criminology, 48, 417-441.

28.

Jamieson

P. E.

Romer

(2008). Unrealistic fatalism in U.S. youth ages 14 to 22: Prevalence and characteristics. Journal of Adolescent Health, 42, 154-160.

29.

Jennings

W. G.

Piquero

A. R.

Reingle

J. M.

(2012). On the overlap between victimization and offending: A review of the literature. Aggression and Violent Behavior, 17, 16-26.

30.

Jensen

G. F.

Brownfield

(1986). Gender, lifestyles, and victimization: Beyond routine activity. Violence and Victims, 1, 85-99.

31.

Katz

(1988). Seductions of crime: Moral and sexual attractions in doing evil. New York, NY: Basic Books.

32.

LaGrange

R. L.

Ferraro

K. F.

(1989). Assessing age and gender differences in perceived risk and fear of crime. Criminology, 27, 697-719.

33.

LaGrange

R. L.

Ferraro

K. F.

Supancic

(1992). Perceived risk and fear of crime: Role of social and physical incivilities. Journal of Research in Crime & Delinquency, 29, 311-334.

34.

Laub

J. H.

Sampson

R. J.

(1993). Turning points in the life course: Why change matters to the study of crime. Criminology, 31, 301-325.

35.

Lauritsen

J. L.

Laub

J. H.

(2007). Understanding the link between victimization and offending: New reflections on an old idea. Crime Prevention Studies, 22, 55-75.

36.

Lauritsen

J. L.

Sampson

R. J.

Laub

J. H.

(1991). The link between offending and victimization among adolescents. Criminology, 29, 265-292.

37.

Lee

M. R.

Earnest

T. L.

(2003). Perceived community cohesion and perceived risk of victimization: A cross-national analysis. Justice Quarterly, 20, 131-157.

38.

Melde

(2009). Lifestyle, rational choice, and adolescent fear: A test of a risk-assessment framework. Criminology, 47, 781-812.

39.

Melde

Esbensen

(2009). The victim–offender overlap and fear of in-school victimization. Crime & Delinquency, 55, 499-525.

40.

Miethe

T. D.

Stafford

M. C.

Long

J. S.

(1987). Social differentiation in criminal victimization: A test of routine activities/lifestyle theories. American Sociological Review, 52, 184-194.

41.

Mustaine

E. E.

Tewksbury

(2000). Comparing the lifestyles of victims, offenders, and victim–offenders: A routine activity theory assessment of similarities and differences for criminal participants. Sociological Focus, 33, 339-362.

42.

Nagin

D. S.

Pogarsky

(2004). Time and punishment: Delayed consequences and criminal behavior. Journal of Quantitative Criminology, 20, 295-317.

43.

Nguyen

Q. C.

Hussey

J. M.

Halpern

C. T.

Villaveces

Marshall

S. W.

Siddiqi

Poole

(2012). Adolescent expectations of early death predict young adult socioeconomic status. Social Science & Medicine, 74, 1452-1460.

44.

Nguyen

Q. C.

Villaveces

Marshall

S. W.

Hussey

J. M.

Halpern

C. T.

Poole

(2012). Adolescent expectations of early death predict adult risk behaviors. PLoS ONE, 7(8), e41905. doi:10.1371/journal.pone.0041905

45.

Orchowski

L. M.

Creech

S. K.

Reddy

M. K.

Capezza

N. M.

Ratcliff

(2012). College women’s perceived risk to experience sexual victimization: A prospective analysis. Violence and Victims, 27, 194-214.

46.

Otis

M. D.

(2007). Perceptions of victimization risk and fear of crime among lesbians and gay men. Journal of Interpersonal Violence, 22, 198-217.

47.

Piquero

A. R.

Paternoster

Pogarsky

Loughran

(2011). Elaborating the individual difference component in deterrence theory. Annual Review of Law and Social Science, 7, 335-360.

48.

Randa

Wilcox

(2010). School disorder, victimization, and general v. place-specific student avoidance. Journal of Criminal Justice, 38, 854-861.

49.

Rinehart

J. K.

Yeater

E. A.

(2012). The effects of male attractiveness and sexual attitudes on women’s risk perception. Violence Against Women, 18, 249-263.

50.

Sampson

R. J.

Laub

J. H.

(1992). Crime and deviance in the life course. Annual Review of Sociology, 18, 63-84.

51.

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, 819-826.

52.

Schreck

C. J.

(1999). Criminal victimization and low self-control: An extension and test of a general theory of crime. Justice Quarterly, 16, 633-654.

53.

Swatt

M. L.

Varano

S. P.

Uchida

C. D.

Solomon

S. E.

(2013). Fear of crime, incivilities, and collective efficacy in four Miami neighborhoods. Journal of Criminal Justice, 41, 1-11.

54.

Tillyer

M. S.

Engel

R. S.

Lovins

(2012). Beyond Boston: Applying theory to understand and address sustainability issues in focused deterrence initiatives to reduce violence. Crime & Delinquency, 58, 973-997.

55.

Tillyer

M. S.

Fisher

B. S.

Wilcox

(2011). The effects of school crime prevention on students’ violent victimization, risk perception, and fear of crime: A multilevel opportunity perspective. Justice Quarterly, 28, 249-277.

56.

Topalli

Brezina

Bernhardt

(2013). With God on my side: The paradoxical relationship between religious belief and criminality among hardcore street offenders. Theoretical Criminology, 17, 49-69.

57.

Vazsonyi

A. T.

Cleveland

H. H.

Wiebe

R. P.

(2006). Does the effect of impulsivity on delinquency vary by level of neighborhood disadvantage? Criminal Justice and Behavior, 33, 511-541.

58.

Walsh

DiLillo

Messman-Moore

T. L.

(2012). Lifetime sexual victimization and poor risk perception: Does emotion dysregulation account for the links? Journal of Interpersonal Violence, 27, 3054-3071.

59.

Warr

(1984). Fear of victimization: Why are women and the elderly more fearful? Social Science Quarterly, 65, 681-702.

60.

Warr

(2000). Fear of crime in the United States: Avenues for research and policy. In Duffee

(Ed.), Measurement and analysis of crime: Criminal justice 2000 (Vol. 4, pp. 451-489). Washington, DC: Office of Justice Programs, U.S. Department of Justice.

61.

Warr

Stafford

(1983). Fear of victimization: A look at the proximate causes. Social Forces, 61, 1033-1043.

62.

Wilcox

May

D. C.

Roberts

S. D.

(2006). Student weapon possession and the “fear and victimization hypothesis”: Unraveling the temporal order. Justice Quarterly, 23, 502-529.

63.

Wilcox

Quisenberry

Jones

(2003). The built environment and community crime risk interpretation. Journal of Research in Crime & Delinquency, 40, 322-345.

64.

Wilcox Rountree

(1998). A reexamination of the crime–fear linkage. Journal of Research in Crime & Delinquency, 35, 341-372.

65.

Wilcox Rountree

Land

K. C.

(1996a). Burglary victimization, perceptions of crime risk, and routine activities: A multilevel analysis across Seattle neighborhoods and census tracts. Journal of Research in Crime & Delinquency, 33, 147-180.

66.

Wilcox Rountree

Land

K. C.

(1996b). Perceived risk versus fear of crime: Empirical evidence of conceptually distinct reactions in survey data. Social Forces, 74, 1353-1376.

67.

Wilson

Daly

(1997). Life expectancy, economic inequality, homicide, and reproductive timing in Chicago neighborhoods. British Medical Journal, 314, 1271-1274.