Racial Disparities in Age at Time of Homicide Victimization

Abstract

This study sought the factors associated with race/ethnicity disparities in the age at which homicide deaths tend to occur. We used the multiple disadvantage model to take race into account as we evaluated associations between age at time of homicide victimization and several social structural, mental health−related, and lifestyle factors. Data were derived from the 1993 National Mortality Followback Survey, a cross-sectional interview study of spouses, next of kin, other relatives, and close friends of individuals 15 years and older who died in the United States in 1993. Our results showed age at time of homicide mortality to be related to the three types of factors; race moderated some of these relationships. In general, being employed, married, and a homeowner appeared associated with reduced victimization while young. The relationship of victimization age and employment was not uniform across racial groups, nor was the relationship of victimization age and marital status uniform across groups. Among Blacks, using mental health services was associated with longer life. Homicide by firearm proved important for our Black and Hispanic subsamples, while among Whites, alcohol’s involvement in homicide exerted significant effects. Our results suggest that programs and policies serving the various racial/ethnic groups can alleviate multiple disadvantages relevant in homicide victimization at an early age.

Keywords

multiple disadvantage model homicide victimization age racial disparities

Introduction

As the nation’s 15th leading cause of death, homicide shortens the life expectancy of Americans, especially young, male Americans from racial minorities (Kochanek, Xu, Murphy, Minino, & Kung, 2011). Between 1991 and 2007, homicide remained among the top four causes of death among Americans 1 to 40 years old, and it claimed a disproportionately high number of Black males 15 to 34 years old (Logan, Smith, & Stevens, 2011). During 1984 to 1991, the murder rate for young Black male victims almost doubled (Levitt, 2004). A drop in homicides began in the mid-1990s due to declining popularity of crack-cocaine, solidification of drug-trafficking “markets,” selective incapacitation, and increased police presence. Despite that drop, empirical evidence indicates ongoing Black–White disparity in the rate of homicide victimization (Levitt, 2004). During 1990 to 1999, for example, Blacks consistently bore the higher rate of homicide victimization and of offending, as compared with Whites (Grogger & Willis, 2000). Among Blacks in 2009, the age-adjusted homicide victimization rate was 26.6% higher than that of Whites (Kochanek et al., 2011; Logan et al., 2011).

As the World Health Organization (Krug, Dahlberg, Mercy, Zwi, & Lozano, 2002) has widely reported, when interpersonal violence results in large numbers of deaths, the matter is one of public health and demands attention from researchers, policymakers, and the public alike. To minimize premature deaths from interpersonal violence, researchers have worked to identify homicide’s risk and protective factors. They have found that in the United States, homicide victims (just like homicide perpetrators) often are male, young, and socially disadvantaged, having low socioeconomic status (SES) and belonging to racial/ethnic minorities (Piquero, MacDonald, Dobrin, Daigle, & Cullen, 2005; South & Messner, 2000). Consistently, studies have shown that the association of (relative) youth with homicide victimization and with greater truncation of life expectancy is stronger among Blacks than majority Whites (Dobrin, Lee, & Price, 2005; Felson & Painter-Davis, 2012). In general, though, the literature lacks studies explicitly designed to explain persistent racial disparity in homicide victimization (Chauhan et al., 2011; Krug et al., 2002).

In addition, the variable age is highly relevant to social disadvantage and, thus, should be considered during efforts to explain racial disparity in homicide victimization. National data on homicide victimization have consistently found the young to be likely victims, with recent data showing highest likelihood of victimization among 15- to 24-year-olds (Lo, Howell, & Cheng, 2013; National Center for Health Statistics, 2012). Homicide mortality can be viewed as a health issue as well as a crime issue, signifying both preventable deaths and truncated life spans (Lo et al., 2013; South & Messner, 2000). As with deaths from preventable illnesses—diabetes, lung cancer—the lower the age at demise, the more truncated the life expectancy, and the more serious the homicide victimization. It is important to understand exactly how individuals become homicide victims. Attending to the concept of age at time of victimization promises to add a dimension of insight into that understanding: seriousness of premature death by homicide and the factors explaining various levels of seriousness (Lo et al., 2013).

Approaching homicide victimization as both a public health problem and the consequence of criminal violence, the present study provided a partial test of the multiple disadvantage model (Lo et al., 2013) as it explains disjuncture between the homicide mortality of Whites and that of Blacks and Hispanics (the principal U.S. minority groups). Including Hispanics—a rapidly growing U.S. minority—in the present study was warranted because their rates of violent crime and of homicide victimization tend to exceed Whites’, but fall short of Blacks’ (Lo et al., 2013; Steffensmeier, Feldmeyer, Harris, & Ulmer, 2011). The multiple disadvantage model attributes homicide victimization to disadvantaging structural factors that constrain individuals’ health and lifestyles across the life span, putting them at increased risk for victimization. Important structural conditions, such as low SES and residence in crime-ridden neighborhoods, are associated with substance use and gun carrying for protection, leaving individuals vulnerable to homicide victimization (Allen & Lo, 2012; Lo & Cheng, 2012). Of major racial/ethnic groups, Blacks and Hispanics are more likely than Whites to be poor and to occupy isolated neighborhoods where job opportunities for adults and youth are limited (Anderson, 1999; Lo & Cheng, 2012; Wilson, 1987). Unsurprisingly, such economic and resource deprivation increases youth gang involvement, along with exposure to violence, that can eventually produce victimization (Piquero et al., 2005).The multiple disadvantage model assigns social and economic inequality a significant role in racial differences in preventable deaths, including homicides (Phelan & Link, 2005; Phelan, Link, Diez-Roux, Kawachi, & Levin, 2004). Research evidence confirms that SES (often measured by income and educational attainment) is the irreducible reason that racial disparity in health and health-related behaviors persists over time (Link & Phelan, 1995). Phelan and Link (2005) studied rates of mortality from preventable diseases over time, attributing decreases in rates to improved medical capacity benefitting both Whites and Blacks. But White–Black disparities in mortality nevertheless persist due to the groups’ differential access to means (knowledge, resources, power). The current American racial majority, Whites, tends to occupy a higher SES echelon, on average, than Blacks and Hispanics. As a result, compared with Blacks, Whites tend to have greater means of reducing risk of victimization and prolonging the life span, and racial disparity characterizes homicide victimization (Gjelsvik, Zierler, & Blume, 2004; Phelan et al., 2004).

Fully understanding racial disparity in age at time of homicide victimization requires investigating social mechanisms that underlie it. Experiencing persistent social disadvantages constitutes chronic stress that increases emotional problems and mental health disorders (Pearlin, 1989). Moreover, mental health problems deriving from social disadvantages that characterize many minority neighborhoods may go untreated due to barriers to high-quality treatment commonly facing minority populations; such problems then worsen (Lo & Cheng, 2011; Lo, Cheng, & Howell, 2014). Minority Americans occupy social locations that foster both emotional difficulty and negative lifestyle choices that facilitate victimization (Aneshensel, 1992; Cockerham, 2005). In the literature, alcohol and gang activity are associated with homicide, whereby these factors increase the risk of homicide victimization across ethnic groups (Ezell & Tanner-Smith, 2009; Kuhns, Wilson, Clodfelter, Maguire, & Ainsworth, 2011). The multiple disadvantage model argues, furthermore, that disadvantaged social locations, emotional difficulty, and negative lifestyle factors wield stronger effects on homicide victimization among members of racial minorities than of the majority group.

Applying the multiple disadvantage model, the present study asked whether and how age at time of homicide victimization is explained—for Whites, for Blacks, and for Hispanics—by social structural factors, emotional difficulty, and lifestyle factors. It also sought to determine whether the strength of the relationships between these variables and age at homicide victimization differed by race. We expected that knowledge the study generated could prove helpful in prolonging life expectancy of Black, Hispanic, and White Americans.

Methods

Data for our study came from the 1993 National Mortality Followback Survey (NMFS), which sampled and collected retrospective, cross-sectional interview data from spouses, next of kin, other relatives, and close friends of individuals 15 years and older who died in the United States in 1993. Interviewees were asked about a decedent’s access to and use of medical services in the year preceding death; disabilities a decedent had in that year; and mortality risk factors describing a decedent in that year. In addition, NMFS collected demographic data from 49 state vital registration agencies (excluding South Dakota) and from analogous agencies serving the District of Columbia and New York City. Death certificates obtained from these agencies documented cause of death (including homicide), age at time of death, and gender. The 1993 NMFS oversampled women, Blacks, and people younger than 35. The present study’s final analyses included data for 1,409 homicide victims (383 non-Hispanic Whites, 765 non-Hispanic Blacks, and 261 Hispanics).

Measures

Our outcome variable, age at time of homicide victimization (or homicide age), was a continuous measure that ranged from 15 to 90 years. We used a logarithmical transformation of homicide age to reduce skewness. Our final analyses’ independent variables comprised social structural, mental health−related, and lifestyle variables closely aligned with the elaboration of the multiple disadvantage model.

Social structural variables included U.S. born, male gender, married, received Aid to Families With Dependent Children (AFDC), employed, homeowner, and at least some college education; they were measured dichotomously, 1 indicating the factor’s applicability to the decedent and 0 (the reference group) indicating its inapplicability. We also used one dichotomously measured mental health−related variable, received mental health care, for which 1 indicated that, in the year preceding death, the decedent received care of any kind as an overnight patient in any alcohol- or drug-treatment facility, psychiatric facility, or mental health facility, and 0 provided the reference.

The four dichotomously measured lifestyle variables used were non-accidental firearm discharge involved in death, alcohol-related death, prescription drug misused, and street drug used. The firearm discharge variable was a proxy, with 1 indicating decedent’s exposure to firearms and 0 providing the reference. Alcohol-related death measured whether respondents (decedent’s relatives and close friends) perceived that alcohol was somehow involved in the death (1 indicating yes, 0 indicating no). Prescription drug misused measured the same respondents’ perceptions that a decedent had—either without being medically prescribed the substance, or well exceeding a prescribed amount of it—consumed pain killers, sedatives, tranquilizers, antidepressants, steroids, or methadone (1 equal to yes, 0 to no). Street drug used measured respondents’ perceptions that a decedent had used heroin, stimulants, marijuana, cocaine, and/or hallucinogens in the year preceding death (1 equal to yes, 0 to no). As the lifestyle variable, street drug, could have been strongly related to our received mental health care variable (with its category for substance treatment), we performed correlation analysis for the two, finding only slight correlation (r = .2, p < .01).

Our choice of these 12 independent variables was guided by NMFS’s data collection process, though the availability of measures suiting an application of the multiple disadvantage model was clearly also important. As our respondents were individuals familiar with a decedent, but not in fact the decedent, we chose behavioral variables that could be readily observed. For each independent measure, however, substantial data were missing. Indeed, opting not to properly recode missing data would have meant using fewer than 20% of the cases to conduct our data analysis. So we chose to recode missing data as the reference category, assuming that in not reporting a readily observable behavior or condition (e.g., employed, married), a respondent was indicating the absence of that behavior or condition. Again, this study’s independent variables were chosen with an eye to their objective (not opinion-based) quality, meaning respondents should have tended to report a measured behavior they had observed.

Data Analysis

Our final analyses (discrete for Blacks, for Hispanics, and for Whites) of relationships between the independent factors and age at time of homicide victimization excluded all NMFS data involving deaths not due to homicide. Our systematic deletion of data from our outcome equation allowed a potential selection bias problem to emerge. We used a two-stage process to address this threat to our results’ validity. First, we developed a selection equation based on our 12 initial factors and 3 more: firearms availability, binge drinking frequency, and assailant not stranger. The equation predicted probability of homicide victimization versus any other means of death (disease, suicide, accident). We used the equation to determine a hazard rate for each homicide victim. The factors firearms availability and assailant not stranger were measured dichotomously. The former indicated whether respondents perceived the decedent had, in the year preceding death, frequented surroundings where firearms were present (1 indicating yes, 0 no). The latter indicated whether respondents perceived that the assailant committing the homicide was known to the decedent (1 indicating yes, 0 no). We measured decedents’ binge drinking frequency (in the year preceding death) as a continuous variable, using six offered responses (never to daily). The term binge drinking indicated consumption of five or more drinks on one occasion.

We included firearms availability, assailant not stranger, and binge drinking frequency in our selection equation because previous research indicated their association with homicide victimization (Allen & Lo, 2012; Cooper & Smith, 2011; Shaw et al., 2006) and because additional variables would help ensure an identified outcome model (Beck, 1983). We included the obtained hazard rate measures as an additional variable in our final regression model. The hazard rate stated the likelihood a particular decedent would be included—as a homicide victim—in the final model (Beck, 1983).

We wanted to evaluate race/ethnicity’s role in associations between age at homicide victimization and each independent variable. To do so, we compared those coefficients—generated separately for each racial/ethnic subsample—obtained for factors yielding at least two significant coefficients. We regressed victimization age on (a) the main effects of the dummy variables Black and Hispanic (White was the reference), (b) all independent variables, and (c) interactions involving the independent variable in question and the two race/ethnicity variables. A significant interaction observed via this testing process indicated that the independent variable had affected homicide victimization age differently within the minority subsamples versus the White subsample. To facilitate unbiased estimation of regression results, we used the sample weight provided by the data set, along with STATA software.

Results

Table 1 presents the simple statistics describing all of the included variables. Black homicide victims in our total sample (N = 1,409) died significantly younger than White ones, while Hispanic homicide victims were the youngest of all. With three exceptions, our independent variables’ associations with race/ethnicity differed significantly from one racial/ethnic subsample to another. The three exceptions were the variables received mental health care, prescription drug misused, and street drug used.

Table 1.

Simple Statistics Describing All Included Variables.

	White	Black	Hispanic	Total	p value
Variables	M or %	M or %	M or %	M or %	p value
Ln (homicide victimization age)	3.59	3.34	3.30	3.41	0.01
U.S. born	96.6%	96.1%	53.3%	88.3%	0.01
Male	83.0%	80.9%	96.2%	84.3%	0.01
Received AFDC	3.7%	8.4%	3.1%	6.1%	0.01
At least some college	24.8%	15.7%	9.2%	17.0%	0.01
Employed	58.2%	44.8%	55.6%	50.5%	0.01
Married	25.1%	13.9%	23.8%	18.7%	0.01
Home owner	11.5%	2.9%	1.9%	5.0%	0.01
Firearm discharge	61.6%	69.0%	70.9%	67.4%	0.02
Received mental health care	4.2%	3.3%	3.4%	3.5%	0.73
Alcohol-related death	12.3%	3.9%	11.5%	7.6%	0.01
Street drug used	22.7%	19.7%	21.5%	20.9%	0.49
Prescription drug misused	5.7%	3.1%	5.7%	19.5%	0.06
N	383	765	261	1,409

Note. Chi-square tests were used to evaluate relationships between Race and each dichotomous variable. F tests were used to examine the relationship between race and the continuous variable. AFDC = Aid to Families With Dependent Children.

In addition, results indicated that, compared with the minority subsamples, a larger percentage of our White subsample had been (in the final year of life) U.S. born, married, employed, homeowners, and possessed of some college education, and moreover experienced an alcohol-related death. Victims in our Black subsample were more likely than White or Hispanic victims to have received AFDC; Hispanic victims were more likely than White or Black victims to be male and to have had a death involving firearm discharge.

Table 2 presents the multivariate regression results. For each ethnic subsample separately, the model regressed age at victimization on the 12 independent variables and on the hazard rates that the selection equation generated. Including no mortalities other than homicide mortalities in the multivariate analyses did not lead to a significant selection bias for Hispanics or Whites; within the Black subsample, however, it did lead to selection bias. Including hazard rates in our final multiple regression analyses provided appropriate correction of this selection bias.

Table 2.

Linear Regression Explaining Ln (Age of Homicide Victimization).

	White		Black		Hispanic
Variables	b	β	b	β	b	β
Hazard rate	−0.03	−0.03	0.10*	0.09	−0.06	−0.06
U.S. born	0.05	0.02	0.06	0.03	−0.06	−0.08
Male	−0.02	−0.03	−0.06	−0.06	0.02	0.02
Received AFDC	−0.07	−0.04	−0.07	−0.05	−0.04	−0.02
At least some college	0.08	0.08	0.05	0.05	0.07	0.06
Employed	−0.17**	−0.21	0.05*	0.07	0.14**	0.20
Married	0.18**	0.21	0.33**	0.32	0.19**	0.22
Home owner	0.45**	0.34	0.46**	0.21	0.54**	0.20
Firearm discharge	0.00	0.00	−0.17**	−0.21	−0.17**	−0.21
Received MH care	0.01	0.00	0.16*	0.08	0.14	0.09
Alcohol-related death	−0.29**	−0.23	−0.13*	−0.07	−0.08	−0.07
Street drug used	−0.10	−0.10	0.01	0.01	0.07	0.08
Prescription drug misused	0.10	0.06	0.07	0.03	0.05	0.03
Constant	3.60**		3.29**		3.31**
F test	9.78**		15.62**		6.57
Weighted cases	383		765		261
Adjusted R²	23.0%		19.9%		21.8%

Note. AFDC = Aid to Families With Dependent Children.

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

For Whites in our study, 4 of 12 independent variables exerted significant impacts on homicide age (see Table 2). Married people and homeowners in the White subsample tended to be older than unmarried people and non-homeowners when victimized. Employed people and those whose homicides involved alcohol tended to be younger when victimized. For Blacks, 6 of 12 independent variables proved significant. Among Blacks, in the final year of life, an older homicide victim was likelier than a younger one to be married, employed, a homeowner, and to have received mental health care; a younger Black homicide victim, in turn, was likelier to have died by firearm and to have had an alcohol-related death. For Hispanics, 4 of 12 independent variables yielded significant results. A Hispanic decedent who had been victimized at an older age tended to be employed, married, and a homeowner, while one victimized relatively younger tended to have died by firearm. Our model explained about 23%, 20%, and 22%, respectively, of the variance in homicide age for the White, Black, and Hispanic subsamples.

Again, we intended to evaluate race/ethnicity’s role in victimization age’s associations with all 12 independent variables. The evaluation procedure involved comparing coefficients generated for each racial/ethnic subsample (as described in the “Data Analysis” subsection). Boldfaced figures in Table 2 signify significant results, which, in part, revealed that employment’s impact on victimization age was much stronger for Whites than for Blacks or Hispanics: A negative association was observed between the two variables for Whites, positive associations for Blacks and Hispanics. We also found marriage to be associated with older victimization ages, an association much stronger among Blacks than Whites. Among Blacks and Hispanics, firearm discharge involved in death was associated significantly with younger victimization ages, with the two subsamples’ significant firearm–age relationships much stronger than that of Whites. Concerning victims reported to have died in alcohol-related homicides, Whites were more likely than Blacks or Hispanics to have been relatively young. The White subsample’s intercept term was significantly larger than that of the Hispanic or Black subsamples, which suggests marked Black–White and Hispanic–White disparities in victimization age. Indeed in our study, after controlling the effects posed by the independent factors and accounting for the hazard rate (i.e., selection bias), a White homicide victim lived 10 years longer, on average, than a Black victim and 9 years longer, on average, than a Hispanic victim.

Discussion

Applying the multiple disadvantage model, we evaluated cross-sectional data from the 1993 NMFS in an effort to identify associations between age at time of homicide victimization (in 1993), race, and 12 social structural, mental health−related, and lifestyle variables. We addressed possible selection bias linked to our excluding from analyses all decedents who died from causes other than homicide. We ensured such bias would not compromise our findings’ validity. In general, findings supported the multiple disadvantage model’s premise that social structural, mental health−related, and lifestyle factors are associated with victimization age. Findings also constituted empirical evidence of race/ethnicity’s capacity to moderate the relationships between victimization age and several social structural and lifestyle factors.

Observed associations generated distinct profiles for each of the racial/ethnic groups studied. Overall, younger Black victims were unmarried, unemployed, not homeowners, murdered with a firearm, and involved in deadly alcohol-related incidents. The same was true of younger White victims, with the exception of firearm involvement, and of younger Hispanic victims, with the exception of the involvement of alcohol. We found race to play a moderating role in homicide victims’ age at time of demise.

Several important points can be made in light of our results. First, our results confirmed that social structural factors are inherently valuable for understanding homicide victimization and that race moderates associations between victimization age and employment, marital status, and home ownership. Specifically, employment during the final year of life was a protective factor extending Black and Hispanic homicide victims’ life spans, although no such effect was observed among Whites. More Whites than minorities have assets that allow them to go without employment or to retire from employment—to be, that is, neither employed nor unemployed (Killewald, 2013; Lo et al., 2014). Thus, identifying an employment−victimization age relationship among Whites that is the inverse of that found for minorities provides evidence of advantage, where life expectancy is concerned, in being White.

Confirming previous research, the present results illustrated a protective role for marriage against homicide victimization at early ages (Sampson, Laub, & Wimer, 2006). Theoretically, marriage is a social relationship factor promoting spousal financial, emotional, and social support (House, 1987); it also typically links individuals to conventional society and to others, via responsibilities to the spouse (Sampson & Laub, 1993). Empirical evidence has correlated marriage with lower drinking levels and better health (Tenorio & Lo, 2011). Interestingly, in our study, marriage’s life-prolonging effect was strongest in the Black subsample—the one whose marriage rate was lowest. This result may connote Blacks’ strong sense of obligation, where family responsibilities are concerned, should marriage take place.

In addition to being a financial asset, ownership of a house results from having financial assets. Much research has demonstrated that Whites are likelier than Blacks or Hispanics to have financial assets, including houses (Kim & Richardson, 2012). Our study found a positive relationship between home ownership and age at time of homicide victimization, a result having two possible implications. First, financial stability leads to home ownership, as well as to reduced risk of victimization early in life. Second, home ownership may have provided decedents in our study with substantial protection against early victimization. This is important knowledge, as Black Americans who fall victim to homicide tend to do so earlier than do White or Hispanic Americans who are murdered. And, our analyses showed all three factors—employment, marriage, and home ownership, or its correlate, financial stability—to have protective roles where Black homicide victims are concerned, delaying, at least, violent demise. In light of such findings, policies and programs fostering Black Americans’ employment and home ownership, and nurturing Black individuals as they adjust to divorce or widowhood, deserve attention and support (Bloom, Hodges, Kern, & McFaddin, 1985).

Our results confirmed, second, that mental health−related factors and lifestyle factors typifying a given racial/ethnic group influence the general age range tending to describe homicide victims from that group. For example, results for the White subsample showed involvement of alcohol to more markedly imperil those individuals than minority individuals. The literature, too, shows Whites to be likelier than Blacks to drink heavily while young—a rite of passage of sorts, as their drinking tapers off in adulthood (Kanny, Liu, & Brewer, 2011). Thus, our significant finding that alcohol consumption is related to homicide victimization of relatively young White Americans clearly aligns with the literature. We found firearm involvement, in contrast, to be irrelevant to Whites’ victimization age, despite its significant power to truncate Blacks’ and Hispanics’ lives. Firearm availability and widespread substance abuse have been empirically linked to homicide victimization as important risk factors (Chauhan et al., 2011; Logan et al., 2011; Shahpar & Li, 1999). The two risks have also been used to explain high rates of homicide in the period 1985 to 1994 (Shahpar & Li, 1999). Our study’s outcome, however, constituted age at time of homicide victimization, and our results identified firearm involvement as one of the factors truncating Black and Hispanic murder victims’ lives. Again, alcohol—but not other substances—affected our White subsample more than our minority subsamples.

Our model included just one mental health−related factor: receipt of mental health care. In the Black subsample, it proved to be associated with older victimization age; no such association was observed in the other subsamples. Being treated in a facility serving, specifically, psychiatric and alcohol/drug patients was clearly associated, in our study, with Blacks’ longer lives. Improving Black Americans’ access to and utilization of mental health care, then, is important.

The third implication stems from the average victimization ages in our subsamples. We obtained average victimization ages for Blacks (n = 765) and for Hispanics (n = 261) were significantly lower than that observed for the White subsample (n = 383): 28 and 27 years versus 36 years. This accords with much published research from the United States showing murdered young Blacks outnumber murdered young Whites (Dobrin et al., 2005; Felson & Painter-Davis, 2012). When we controlled all of the model’s variables, the model intercepts for the three subsamples remained significant. These results suggest that variables not considered by this study could help explain racial differences in age at time of homicide victimization.

Our results clearly demonstrate the moderating role that—as the multiple disadvantage model suggests—race/ethnicity plays in association between age at time of homicide victimization and social structural, mental health−related, and lifestyle factors (Lo et al., 2013). Future testing of the multiple disadvantage model might include mental health−related and lifestyle factors as variables mediating social structural factors’ effects on age at time of homicide victimization. This could further elaborate social dynamics underlying the fact that having racial/ethnic minority status, along with social structural disadvantages, poses increased risk of homicide victimization while young.

Three study limitations should be mentioned. First, using secondary data prevented us from including all factors relevant to the multiple disadvantage model. Many such factors, but especially neighborhood deprivation, have previously been linked to homicide victimization (Ezell & Tanner-Smith, 2009; Gartner, 1990; Gjelsvik et al., 2004). Given the limited number of variables generated from the multiple disadvantage model used for this study, and the fact that we tested some basic directional and magnitude-specific hypotheses, our study should be considered only a partial test of this theoretical model. Second, the data we extracted from NMFS were retrospective data. The NMFS data collection instrument required interviewees to recall events, circumstances, and details pertaining to a decedent. Although our independent variables were selected for their factual, behavioral, readily observable nature, the data remain inherently limited. The degree of an interviewee’s familiarity with a decedent may or may not support accurate response to survey items; as well, grief at a loved one’s death may distort memories of him or her. Despite these things, our study captured important profiles of Black, Hispanic, and White victims of homicide, thereby furthering the field’s understanding of racial disparity in violent victimization.

The third limitation centers on the age of the data set. The 1993 survey included important variables but is less current than we would like. We will certainly seek more up-to-date data that include variables suitable for testing the multiple disadvantage model. Still, our real aim was to test basic hypotheses using the multiple disadvantage model, and the results we obtained illustrate key factors in racial disparity in homicide victimization. Our findings shed some light on how policies and programs may be modified to diminish, if not eliminate, disparity. The final limitation to be discussed involves the appearance, based on our results, that age at time of homicide victimization is an effect in a causal model of social structural, mental health−related, and lifestyle factors. Yes, these factors and homicide death clearly have a temporal order. Nevertheless, because we cannot be certain no spurious variables were present, the study results yield only hypothetical causal inferences.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Celia C. Lo, PhD, is a professor in the School of Social Work at the University of Alabama, Tuscaloosa. Her research interests include the sociology of drugs and alcohol, disparities in mental health and access to treatment services, and juvenile delinquency.

Rebecca J. Howell, PhD, is an assistant professor in the Department of Criminal Justice at Charleston Southern University. Her research interests include the etiology of delinquency and drug use, program evaluation, and criminological theory testing.

Tyrone C. Cheng, PhD, is an associate professor in the School of Social Work at the University of Alabama, Tuscaloosa. His research interests include welfare reforms, Medicaid policies, and child welfare and drug use.

References

Allen

A. N.

C. C.

(2012). Drugs, guns, and disadvantaged youths: Co-occurring behavior and the code of the street. Crime & Delinquency, 58, 932-953.

Anderson

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

Aneshensel

C. S.

(1992). Social stress: Theory and research. Annual Review of Sociology, 18, 15-38.

Beck

R. A.

(1983). An introduction to simple selection bias in sociological data. American Sociological Review, 48, 386-398.

Bloom

B. L.

Hodges

W. F.

Kern

M. B.

McFaddin

S. C.

(1985). A preventive intervention program for the newly separated: Final evaluations. American Journal of Orthopsychiatry, 55, 9-26.

Chauhan

Cerda

Messner

S. F.

Tracy

Tardiff

Galea

(2011). Race/ethnic-specific homicide rates in New York city: Evaluating the impact of broken windows policing and crack cocaine markets. Homicide Studies, 15, 268-290. doi:10.1177/1088767911416917

Cockerham

W. C.

(2005). Health lifestyle theory and the convergence of agency and structure. Journal of Health and Social Behavior, 46, 51-67.

Cooper

Smith

E. L.

(2011). Homicide trends in the United States, 1980–2008: Annual rates for 2009 and 2010. Washington, DC: U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics.

Dobrin

Lee

Price

(2005). Neighborhood structure differences between homicide victims and non-victims. Journal of Criminal Justice, 33, 137-143. doi:10.1016/j.jcrimjus.2004.12.005

10.

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. doi:10.1177/1088767908330493

11.

Felson

R. B.

Painter-Davis

(2012). Another cost of being a young black male: Race, weaponry, and lethal outcomes in assaults. Social Science Research, 41, 1241-1253. doi:10.1016/j.ssresearch.2012.04.006

12.

Gartner

(1990). The victims of homicide: A temporal and cross-national comparison. American Sociological Review, 55, 92-106. doi:10.2307/2095705

13.

Gjelsvik

Zierler

Blume

(2004). Homicide risk across race and class: A small area analysis in Massachusetts and Rhode Island. Journal of Urban Health-Bulletin of the New York Academy of Medicine, 81, 702-718. doi:10.1093/jurban/jth152

14.

Grogger

Willis

(2000). The emergence of crack cocaine and the rise in urban crime rates. Review of Economics and Statistics, 82, 519-529. doi:10.1162/003465300558957

15.

House

J. S.

(1987). Social support and social structure. Sociological Forum, 2, 135-146.

16.

Kanny

Liu

Brewer

R. D.

(2011). Binge drinking: United States, 2009. In Morbidity and Mortality Weekly Report (Ed.), CDC health disparities and inequalities report–United States, 2011 (Vol. 60, pp. 101-104). Atlanta, GA: National Center for Chronic Disease Prevention and Health Promotion, CDC. Retrieved from http://www.cdc.gov/mmwr/pdf/other/su6001.pdf

17.

Killewald

(2013). Return to being black, living in the red: A race gap in wealth that goes beyond social origins. Demography, 50, 1177-1195. doi:10.1007/s13524-012-0190-0

18.

Kim

Richardson

(2012). The impact of socioeconomic inequalities and lack of health insurance on physical functioning among middle-aged and older adults in the United States. Health and Social Care in the Community, 20, 42-51. doi:10.1111/j.1365-2524.2011.01012.x

19.

Kochanek

K. D.

Murphy

S. L.

Minino

A. M.

Kung

H. C.

(2011). Deaths: Preliminary data for 2009. National Vital Statistics Reports, 59(4). Retrieved from http://www.cdc.gov/nchs/data/nvsr/nvsr59/nvsr59_04.pdf

20.

Krug

E. G.

Dahlberg

L. L.

Mercy

J. A.

Zwi

A. B.

Lozano

(2002). World report on violence and health. Geneva, Switzerland: World Health Organization.

21.

Kuhns

J. B.

Wilson

D. B.

Clodfelter

T. A.

Maguire

E. R.

Ainsworth

S. A.

(2011). A meta-analysis of alcohol toxicology study findings among homicide victims. Addiction, 106, 62-72. doi:10.1111/j.1360-0443.2010.03153.x

22.

Levitt

S. D.

(2004). Understanding why crime fell in the 1990s: Four factors that explain the decline and six that do not. Journal of Economic Perspectives, 18, 163-190. doi:10.1257/089533004773563485

23.

Link

B. G.

Phelan

(1995). Social conditions as fundamental causes of disease. Journal of Health and Social Behavior, 35(Extra issue), 80-94.

24.

C. C.

Cheng

T. C.

(2011). Racial/Ethnic differences in access to substance abuse treatment. Journal of Health Care for the Poor and Underserved, 22, 621-637.

25.

C. C.

Cheng

T. C.

(2012). Discrimination’s role in minority groups’ rates of substance-use disorder. American Journal on Addictions, 21, 150-156.

26.

C. C.

Cheng

T. C.

Howell

R. J.

(2014). Access to and utilization of health services as pathway to racial disparities in serious mental illness. Community Mental Health Journal, 50, 251-257. Available online January 13, 2013.

27.

C. C.

Howell

R. J.

Cheng

T. C.

(2013). Explaining Black-White differences in homicide victimization. Aggression and Violent Behavior, 18, 125-134. doi:10.1016/j.avb.2012.11.006

28.

Logan

J. E.

Smith

S. G.

Stevens

M. R.

(2011). Homicides-United States, 1999–2007. In Morbidity and Mortality Weekly Report (Ed.), CDC Health disparities and inequalities report—United States, 2011 (Suppl. 60, pp. 67-70): CDC, U.S. Department of Health and Human Services. Retrieved from http://www.cdc.gov/mmwr/pdf/other/su6001.pdf

29.

National Center for Health Statistics. (2012). Health, United States, 2011: With Special Feature on Socioeconomic Status and Health. Hyattsville, MD: Center for Disease Control and Prevention, U.S. Department of Health and Human Services.

30.

Pearlin

L. I.

(1989). The sociological-study of stress. Journal of Health and Social Behavior, 30, 241-256.

31.

Phelan

J. C.

Link

B. G.

(2005). Controlling disease and creating disparities: A fundamental cause perspective. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 60, 27-33.

32.

Phelan

J. C.

Link

B. G.

Diez-Roux

Kawachi

Levin

(2004). “Fundamental causes” of social inequalities in mortality: A test of the theory. Journal of Health and Social Behavior, 45, 265-285.

33.

Piquero

A. R.

MacDonald

Dobrin

Daigle

L. E.

Cullen

F. T.

(2005). Self-control, violent offending, and homicide victimization: Assessing the general theory of crime. Journal of Quantitative Criminology, 21, 55-71. doi:10.1007/s10940-004-1787-2

34.

Sampson

R. J.

Laub

J. H.

(1993). Crime in the making: Pathways and turning points through life. Cambridge, MA: Harvard University Press.

35.

Sampson

R. J.

Laub

J. H.

Wimer

(2006). Does marriage reduce crime? A counterfactual approach to within-individual causal effects. Criminology, 44, 465-508. doi:10.1111/j.1745-9125.2006.00055.x

36.

Shahpar

G. H.

(1999). Homicide mortality in the United States, 1935-1994: Age, period, and cohort effects. American Journal of Epidemiology, 150, 1213-1222.

37.

Shaw

Hunt

I. M.

Flynn

Amos

Meehan

Robinson

. . . Appleby

(2006). The role of alcohol and drugs in homicides in England and Wales. Addiction, 101, 1117-1124. doi:10.1111/j.1360-0443.2006.01483.x

38.

South

S. J.

Messner

S. F.

(2000). Crime and demography: Multiple linkages, reciprocal relations. Annual Review of Sociology, 26, 83-106. doi:10.1146/annurev.soc.26.1.83

39.

Steffensmeier

Feldmeyer

Harris

C. T.

Ulmer

J. T.

(2011). Reassessing trends in Black violent crime, 1980-2008: Sorting out the “Hispanic effect” in Uniform Crime Reports’ arrest, national crime victimization survey offenders’ estimates, and U.S. prisoner counts. Criminology, 49, 197-251. doi:10.1111/j.1745-9125.2010.00222.x

40.

Tenorio

K. A.

C. C.

(2011). Social location, social integration, and the co-occurrence of substance abuse and psychological distress. American Journal of Drug and Alcohol Abuse, 37, 218-223.

41.

Wilson

W. J.

(1987). The truly disadvantaged: The inner city, the underclass, and public policy. Chicago, IL: The University of Chicago Press.