Where You From? Examining the Relationship between Gang Migrants and Gang-Related Homicide

: Research has frequently focused on the increased likelihood of violence and homicide among gang-involved individuals, as well as on the factors that contribute to this violence. Such work has examined the relationship between immigration and the frequency of crime, as well. However, there is a dearth of research examining the likelihood of gang-related homicide and the presence of both gang migrants from within the U.S. and those from abroad in a given community. The current paper utilizes National Youth Gang Survey data to examine the relationship between law enforcement perceptions of gang migrants in their jurisdiction and the frequency of gang-related homicide. The results reveal that gang-related homicides have a signiﬁcant and negative association with the presence of gang migrants. These ﬁndings have important policy implications for understanding and addressing serious gang violence and homicide at the community level.


Introduction
Research has regularly revealed that gang members are more likely to become involved in crime and violence [1], as well as that gang members are more likely to engage in homicide. Due to this consistent relationship between gang involvement and violence, scholarship has examined specific characteristics of gangs and their members in order to improve comprehension of gang member criminality [2][3][4]. One area in which very little recent work has been conducted is on gang migrants from inside and outside of the United States. 1 Work conducted by Maxson (1997) revealed that gang migrants move for a variety of reasons, with the most frequent reason for moving being social [5]. Additional work has examined gang growth and migration in select jurisdictions in the United States, with varying results [6][7][8][9]. However, little work has examined the relationship between gang migrants and serious crime, including homicide.
Research has examined the relationship between immigration and crime with mixed results. While there are works that have found a significant and positive relationship between immigration and crime, suggesting that immigration brings with it an increased likelihood of crime in a community [10], most research has either found no significant relationship between immigration and crime [11] or a significant and negative association [12,13]. Scholarship examining immigration and gangs has revealed that in select areas, immigrants are more likely to join gangs [1]; however, this does not necessarily mean gangs with immigrants are likely to commit more crime. In fact, Duran's (2018) work highlights high levels of disproportionate minority contact with the system in areas close to the border. He specifically argues that the practices of "White Diversion" and "Minority Delinquentization" are occurring, and that disproportionate minority contact coincides with increased gang involvement [14].
Research has not yet examined whether the presence of gang migrants impacts the likelihood of serious gang violence such as homicide, or whether areas where police report high concentrations of gang migrants contributing to gang violence influence the probability of gang-related homicide. Given that research regularly finds a negative relationship between immigration and crime and a positive association between gang involvement and crime, it is critical to understand whether the concentration of gang member migrants in a community influences the likelihood of gang-related homicides. The current study contributes to this gap in the literature by analyzing the relationship between police reports of the percentage of gang member migrants in their jurisdiction, gang member migrants contributing to gang violence in their jurisdiction, and number of reported gang-related homicides. The results have implications for policy and can specifically inform approaches for communities to collaborate in order to address and prevent serious gang crime and violence more effectively.
The paper begins with a discussion of the relationship between immigration and crime, transitions into addressing work on gangs and crime, and concludes by addressing scholarship related to immigration and gangs. The paper then moves into the proposed hypotheses, analyses, and results and concludes with a discussion of the theoretical and policy implications. Specifically, the paper discusses the importance of police and community collaboration, as perceptions of serious violence and homicide vary between police, the public, and other data sources. These results have implications for collaborative community approaches seeking to reduce the likelihood of gang-related homicides and other serious gang violence, improve perceptions of safety in these communities, and foster positive relationships among community members.

Gang Violence and Homicide
Research regularly finds that gang members are more likely than non-gang members to participate in violence [1,[15][16][17] and that gang members engage in violent behavior that is related to both gang and non-gang issues [18,19]. Research further reveals that gang violence may result from racial issues in combination with a need for group protection while trying to achieve control of territory in the same community [20][21][22][23]. Gang violence frequently occurs both between gangs [3,24] and within gangs [25]. More specifically, gangs regularly engage in violence due to internal gang conflict or rivalries as well as because of drugs, domestic violence, or robbery [26].
According to Decker and Curry (2002), gang homicides have unique characteristics that include the race of victim and perpetrator, drug involvement, victim-offender relationship, weapon use, spatial concentration, and sex. It is important to examine the likelihood of gang homicides among areas that police report gang migrants contributing to gang violence, as the nature of gang homicides are different than non-gang homicides [27][28][29][30]. Specifically, gang homicides often include incidents of both drive-by and walk-by shootings, homicide as a method of retaliation, and higher rates in concentrated spaces/areas of a community [3,15,[31][32][33]. Similarly, Maxson et al. (1985) compared gang and non-gang homicides and found that gang homicides had a higher probability of involving gun use, vehicles, and occurring in public spaces [34]. Additionally, law enforcement has characterized gang homicides as having a greater likelihood of being violent and having more individuals involved as offender, accomplice, and victim. Furthermore, they found differences in a variety of individual-level characteristics, including age, ethnicity, and the relationship between involved parties.
Problems specific to gang homicides include the effect on families and communities [35]. Gangs influence the mentality of individual gang members in such a way that they may be willing to give their own life or engage in murder to show their commitment [36]. Research reveals that gang violence is impacted by multiple factors [37] which affect the likelihood of gang homicide, such as the frequency of gang membership in an area, population density, and social and economic deprivation [38]. Scholarship has not, however, specifically examined the influence that the presence of gang migrants has on the likelihood of gang-related homicide.

Gang Member Migration and Crime
Research has analyzed gang migrants both within the U.S. and abroad. Specifically, research has focused on why gang members move from one city to another and the impact of this on gang proliferation and crime. The occurrence of gang members migrating has been identified for several decades [39][40][41][42]. To this day, there has been very little work on gang migration, and most of this is community specific. Skolnick et al. (1988) found in their research with inmates and correctional staff that gangs expand to new areas to sell illegal drugs. Gangs in Milwaukee, for example, were primarily an outgrowth of Milwaukee, with only a small portion of the gang members coming from the Chicago area [6]. Although there are cases of gang members or gangs migrating to different communities in Southern California, it is not common [8]. Arguably, most gangs are groups of youth that do not have the resources to establish themselves outside of their own turf. Research in Kenosha, Wisconsin highlights that although police perceive that gangs in the area are migrating from Chicago, most of the gangs and gang problems are occurring due to local economic and social issues [9]. Furthermore, research examining why gang members move reveals that while they move for a variety of reasons, including illegal attractions such as drug market expansion, the most frequent reasons for moving were social and family-related [5].
According to the NYGS, the majority of jurisdictions reporting issues with gangs identified gang member migrants in the community [43]. Although the media often report migrating gang members as an issue that is getting worse, there is little data to support this, and there is little work on the topic. The general public views gang member migrants as contributing to gang violence, drugs, and conflict [44,45]. Research suggests that street gangs are connected to gangs in different communities due to gang migration [46]; additional work has revealed a relationship between gang migration and heightened gang activity, violent crime, and drug crime [47]. Other work, however, is contradictory, revealing a lack of significant association between gang member migrants and gang-related crimes [43].
Gang members migrate for both illegitimate and legitimate reasons, with legitimate social reasons the most frequent rationale reported for gang member migration [5]. It is possible that international gang member migration is connected to the adoption of different gang styles instead of gang substance [22]. Specifically, McGuire (2006) argues that although gang members may move, it is possible that gang style and practices are migrating as well [48]. Conversely, there have been reports of gang members migrating to different cities in Canada [46]. This includes local Canadian gangs establishing themselves in various areas in Canada. Street gangs in Toronto have arrived from the United States and Jamaica, and work has revealed gangs that moved from smaller to larger areas in Canada. Additionally, local gangs have been established in multiple communities, and there have been reports of gangs migrating from the U.S. to major cities in Canada such as Toronto [46]. Gang member migration is therefore relatively common and occurs for a variety of reasons. Results are mixed on the relationship between gang member migrants and gang-related crime, with suggestions that gang members may move to expand criminal enterprises, while other work identifies moving for social reasons [5]. Little research has examined the relationship between gang migrants and serious gang crime such as homicide.

Immigration, Crime, and Gangs
Given that the current study examines gang migrants from inside and outside the United States, it is important to discuss the relationship between immigration and crime. Research examining the relationship between immigration and crime regularly finds no significant association or a negative association with criminal behavior [49][50][51][52][53][54][55]. For example, immigrant youth are more afraid of being victimized by someone with a weapon compared to non-immigrant youth [56]. This suggests that there is a negative relationship between low acculturation and a high level of fear towards crime.
Communities located on the border in Texas tend to have Latino homicide rates that are lower than 50% of those in communities located away from the border [57]. Interestingly, in analysis of homicides in El Paso, TX, Emerick et al. 2014 found no significant association between immigration and homicide; however, there was a significant and positive relationship between the percentage of Latinos and gang-related homicides. Research reveals lower levels of homicide in areas with higher immigrant concentration; however, when focusing specifically on gang-related homicides there is a significant and positive association with immigration [58]. Other research has revealed that expatriate Latinos living in the U.S. have an increased likelihood of cooperating with law enforcement compared to Latinos born in the U.S. [12]. However, Latinos that have experienced police assault or have gang-involved friends have a lower probability of cooperation with law enforcement. Other studies have revealed that foreign-born Latinos have an increased probability of cooperating with police compared to their U.S.-born counterparts, and that this is consistent even when examining violent gang crimes [13,59].
Borders have unique conditions that can contribute to the production of gangs that are different than gangs in other areas, such as a "bi-national barrio-prison-cartel hybrid" gang located in El Paso [60]. Due to the already large number of gang members in the area, there is a large recruitment pool for gangs to grow. The different methods utilized to identify gangs and gang members could lead to both over-policing and criminalization of young immigrants [61]. Research by Esbensen and Carson (2012) revealed that when determining the major characteristics of gang youth, immigrant status was not found to be one of them, especially during a youth's time in middle school. More specifically, youth who reported being born outside of the United States had a lower probability of being gang-involved when younger; however, once these youth turn 15, immigrant youth make up a noticeable portion of the gang youth in their sample.
Gang members born outside of the United States are less likely to participate in crime compared to gang members born in the United States [62]. However, gang-involved youth are found to be more likely to participate in delinquency compared to non-gang youth irrespective of whether or not they were born outside of the United States. Consistent with these findings, Valdez et al. (2009) examined 28 homicides with Mexican-American gang members and found no immigrant youth involvement. Although similarly aged immigrant youth lived in the same areas where the violent incidents occurred, their involvement in gangs and violence was minimal [63]. Conversely, Hollis (2018) found a significant and positive relationship between Latino immigration and a greater amount of crime; however, this relationship is specific to non-gang crime [64]. When examining gang crime specifically, there is no significant association with Latino immigration. These findings suggest that even though gangs and gang involvement increase the likelihood of violence, the presence of immigrants may aid in reducing this probability of gang violence.

Research Questions and Hypotheses
Although research has examined the relationship between immigration, gangs, and crime and gang migrants and crime, there has been little work that has specifically examined the relationship between gang migrants from within and outside the United States and gang violence, including gang-related homicide. Given the mixed and limited findings in the research examining immigration, gang migrants from inside and outside of the United States, and gang crime, the current study asks the following question: 1. What is the relationship between gang migrants and gang-related homicide? Specifically, this paper hypothesizes that: a. In jurisdictions where law enforcement reports that gang migrants from inside the United States significantly influence gang-related violence, gang related-homicides are less likely to occur; b.
In jurisdictions where law enforcement reports that gang migrants from outside the United States significantly influence gang-related violence, gang-related homicides are less likely to occur; c.
The greater the percentage of migrant gang members in a jurisdiction, the less likely gang-related homicides are to occur.

Data Description
The current study utilizes data from the 2012 National Youth Gang Survey (NYGS), which was the last year in which the NYGS was distributed. The NYGS was originally implemented to assess the severity of gang issues in the United States by examining law enforcement perspectives on where gangs exist, the different characteristics they have, and how their behaviors vary across communities. The items utilized in the NYGS are supported by scholarship [65]. Data collection occurred through the distribution of surveys to various law enforcement jurisdictions across the U.S. Specifically, all departments in jurisdictions of at least 50,000 and all suburban counties were sent surveys. Rural jurisdictions provided their perspective as well; surveys were provided to a random sample of law enforcement offices in communities with populations between 2500 and 49,000, in addition to counties categorized as rural. The strategy resulted in a representative sample with an 85% response rate from law enforcement jurisdictions across the United States, making it a good fit for the present study. The data allow for the ability to examine gang related homicide frequencies across jurisdictions that report varying levels of issues of gang migrants from the law enforcement perspective.

Analytical Methods
The analyses for the current study included both bivariate and multivariate methods. Due to the use of a variety of variable types, including dichotomous and ordinal independent variables and a count dependent variable, one-way analysis of variance was utilized. This allowed for comparison of gang migrant characteristics with gang-related homicides in different jurisdictions. In order to account for additional variables and decreased spuriousness, multivariate analyses were utilized. Due to the skewed count level dependent variable, negative binomial regression was used. Additionally, in order to account for missing cases in the data, multiple imputation was utilized in STATA (Release 16). By utilizing multiple imputation, values are able to imputed in the data where limited information is provided. When using multiple imputation, the assumption is made that there is no relationship between the likelihood of missing data on one variable and the variable's real value [66,67].

Variables
The number of gang-related homicides in a jurisdiction during the past year that were reported by law enforcement was included as the dependent variable. The number of gang-related homicides in the last year was determined through surveys distributed to various law enforcement jurisdictions across the United States.
The analyses included three independent variables. First, law enforcement was asked to provide the percentage of gang migrants in their community that were gang members. This was a scale level variable from 0-4, where 1 = 1-25%, 2 = includes 26-50%, 3 = 51-75%, and 4 = 76-100%. Second, two dichotomous variables asking law enforcement whether either gang members migrating from outside the United States or gang members migrating from inside the United States were significantly contributing to gang violence in their jurisdiction were included as independent variables.
Various control variables were included in the analyses. The number of gang-related homicides may differ across regions due to variations in methods of gang member identification [68]; thus, different regions of the United States (Northeast, Midwest, West, and South) were included as control variables. Moreover, given the potential relationship between gang presence and homicides, both the number of gangs and number of gang members were included in the analyses. Additional control variables included jurisdiction population, the length of time measured in years that law enforcement have reported gang problems, and whether or not the jurisdiction has a gang unit. Lastly, variables measuring whether there was a greater amount of offending before an individual joined a gang and/or more offending once they were committed to a gang were included as control variables.

Results
The data in Table 1 includes descriptive information on the average number of gangrelated homicides in relation to the control variables included in the analyses. When looking at the number of active gangs reported in a jurisdiction, the average number of gang homicides reported increases with the number of active gangs in a jurisdiction. The mean number of gang-related homicides peaks at about 50 active gangs. This is similar when looking at the number of active gang members. The mean number of gang-related homicides peaks at about 500 active gang members. The average number of gang-related homicides tends to increase with the number of years police report that gangs have been a problem. Similarly, the more populated an area is, the higher the average number of gang-related homicides. Jurisdictions that report increased levels of offending after gang joining report a higher number of gang-related homicides on average compared to jurisdictions that do not report increased levels of offending after gang joining. The data highlight regional variations in the average number of gang homicides; jurisdictions with gang units report a higher number of gang-related homicides on average compared to jurisdictions without a gang unit.

Bivariate Analyses
The results of the bivariate analyses between police perceptions of gang migrants influencing gang related violence, reported percentage of gang migrants in a jurisdiction, and gang related homicides are displayed in Table 2. The average number of gang-related homicides is over twice as high in jurisdictions that do not report gang migrants from other U.S. jurisdictions influencing violence compared to jurisdictions that do report this influence. Conversely, there is minimal difference in the average number of gang-related homicides reported in jurisdictions where police identify gang migrants from outside the U.S. influencing gang violence compared to those that do not. Lastly, when police report that the percentage of gang members in a jurisdiction is greater than 25%, the average number of gang-related homicides begins to decrease. The results of the bivariate analyses do not reveal any statistically significant findings; however, these findings highlight the need to conduct multivariate analyses in order to improve comprehension of the relationship between gang migration and gang-related homicides.

Multivariate Results
The results in Table 3 include two models; Model 1 includes police reporting on gang migrants from the U.S. contributing to gang related violence in their jurisdiction as an independent variable, and Model 1 include police reporting on gang migrants from outside the U.S. contributing to gang related violence in their jurisdiction.
The findings in Model 1 do not support the first hypothesis. There is a significant and negative association between reporting that gang migrants within the U.S. influence gang-related violence and gang-related homicides. Therefore, in jurisdictions where police report that gang migrants within the U.S. significantly influence gang related violence, gang-related homicides are significantly less likely to occur. Additional findings reveal a significant and positive association between the number of active gangs and gangrelated homicides, and a significant and negative association between the number of active gang members and gang-related homicides. Furthermore, the number of years there has been a gang problem, jurisdiction population, and increased likelihood of offending upon gang joining were all significant and positively associated with the likelihood of gang-related homicides. The results in Model 2 do not support the second hypothesis. The relationship between police reporting that gang migrants from outside United States significantly influence gangrelated violence and gang-related homicides is negative, although it is not significant. Additional results reveal a significant and positive relationship between the number of active gangs and gang-related homicides, and a significant and negative association when comparing the number of active gang members and gang-related homicides. Moreover, the number of years there has been a gang problem, jurisdiction population, and increased likelihood of offending upon gang joining were all found to significantly increase the probability of gang-related homicide.
The results in Table 4 include Model 3, which consists of the percentage of gang migrants as the independent variable, and Model 4, which is the full model with percentage of gang migrants, police reporting of gang migrants from inside the U.S., and police reporting of gang migrants from outside the U.S. all significantly influencing gang-related violence. The results in Model 3 show support for the third hypothesis. There is a significant and negative association between the percentage of gang migrants in a jurisdiction and gang-related homicides. Additional findings reveal that the number of years there has been a gang problem, jurisdiction population, and increased likelihood of offending upon gang joining were all significant and positively associated with the likelihood of gang-related homicide. Furthermore, there is a significant and positive association between the number of active gangs and gang-related homicides, and a significant and negative association between the number of active gang members and gang-related homicides.
While the results in Model 4 show support for the third hypothesis, they do not support either the first or second hypotheses. There is a significant and negative association between the percentage of gang migrants in a jurisdiction and gang-related homicides. There is a negative association between police reporting that gang migrants within the U.S. influence gang-related violence and gang-related homicides, although it is not significant. The relationship between police reporting that gang migrants from outside the U.S. significantly influences gang-related violence and gang-related homicides is negative, although thus is not significant either. Additional findings reveal that the number of years there has been a gang problem, jurisdiction population, and increased likelihood of offending upon gang joining were all found to significantly increase the probability of gang-related homicides.
Moreover, there is a significant and positive relationship between the number of active gangs and gang-related homicides, and a significant and negative association when comparing the number of active gang members and gang-related homicides.

Summary of Findings and Discussion
Although Model 1 supported the first hypothesis, the full model (Model 4) did not show significant support for or against hypothesis one or two. Both Models 3 and 4 showed significant support for the third hypothesis, demonstrating that there is a significant and negative association between the percentage of gang members that are gang migrants in a community and the probability of gang-related homicides. The results suggest that gang migrants may aid in reducing the likelihood of serious crimes such as homicide. This is consistent with other research which has found that immigrants coming into a community can help to revitalize an area by strengthening social ties and support with one another, arguably reducing the likelihood of crime [11]. These findings are consistent with work examining gang migrants within the U.S. For example, Maxson (1997) found that although at times gang members may move for illicit reasons, most of the time they move for family-related or other social reasons. This suggests that they are not migrating in order to contribute to serious gang violence and homicide in the community.

Policy Implications
The results of this study highlight the need for community collaboration. Given that the presence of gang migrants significantly reduces the likelihood of gang-related homicides, communities should arguably be working to embrace and integrate these individuals into the area with the goal of reducing gang joining and other gang-related crimes. In areas where police may perceive and/or observe that gang migrants from within or outside the U.S. significantly contribute to gang-related violence, it is important to acknowledge that this does not necessarily result in an increase in the likelihood of gang-related homicides. Community education and exposure to gang migrant behavior could potentially bring perceptions in line with reality and help to reduce the likelihood of gang-related homicides as well as serious gang violence. It is important to adopt practices to aid in changing community perceptions of gang migrants. Therefore, it is essential to develop and utilize intervention methods that are culturally relevant [69]. This includes implementing methods that are family focused, adjusting perspectives, and providing resources for support.
The results highlight the need for community collaborative approaches to improve comprehension of gang migrant behavior. Specifically, if the presence of gang migrants significantly reduces the likelihood of gang-related homicides in a community, it is essential to understand why members are getting involved in gangs and the illegal behaviors they are engaging in. This could potentially be accomplished through community policing. Implementing community partnerships and collaborations that focus on communication, education, and the identification of the issues within specific areas will aid in improving community and police perceptions of gang migrants. By understanding their behaviors, the community can help address the needs of individuals who have recently relocated to their area, which may ultimately reduce the likelihood of other gang-related crimes as well.
Regional variation exists in gang member identification methods used by law enforcement in the United States [68]. More specifically, the use of gang signs and symbols to identify gang members is common [70]. Given that the presence of gang migrants decreases the likelihood of gang-related homicides, law enforcement may want to incorporate community collaboration into their gang member identification practices and protocols. Specifically, before officially identifying individuals as criminal gang members officers could communicate with members of the community to determine individual identity and status. This would strengthen overall community cohesion and potentially reduce both gang involvement among migrants and gang crime in general.
One of the biggest obstacles to this line of research is adjusting and improving both community and police perceptions of gang migrants. Community center or service projects may help with this. Bringing the community together for various social and/or service gatherings to improve an area, such as cleanup, can arguably help to change perceptions [71,72]. Research shows that individuals may join gangs due to feeling marginalized [73]. Youth born in the United States who experienced discrimination-related stress have a higher probability of gang involvement [74]. Causes of gang joining may vary for Latinos born in the U.S. compared to Latinos born elsewhere; this is because stress due to discrimination or adaptation is not a predictor of gang membership for immigrant youth, with economic inequality rather being reported as a major reason for gang joining. Based on the results of the current study, gang migrants do not relocate to become involved in serious violence, and preventing social and economic marginalization may aid in better relationships overall and potentially reduce the likelihood of migrants joining gangs as well as other gang-related crime.

Limitations and Conclusions
The data available on this topic are limited for a variety of reasons. This study is unable to establish causation. The data used were taken from the most recent NYGS and are crosssectional. Thus, various control variables were utilized in a multivariate analytic approach in order to reduce the amount of spuriousness and to strengthen the findings. Although this is the most current available national data in the United States relating to the perspectives of law enforcement, future work will need to be conducted utilizing more current data in order to support the current findings. Furthermore, the percentage of gang migrants and of gang migrants influencing gang-related violence were determined through the perceptions of law enforcement. There are various limitations to using law enforcement perceptions to examine gang migrants, gang violence, and homicide. Law enforcement perspectives do not necessarily represent official crime data in a community, and rather represent the subjective opinions of law enforcement [75]; gang and member representations in the media may influence officer perspectives and result in implicit bias [76]. Future research will need to examine this issue from different community perspectives to better comprehend the relationship between gang migrants and gang related homicides and other crimes. Furthermore, it would be beneficial for future research to examine the relationship between gang migrants from both inside and outside of the United States and crime more generally.
In summation, the results of this study are consistent with past research examining immigration and crime. With a focus on gang migrants, this research contributes to the current literature by revealing that gang migrants reduce the likelihood of gang-related homicides in a jurisdiction. Given that gangs and gang members tend to be more likely to participate in crimes, including serious violence, compared to non-gang individuals, this work adds to the complexity of scholarship on gangs and gang involvement. Research has consistently argued for the need to have specialized gang policies and programs [77]. The current findings support this argument and suggest that focusing on collaboration between community groups and law enforcement will contribute to overall understanding of gang involvement, reduce the likelihood of serious gang violence, and improve relationships among newly-arrived and longtime members of the community.
Funding: This research received no external funding.
Institutional Review Board Statement: Ethical review and approval were not required for this study, due to the data being publicly available with no individual identifiers.

Informed Consent Statement:
The current study is secondary data analysis, but informed consent was obtained when the data was originally collected. Data Availability Statement: Data for the National Youth Gang Survey can be accessed through the Inter-university Consortium for Political and Social Research (icpsr.umich.edu).

Conflicts of Interest:
The authors declare no conflict of interest.

1
Although "gang migrants" frequently refers to gang members moving from one jurisdiction to another within the same country, it can refer to gang members moving from one country to another as well; see van Gemert et al., 2008. 2 Outliers dropped.