Border Residents’ Perceptions of Crime and Security in El Paso, Texas

: Research has continued to show the overall safety of the U.S. border region contrary to the widespread belief about the insecurity of the U.S.-Mexico border and frequent claims for the need to secure the border in order to prevent the spread of violence into the rest of the country. Rarely do we ask how border residents feel about safety and crime, which could shed significant light on the claims that the border is an insecure warzone posing a threat to the entire country. While calls to secure national borders are common, outsiders’ perceptions of an unsafe border are not supported by official crime rates and statistics, Border Patrol apprehensions, or the everyday experiences of people in American cities along the U.S.-Mexico border. This paper investigates the perception of crime and security, as expressed by the residents of El Paso, Texas, a large city located along the U.S.-Mexico border and directly across from Ciudad Juárez, Mexico. Data come from a National Institutes of Health (NIH) funded survey that asked 919 residents about their perceptions of crime, sense of security and safety in their neighborhood and the city in general. The results show that the overwhelming majority of border city residents feel safe and that those who are undocumented and raised in El Paso are the most likely to report feeling safe or very safe. We also find that the foreign-born population had a statistically significant lower felony conviction rate than those who were U.S.-born, an important qualifier in discussions over immigration and its connection with violence and crime. Contrary to sensationalized claims about border violence, residents of El Paso do not display any of the sense of insecurity experienced in neighboring Ciudad Juárez, Mexico. We present hypotheses about possible causes for these low levels of violence in the U.S.-side of the border and discuss the dissonance between the reality on the border and perception outside of the border region.


Introduction
Anti-immigrant rhetoric seems to be at an all-time high, with the current administration and its depictions of the border as insecure, violent, and chaotic. Donald Trump spoke of the 'migrant caravan' during the 2018 elections and threatened to go as far as to close the entire southern border (Blitzer 2018). He engaged in a government shutdown over disagreements with the Democratic Party around immigration, demanded funds for a wall in his State of the Union addresses, and declared a national emergency at the southern border in a speech to the nation on February 15, 2019. However, this perception of an unsafe border is not supported by official crime statistics (Martinez and Stowell 2012), Border Patrol apprehensions, and-perhaps most importantly--it contradicts the everyday experiences of residents living in cities along the U.S.-Mexico border, as this paper demonstrates.
This phenomenon predates Trump, however. Politicians and the media portray the southwest border as insecure, out of control, and crime-ridden. To mention but a few examples, former Arizona Senator John McCain said,"[T]he border is not only unsafe, but it is also being patrolled in an ineffective manner" (Goff and Sáenz 2014). A piece published in Foreign Policy Magazine placed the binational U.S.-Mexico border region amongst the most dangerous in the world (Walker 2011). News stories claim that border cities provide a haven for Mexico's drug violence (Gillum et al. 2011). Influential right-wing commentator Ann Coulter writes that immigrants cost taxpayers money in terms of "the number of police, missing persons operators, hospital emergency room doctors, surgeons, prosecutors, judges, court clerks, prison guards, and rape counselors made necessary by criminal aliens" (Coulter 2015:Chapter 7). Like other commentators, Coulter invokes images of the sexual predator, child abductor, and criminal. As Beinart (2016) writes, "As antiblack racists have for centuries, today's anti-Latino commentators depict their targets as a sexual threat." This stems from a long-held stereotype of Latin people as hypersexual. This is implied in the racist term macho as used in conversations in English to convey hegemonic masculinity as standard for all Latinos. President Trump's discourse reproduces racist anti-Mexican rhetoric (this author) and further amplifies and legitimizes the wrongful confounding of migration and crime. This process is what some scholars have termed crimmigration, and results in racial profiling as well as the incarceration of minorities for not being able to obtain immigration papers, rather than committing any crime (Armenta 2017). Trump's announcement of his presidential campaign included the now-infamous statement: When Mexico sends its people, they're not sending their best. They're not sending you. They're not sending you. They're sending people that have lots of problems, and they're bringing those problems with us. They're bringing drugs. They're bringing crime. They're rapists. (Press Conference, Trump Tower, New York City, 16 June 2015).
Trump's claim about Mexican rapists and criminals coming through the border is an example of the wide-diffusion and purchase of this border myth, necessary to Trump's political rally cry of building a wall (this author). Amidst a continuous push for a border wall under the pretense of widespread lawlessness and crime, attention has turned to El Paso, Texas. Nevertheless, its low crime rates seem to contradict this narrative. When Trump tried to rally support for his border wall plans in El Paso in 2019, Beto O'Rourke, a former Congressman from El Paso, brought together El Paso residents in a large town hall meeting to challenge this fear-based immigration rhetoric (Landler and Romero 2019). Many residents spoke out about how safe their community was, due to the residents themselves and their tight-knit communities, with over 83 percent of El Paso residents being Latin people or immigrants from Mexico. Residents did not consider the wall between El Paso and Juarez as contributing to this sense of safety.
Despite attention to El Paso in the news, there has been relatively little research done about how residents living in the U.S.-Mexico border perceive crime. Asking the border residents-those who experience the reality of living in the borderlands-how they perceive the safety of their neighborhoods allows for a more authentic interpretation of border security. Compared to the robust findings on the links between immigration and crime based on crime statistics, the existing research on how residents living in mixed communities perceive crime is lacking. This paper responds to this gap, directly addressing how border residents in El Paso perceive crime in their community and neighborhoods.

Empirical Links between Immigration and Crime
While many urban residents outside of the border region may equate increased immigrant presence with dangerous neighborhoods (this author), this perception may indeed be unfounded as research has continuously shown that when the number of immigrants increases in a neighborhood, crime decreases (Miller and Peguero 2018). As long as fears about immigration have been around, empirical research has questioned and challenged this link (Hourwich 1912). The literature on immigration and crime shows similar or lower crime rates for the foreign-born as compared to the US-born (Ceobanu 2011;Davies and Fagan 2012;Kubrin 2007;Lee and Martinez 2009;Martinez and Stowell 2012;Stansfield 2014;Stowell et al. 2009;Wadsworth 2010). Immigration has been shown to reduce violent crime. This challenges old functionalist theories about the disorganizing nature of immigration in local communities. Martinez et al. (2016) investigated the relationship between homicides and immigration in San Diego, an immigrant gateway similar to El Paso, between 1960 and 2010. They found that when San Diego neighborhoods experienced an influx of immigrants, there were fewer homicides. Similarly, Ferraro (2016) conducted fixed effects regression analyses of data from 1252 cities and towns between 2000 to 2007, which included 194 new destinations (areas with populations over 1000 and where immigration increased by at least 150 percent within the last ten years). Ferraro (2016) found that new immigration destinations experienced greater decreases in crime in comparison to other areas. Martinez et al. (2004) examined whether immigrant enclaves and ethnic and racial composition affect drug homicides in San Diego. They found that the economic deprivation and concentration of skilled workers were significant predictors of drug homicides in San Diego, not the number of immigrants, showing that crime rates are a class issue, not a racial one. Martinez and Stowell (2012) also analyzed the relationship between homicides and immigration in Miami and San Antonio in the 1980s and 1990s. They found that immigration had no effect on homicide, even when they controlled for Latino regional concentration and homicide motivations. As today, in the 1990s there was also fear of violent and drug-related crime, sometimes along the border, and it was also encouraged by the executive branch's policy agenda and discourses about being tough on crime.
This finding has been shown to hold outside of immigrant gateways. Ousey and Kubrin (2009) conducted a longitudinal analysis over 20 years across 159 U.S. cities, looking at how the violent crime rate is affected by the percentage of foreign-born individuals and the percentage of individuals who do not speak English. They found that immigration decreases felony homicides and drug-related homicides, but increases gang-related homicides (which may be because police categorize homicides within immigrant communities as gang-related), all at statistically significant rates. Using data from the Uniform Crime Reports and the Current Population Surveys, Butcher and Piehl (1998) found that foreign-born youth engage in less crime than U.S.-born youth. Similarly, Reid et al. (2005) used data from the 2000 U.S. Census and the 2000 Uniform Crime Report to investigate the relationship between immigration and crime at the macro-level in and across a random sample of 150 metropolitan areas. They found that immigration does not increase crime rates and that it may decrease crime.
In studies that found that crime increased, this increase was mediated not by increases in immigration but by a different variable-such as reduced labor market opportunities or family structure-and differences in incarceration rates due to pre-detention practices and racial profiling (Hagan and Palloni 1999). Spenkuch (2013) finds that immigration is correlated with an increase in financial-related crime, such as theft, but only when increased immigration is coupled with poor labor market opportunities.
However, something else, such as underreporting, may be behind low crime rates in El Paso and among immigrant communities in general. Zatz and Smith (2012) write that while research has typically focused on the relationship between immigrants and crime in neighborhoods in general, little research has been done to investigate the relationship between immigrants and their communities. Immigrants are particularly vulnerable to victimization and are often not protected by the criminal justice system because they are "afraid to call the police or otherwise draw attention to themselves" (p. 147). Immigrants are more likely to be robbed and extorted but are less likely to report it to the police (McDonald andErez 2007, Nuño 2013).
Nevertheless, the analysis of data from the San Diego Police Department from 1980-2000 shows that immigrants are not more likely to be victims of homicide than those who are native-born (Martinez et al. 2010). Besides homicides, this possible under-reporting bias can be overcome by asking immigrants directly whether they have been victims of crime. Barranco and Shihadeh (2015) write that undocumented Latino immigrants are particularly vulnerable for victimization as they often carry cash and their legal status can make them reluctant to report crimes to the police. They found that an area with many undocumented immigrants can lead to a "spillover" victimization, where attackers cannot differentiate between new immigrants and the rest of the community. Crime increases because attackers wish to attack those undocumented and cannot differentiate between them and others. Some question the widespread nature of this "walking ATM" stereotype. Nonetheless, some can confirm through interviews with street offenders, the targeting of caramel skin people because of this stereotype (Caraballo 2017:54).
One reason why we see a decrease in crime in response to an increase in immigration is that immigrants seek to improve their circumstances in the long-term and are thus not drawn to crime (Tonry 1997). Tonry (1997) also theorizes that the second generation of immigrants might be more likely to engage in crime because of "social strain," and discusses how minorities in many countries are overpoliced and because of this disproportionately likely to be arrested, convicted, and imprisoned for crimes. Another theory is that in response to an influx of new immigrants, residents will put pressure on officials to police more, leading to a decrease in crime overall (Ousey and Kubrin 2009). A third reason is that ethnic networks have substantial social control mechanisms (Zhou and Bankston 1998). Immigrants living in neighborhoods with other co-ethnics are more likely to know neighbors and thus, indirectly may keep others in check.
New immigrants often move into low-income neighborhoods with high levels of crime because rent is lower than other areas or because this is where they can find housing. The immigrant revitalization theory argues that neighborhoods with high immigrant populations experience a reduction in violent crime (Martinez et al. 2010). Vélez (2009) finds that the arrival of new immigrants in an area with concentrated disadvantage reduces the number of homicides in those areas. Vélez analyzed data from 786 census tracts in Chicago that covered homicides and recent immigrant concentrations from 1993-1995 and found that immigrants revitalize neighborhoods by developing strong ties to families, clergy, social service providers, and officials, allowing them to generate social control. They help "reinvigorate an ethnic enclave economy" (p. 327), providing social capital for the residents through creating businesses. Immigration, "reinvigorates local economic opportunity structures and social networks" and revitalizes the neighborhood (p. 325). Third, they strengthen community institutions. There may be a Latino paradox, which Sampson (2006) defines as "Hispanic Americans doing better on a range of social indicators, including propensity to violence." Does the immigrant revitalization theory hold for border communities and immigrant gateways? At one point, increases in border immigration enforcement measured through reported apprehensions by the border patrol and person-hours spent watching the border were correlated with a decrease in crime in urban areas (Coronado and Orrenius 2007). However, the causal relation still remains unclear. Little research addresses crime in U.S. border urban areas using original data.

Perceptions of Crime and Fear of Crime
Particularly as it relates to immigrant communities, perceptions of crime and security are central to understanding how outsiders and residents understand safety and security in their communities. With anti-immigrant rhetoric at an all-time high, a discussion of how perceptions crime and safety are tied to immigration is key. While immigration has been shown to either decrease or not increase crime, Nunziata (2015), using data from Western European counties, found that when immigration increased, fear of crime and negative attitudes towards immigrants increased even though crime did not increase. Fear of crime, particularly that attributed to immigrants, has been shown to encourage voting for anti-immigrant politicians. Some have found that an influx of immigration into a country or local area has been correlated with higher votes for anti-immigration politicians (Golder 2003, Knigge 1998), while others have found the opposite or no change (Van Der Brug et al. 2005). Dinas and van Spanje (2011) found that there is not a link between immigration and anti-immigrant party popularity; but rather, when crime rates are attributed to immigrants either by the media, politicians, or both, these parties gain popularity and legitimation whether or not their charges are based on fact.
Anti-immigrant rhetoric has affected policies and created exclusion. Provine and Doty (2011) find that policy responses "reinforce racialized anxieties" (p. 261). They argue that three federal responses-building up the border, increased partnerships with police, and increased reliance on Immigration and Customs Enforcement (ICE)--have resulted in a racialized language regarding how the U.S. handles immigration. They and many others argue that immigration has been criminalized and that the law has sustained racial inequalities (Armenta 2017, Dowling and Inda 2013, Golash-Boza 2015. As immigrants are more likely to be the victims of crime, rather than the perpetrators, this rhetoric can intensify their vulnerability and provide cover for exploiters and human traffickers (Zatz and Smith 2012).
Anti-immigrant rhetoric or policies that might result in the social and political exclusion of immigrants may increase crime. Lyons et al. (2013) propose that the link between immigration and the decrease in crime found in neighborhoods is mediated by increased social and political participation. Using data from the National Neighborhood Crime Study, they found that when immigrants have access to elected offices, there is a decrease in violence in those neighborhoods.

U.S.-Mexico Border and Crime
Mexico has experienced high levels of violence in the last several years, yet it seemingly has not spilled over into the United States. Ciudad Juárez was considered one of the most dangerous cities in the world, with 282.7 homicides per 100,000 residents. Located directly across Ciudad Juárez, El Paso, Texas, has been ranked for a while as the safest U.S. city with a population of over 500,000 inhabitants (CQ Press 2013). Despite the proximity, border cities like El Paso are among the safest in the country (Lee et al. 2001). Violent crime rates in three Texan U.S. border cities-Laredo, Brownsville, and McAllen--decreased from 2006 to 2010 during the peak of drug war violence in northern Mexico (Correa-Cabrera 2013). If the national crime rates paralleled the rates along the US-Mexico border, there "would have been 5720 fewer homicides, 159,036 fewer property crimes, and 99,205 fewer violent crimes" (Nowrasteh 2019).
However, as shown in an FBI informative video, the rhetoric of danger does not necessarily reflect reality. The narrator starts, "Drug trafficking. Human smuggling. Extortion. Murder. Corrupt public officials. All these crimes represent a multi-billion-dollar industry, and they pose a threat not only to communities on both sides of the border but to our national security as well" (FBI n.d.). The Assistant Director of the FBI's Criminal Investigative Division, Kevin Perkins, said, "What we do have on the U.S. side of the border are kidnappings, corruption issues, and the illegal drug trade that goes through. The cash that's flowing south, the weapons that are flowing south, and the drugs that are flowing north." Crime and immigration statistics are also exaggerated. In January 2019, President Donald Trump said that border agents "apprehended, last year, 17,000 criminals trying to get across the border" (Trump 2019). Sixty-three percent of the 16,831 "criminal aliens" were individuals who failed routine checks at legal entry points (Woolf 2019). The other 37 percent had criminal records, but 47 percent of these records were for illegally entering or re-entering the U.S., 14 percent were for DUIs, and 13 percent were for a violent conviction.

Methods and Data
Data were collected through in-person confidential interviews conducted between 2011 and 2012 in El Paso, Texas. El Paso is the second largest border metropolitan area after the San Diego metropolitan area, and the second in terms of border crossings in the U.S. The data collection followed a purposive maximum variation sampling technique to show the heterogeneity of Latin people. Participants were reached out to by bilingual university students, often with immigrant backgrounds themselves. These students were IRB certified and received training for six weeks in methodology and sensitivity to minorities and vulnerable populations. The survey and sampling frame were modeled on a similar study conducted by Hector Balcazar, Joe Heyman, and colleagues (Lapeyrouse et al. 2012). No identifying information was collected to protect participants. The University of Texas at the El Paso Institutional Review Board and the National Institute on Minority Health and Health Disparities approved the study. The study was funded through a larger National Institutes of Health grant.
The social survey included questions regarding border residents' perceptions of security. The sample analyzed consists of 919 housed Hispanic residents of El Paso. A limitation of this paper is that it only includes the views of Hispanics, and this is partly because Hispanics comprise 83 percent of total residents in El Paso (US Census 2019). However, this is not a significant limitation-much of the border is majority Hispanic, especially in Texas, so the people who live on the border would be the ones that experience the reality and can know whether or not there is a violent crisis with the potential to "spillover." Ethnographic data also tell us that these perceptions of safety are also true for other ethno-racial groups living in the city. The data from this survey were analyzed using IBM SPSS 25 statistical software. Demographics, perceptions of safety, description of neighborhood, and criminal conduct are reported. If an individual chose not to respond to a question, they are not included in the reported results. Felony convictions between foreign-born and US-born respondents were compared using binomial logistic regression. We also conducted a multinomial logistic regression to investigate how demographic variables affected the perception of safety.
The security of a city or neighborhood is often measured by the change in the volume of crime from one period to the next or by comparing the crime rate in one jurisdiction with another one. In contrast, we rely on the subjective feeling of security that residents have (which matches the low crime rate of El Paso), and self-reported felony convictions, which are rarely asked about in large surveys. We contribute to the literature by pointing out, in novel ways, the discrepancy between the perception of safety at the border and anti-immigrant rhetoric that associates immigration with an increase in crime and a border region that is plagued by crime.

Demographics
The sample has a larger percentage of female respondents (55.8 percent) than male respondents (44.2 percent). Table 1 shows the demographic data of respondents. The median age of those surveyed is 29 years old. Over 60 percent were born in the United States, while 39.7 percent are foreign-born. The highest percentage of respondents are second-generation immigrants born in the U.S. to at least one foreign-born parent (35.5 percent). The families of many respondents have been in the U.S. for many generations. In terms of citizenship, 79.7 percent of respondents are U.S. citizens, and only 3.5 percent of respondents identified as undocumented residents of El Paso.

Perceptions of Safety
Survey respondents were asked how safe they felt living in El Paso. Respondents chose from the following options; "very safe," "safe," or "not safe." Table 2 identifies that 96.9 percent of respondents reported feeling either "very safe" or "safe." The majority of those individuals described feeling "very safe." Only a small percentage reported feeling "not safe." We then compared citizenship status with perceptions of safety and found that undocumented residents reported the highest percentage for feeling "not safe." The valid percentages are presented regarding the percentages from those responding.

Neighborhood Descriptions
By design, respondents were selected from neighborhoods across El Paso. Respondents were asked how they described their neighborhood.
The question was open-ended. In order to avoid priming or biasing the answers, the survey did not list any descriptions for the respondents to choose from. Though respondents provided a variety of answers, they were easily categorized. While some of the descriptions can be used somewhat interchangeably, we kept "calm" separate from "good" or "safe" to keep some of the nuances reported. Table 3 illustrates that the majority of respondents (43.2 percent) described their neighborhood as "calm," followed closely by "good" and "safe." The vast majority used terms that characterize their neighborhood in a positive light. Only 3.2 percent of individuals described their neighborhood as having a "bad reputation." Even then, saying that a neighborhood has a bad reputation does not mean that the respondent agrees, but it could mean an acknowledgment that others perceived it as dangerous. Again, El Paso residents demonstrate feeling safe in their neighborhoods and talk about them mainly in positive terms.

Felony Conviction
In response to the question, "Have you encountered any problems with the law [criminal, civil, immigration court]?", 17.3 percent of those born in the U.S. answered in the affirmative while a lower percentage (9.5 percent) of those foreign-born reported any problems with the law.
When asked whether respondents have ever been convicted of a felony, our study found that 97.5 percent reported never being convicted of a felony. Only 2.5 percent of respondents reported having a felony conviction. We compared foreign-born populations with criminal conduct using binomial logistic regression. The foreign-born population had a statistically significantly lower felony conviction (at 1.1 percent) than those who were U.S.-born (3.5 percent, p < 0.05). A little over 10 percent overall reported having "been unjustly detained by the police," 12.9 percent of the U.S.-born Hispanics surveyed said so, while only 6.2 percent of foreign-born Hispanics reported having been detained unjustly by the police. As anti-immigrant rhetoric often points to the increased criminality of immigrants, these figures may tell a different story-lower felony conviction rates and experiencing unjust detention by the police instead allude to overpolicing and detention, not increased criminality.

Results
In this analysis, our dependent variable is safety, whether residents felt "very safe," "safe," or "not safe." This response comes directly from the survey and is thus self-reported and subjective, as is often the case in public opinion surveys. Independent variables were: gender, years spent living in El Paso, whether they were raised in El Paso, if they were foreign-born, the number of years of education they had, and their immigration status (if they were a U.S. citizen, resident, temporary resident, had a work visa, had a student visa, or were undocumented). All nominal variables were recoded as dummy variables. The variables included in this model were chosen because they were identified as predictors of crime in areas with high immigration, as discussed in the literature review.
We used a listwise or case deletion of missing data, the most frequently used approach, to run a multinomial logistic regression analysis using SPSS Version 25. Due to the fact that the missing data do fulfill the requirements of missing completely at random, as there is a large enough sample and power is not an issue, we can use listwise or case deletion (Kang 2013). After the listwise deletion, the breakdown of perceptions of safety paralleled the pre-listwise deletion, where 61.6 percent reported feeling very safe, 35.8 percent reported feeling safe, and 2.6 percent reported feeling unsafe.
In Table 4, the reference category is feeling unsafe in El Paso. Except for Migrant generation, years of education, and years living in El Paso, all variables are coded as dummy variables. Each of the variables in Table 4 is in reference to their opposite attribute (i.e., "male" is in comparison to female, "not raised in El Paso" is compared to those raised in El Paso, "Not a U.S. citizen" is compared to being a U.S. citizen). Only gender, U.S. citizenship status, and having a work visa are significant (p < 0.10). Our results show that men are 6.525 times as likely as women to feel very safe as opposed to not safe and 8.115 times as likely to feel safe as opposed to not safe (p < 0.10). Our results also show that individuals who do not have a work visa are much more likely to feel unsafe than very safe in comparison to those who hold work visas (p < 0.001). Finally, those who are not U.S. citizens are 0.070 times as likely to report feeling safe when compared to those who are U.S. citizens (p < 0.10).
While not statistically significant, it is notable that those who are undocumented are more likely to perceive El Paso as very safe or safe in comparison to those who are not undocumented. Another notable finding, albeit not statistically significant, is that those not raised in El Paso are more likely to perceive the city as very safe or safe. This may be because many of those not raised in El Paso are immigrants, and their referents for safety are more unsafe than El Paso. The homeless respondents, which are not included in the sample used for this paper, are less likely to feel safe than their counterparts.
These results may have one main limitation. Some have suggested that there need be at least ten events to run a logistic regression analysis (Peduzzi et al. 1996). After performing the listwise case deletion, only nine individuals reported feeling unsafe. Because there are only nine cases, the regression coefficients can be biased, the confidence intervals can have improper coverage, and the Wald statistic can be conservative. However, later studies have questioned the minimum of ten events rule (Vittinghoff and McCulloch 2006). One study showed that "the worse instances of each problem were not severe with 5-9 EPV (events per variable) and usually comparable to those with 10-16 EPV" (p. 717). Because the issues described earlier are seen more frequently with Cox regressions than logistic regressions, and because the frequencies in the tables parallel those seen in the total sample pre-deletion, we can still perform the logistic regression and are confident in the results.

Discussion
A clear way of determining how safe the southwest border is can be established through the perceptions of border residents. This is a relatively novel methodological approach and a contribution to the literature. In this paper, we dispel misconceptions about high levels of crime in a border city with data provided by border residents themselves, which parallel papers analyzing crime rates (Ferraro 2016;Lyons et al. 2013;Martinez et al. 2004;Martinez et al. 2016;Ousey and Kubrin 2009). Based on our research, residents of El Paso, Texas, feel safe living next to the border with Mexico. We also found that U.S. citizens are more likely to feel safe than non-U.S. citizens, and the fear of non-citizens may be more related to ICE and deportation than violence and crime. These findings support other studies that find that women are more likely to report a higher fear of crime than their male counterparts (Scott 2003). In summary, these findings parallel other empirical data that find border cities are safer than the media suggests and safer than cities of equivalent size far from the border (Castañeda and Heyman 2012).
Similar to this paper, the Border Network for Human Rights (2010) surveyed 1222 residents from ten major border cities from California to Texas, and found that 67.1 percent of respondents reported feeling safe living in a border community, 69.7 percent felt their border neighborhood is as safe as most U.S. neighborhoods, 51.8 percent felt safe allowing their child to play in a neighborhood park, and 87.5 percent said they felt safe walking or driving in their neighborhood. A survey for Cronkite News, Univision, and Dallas Morning News across cities on both sides of the border found that 92 percent of respondents in the U.S. border felt safe in their neighborhood at night, compared with 54 percent on the Mexican side. Furthermore, only 14 percent of border residents in the U.S. said that the U.S. should definitively build a border wall (Bilker 2016, López 2016.

Border Crime Paradox
Mexico has indeed suffered an increase in violence due to an armed war on drugs. While drug trafficking, supply, demand, and business connections are transnational, drug war violence in Northern Mexico has not spilled over to the U.S.-side of the border. Many questions remain unanswered and contradictory claims about the U.S.-Mexican border make it hard to decipher fact from myth.
Despite claims of violence spilling over from Mexico, surveyed El Paso residents do not report a sense of insecurity. These results add to the existing body of evidence that cities with high immigrant populations display a lower crime rate (Martinez et al. 2010) and that residents perceive this to be the case. Research has also found that recent immigrants are less likely to use illicit drugs (Loza, Castañeda, and Diedrich 2017). There is no empirical evidence that supports the claim that an influx of immigrants is a disorganizing force in communities. On the contrary, research suggests that "crime and violence are down, cities are growing in population, poor urban neighborhoods are being economically revitalized, and immigrants are renewing small towns on the verge of withering away" (MacDonald and Sampson 2012). The question remains as to why this occurs.
One hypothesis is that immigrants want to keep a low profile and avoid criminal activity due to a fear of deportation--regardless of whether they are legal permanent residents, visa holders, or undocumented. A second, a non-mutually exclusive hypothesis would be that immigrants come to work hard, save, and provide a better future for their children. The reference group of first-generation immigrants is their former community of origin, which frequently is poorer and more insecure, especially if affected by the violent drug conflict. This hypothesis might explain why undocumented immigrants are likely to perceive El Paso as safe or very safe. Immigrants aspire to achieve social mobility through hard work, which is aided by the relative ease with which Latinos find employment in the United States (Castañeda 2018). A counterargument would be that criminal immigrants get deported, yet a study using immigration statistics from the Department of Homeland Security shows that out of the 393,000 immigrants removed, more than half of them (67.3 percent) had no prior criminal offense (Parrado 2012). Other research has also shown how most deportees had no other crimes beyond the fact of not being granted immigration papers (Golash-Boza 2015).
A third hypothesis is the heavy presence of law enforcement. El Paso has a high number of federal, state, and local law enforcement agencies as compared to other cities, which may deter criminal activity. The federal agencies located in El Paso metropolitan area include U. There is also a significant military presence in El Paso given that Fort Bliss, White Sands Missile Range, and other federal military installations are located in this area. So, on a given street in some middle-class neighborhoods of El Paso, there may be few residents not working in the security and law enforcement sector.
These security and law enforcement organizations counter crime even within their own agencies. An effective way of combating crime starts within an agency and extends outward. Strict measures are in place to intimidate authorities from corrupt practices. High fines and federal imprisonment are measures put in place to dissuade local law enforcement from engaging in corruption and facilitating the illicit trade of narcotics. The authorities have placed rigorous punishments on those found guilty of corruption. For example, in 2008, a customs agent was charged ten years to life in prison for conspiracy to smuggle narcotics and illegal immigrants (Licón 2009). Border Patrol agents can make a considerable salary, incentivizing agents to keep their jobs and avoid risking them for collaborating with cartels and transnational organized crime. The starting salary for a new border agent varies from $38,619 to $49,029, depending on education and experience. With satisfactory service over three years, Border Patrol Agents can earn up to $93,470 a year, plus benefits and overtime (Arbuckle 2016). These salaries are very competitive in the area and incentivize organizations to work against crime and corruption in their own organizations, potentially building a culture of accountability in El Paso.
A fourth hypothesis would be that even while the drug trade is transnational, the social costs of enforcing the war on drugs have been disproportionately placed on Mexico (Dear 2013). Given the lower rate of criminal prosecution and higher impunity in Mexico, it may be more convenient for criminal organizations to engage in competition, violent enforcement, and intimidation in the Mexican-side of the border, while benefiting all parts of the distribution and consumption networks.
A fifth hypothesis would be that the policing of the line (Dunn 2009), the militarization of the border area (Dunn 1996), and the construction of walls and fences (Castañeda 2019), resulted in a decline in crime. While hard to completely dismiss, this hypothesis is not the most likely answer because crime rates were already low before these policies took place. Studies have not proven a direct connection between increased border enforcement and a decrease in crime in U.S. cities.
Another issue at question is whether Latino and immigrant communities are policed more and cited more often than others. For example, Prieto (2018) shows how police in Central California targets immigrants for car impoundment.
The aim of this paper is not to test all of the previously discussed hypotheses. Nevertheless, the data presented show that residents feel safe and live in a context of low crime. Future research should venture into trying to explain this paradox in crime differences across borders within the same illicit industry. A limitation of this paper is that it relies on self-reported feelings about neighborhood safety, crimes, and felonies committed, thus socially desired answers may have caused underreported of criminal activities, yet triangulation of data on education, legal status, housing situation, and drug use and well as research design lead us to think that the vast majority of respondents were truthful. The reliance on self-reporting and subjective understandings is also a strength since the perception of safety relies upon cultural expectations and the general social context. This paper adds evidence from a large community sample about the lack of affinity between crime and immigrant concentration (Martinez and Stowell 2012). Further research should study how perceptions of crime compare in white and mixed neighborhoods and how they vary by political affiliation.

Conclusions
Little research has been conducted on the perceptions of crime along the U.S.-Mexican border. It is essential to understand the perspectives of border residents. This information allows for an authentic interpretation of border security. Based on the results of our study, El Paso residents do not display any sense of danger living within proximity to the border of Mexico. The high concentration of federal law enforcement agencies helps deter criminal activity, but so do the increasing job opportunities in the U.S. and hopes of social mobility for the children of immigrants.
Further, studies have shown that the border fence has not deterred immigration, but it has increased deaths in the desert and extended the time that undocumented workers pass in the U.S. (De León 2015;Massey et al. 2002). The proposals to build border walls should be reconsidered because they incur high symbolic and economic costs with little returns to security. There is not a border security crisis but a humanitarian crisis caused by violence abroad, family separation, and the encampment of asylum seekers in and outside the U.S.

Coda
On 3 August 2019, an armed man shot and killed 22 people and physically injured dozens of others in a Walmart in one of the commercial hearts of the El Paso/Ciudad Juarez metropolitan region. In a letter that he published before driving the long distance between Dallas to El Paso, he explicitly stated his hatred of Hispanics and what he, echoing others, saw as "the invasion of America" by immigrants from Latin America. Thus, this individual not only left people dead and injured, but it also left a sense of terror for brown people who fear that they could be the next targets whether they are US-born, naturalized citizens, or undocumented. This event alone put to the test the sense of safety that most El Paso residents have. Unfortunately, these mass shootings are not unique to El Paso. White nationalism and racial terrorism are new sources of fear for Latin people in El Paso and all over the United States . Nevertheless, El Paso came together as a community to remember those attacked and to celebrate this border community, as residents have expressed the city is a resilient one, and El Paso will come back up, further strengthened by the reminder about the strong social ties of families, neighbors, and community members across the city and the region (Pérez 2019).
Author Contributions: Writing-original draft, E.C. and C.C. Both authors have read and agreed to the published version of the manuscript.