Police Victimization Among Persons Who Inject Drugs Along the U.S.–Mexico Border

Abstract

Objective:

Problematic policing practices are an important driver of HIV infection among persons who inject drugs (PWID) in the U.S.–Mexico border region. This study identifies factors associated with recent (i.e., past 6 months) police victimization (e.g., extortion, physical and sexual violence) in the border city of Tijuana, Mexico.

Method:

From 2011 to 2013, 733 PWID (62% male) were recruited in Tijuana and completed a structured questionnaire. Eligible participants were age 18 years or older, injected illicit drugs within the past month, and spoke Spanish or English. Multivariable logistic regression analyses identified correlates of recent experiences of police victimization (e.g., bribes, unlawful confiscation, physical and sexual violence).

Results:

Overall, 56% of PWID reported a recent police victimization experience in Tijuana. In multivariable logistic regression analyses, factors independently associated with recent police victimization included recent injection of methamphetamine (adjusted odds ratio [AOR] = 1.62; 95% CI [1.18, 2.21]) and recently received injection assistance by a “hit doctor” (AOR = 1.56; 95% CI [1.03, 2.36]). Increased years lived in Tijuana (AOR = 0.98 per year; 95% CI [0.97, 0.99]) and initiating drug use at a later age (AOR = 0.96 per year; 95% CI [0.92, 0.99]) were inversely associated with recent police victimization.

Conclusions:

Physical drugusing markers may increase PWID susceptibility to police targeting and contribute to experiences of victimization. Interventions aimed at reducing police victimization events in the U.S.–Mexico border region should consider PWID’s drug-using behaviors. Reducing problematic policing practices may be a crucial public health strategy to reduce HIV risk among PWID in this region.

Persons who inject drugs (pwid) on the Mexican side of the U.S.–Mexico border region are vulnerable to police victimization (e.g., extortion, physical and sexual violence). In Tijuana, a Mexican city that borders the United States, accounts of police victimization are widespread among PWID and have been linked to behaviors that increase susceptibility to HIV infection (Beletsky et al., 2012, 2013; Infante et al., 2012; Miller et al., 2008; Pinedo et al., 2014a, 2015; Pollini et al., 2008, 2009, 2010; Strathdee et al., 2005, 2008b; Volkmann et al., 2011). Community and problematic policing practices can operate as social-structural drivers of HIV infection (Blankenship & Koester, 2002; Miller et al., 2008; Sarang et al., 2010; Strathdee et al., 2008b, 2010). To decrease their exposure to police, PWID in this region who have experienced police victimization are more likely to take additional risks that increase their risk for HIV. Such risks include engaging in receptive needle sharing, injecting in high-risk environments (i.e., shooting galleries), seeking injection assistance (i.e., use of a “hit doctor”), and rushing injections (Beletsky et al., 2012, 2013; Philbin et al., 2008; Pollini et al., 2008, 2009, 2010; Strathdee et al., 2005; Volkmann et al., 2011). Community and problematic policing practices can adversely affect the adoption of preventive behaviors of PWID and increase their odds of acquiring HIV. For fear of encountering police, for example, PWID may not access needle exchange programs or purchase clean needles (Kerr et al., 2005; Miller et al., 2008; Pollini et al., 2008; Volkmann et al., 2011). HIV disproportionally affects PWID in Tijuana: 4% of men, 10% of women, and 12% of female sex workers who are PWID are HIV positive, compared with 0.3% among the general adult Mexican population age 15–49 (Patterson et al., 2009; Strathdee & Magis-Rodriguez, 2008; Strathdee et al., 2008a).

Understanding factors that increase susceptibility to police victimization is crucial in the border region to inform policies surrounding policing practices. PWID are common police targets given their engagement in drug-related activities (Beletsky et al., 2011b, 2013; Burris et al., 2004; Miller et al., 2008; Strathdee et al., 2010; Werb et al., 2008). Physical and visual markers may play an important role in police victimization. Drawing from sociological and criminology research, labeling theory provides a useful framework for examining how law enforcement officials respond to PWID in this region (Akers, 2013; Bernburg et al., 2006; Lilly & Ball, 2010; Melossi, 1985; Watson et al., 2004). This perspective posits that labels (i.e., drug user, injection drug user) are attached to individuals or groups who engage in behaviors that are perceived as deviant by the larger society. This, in turn, evokes stereotypes and beliefs associated with those labels and influences how others react to and interpret a labeled person’s behaviors. Negative associations with being a drug user may influence how police perceive daily behaviors, resulting in increased police profiling and targeting (Cooper et al., 2005; Rhodes et al., 2006). For example, such common activities as walking in the street or socializing may be perceived as suspicious or deviant for someone marked with an adverse label (e.g., drug user) (Cooper et al., 2005; Kerr et al., 2005).

Using labeling theory as the analytic framework, we conceptualized four domains of factors that may influence victimization experiences by police: (a) sociodemographic factors, (b) structural/environmental factors, (c) drug-using behaviors, and (d) economic factors. The objective of this study was to identify factors associated with recent police victimization. We hypothesized that physical identifiers and characteristics (e.g., drug-using stigmata) that may facilitate police profiling within each of our domains would be associated with recent police victimization.

Method

Study design and participants

Between 2011 and 2013, we recruited 733 PWID to participate in Proyecto El Cuete IV, a prospective observational cohort study examining the impact of the 2010 Mexican drug reform law on the drug use behaviors and HIV risk of PWID in Tijuana, Mexico (Robertson et al., 2014). Briefly, eligible participants were adults age 18 years or older who had evidence of injecting illicit drugs within the past month (confirmed via visual inspection of physical markers of drug injection use), were not planning to permanently move from the city over the next 24 months, and were able to speak Spanish or English. Participants were recruited through targeted sampling using street-based outreach (Watters & Biernacki, 1989). Eligible participants provided written informed consent before being enrolled into the study. Baseline questionnaires were administered by a trained interviewer using computer-assisted personal interviewing in a private setting. Participants received $20 USD for completing the baseline questionnaire. The Institutional Review Board of the University of California, San Diego, and the Ethics Board of the Colegio de la Frontera Norte, Tijuana, approved all study protocols.

Measures

The survey elicited data on sociodemographics, migration history, lifetime and recent HIV risk behaviors, and experiences with police, among other domains. We included measures that fit within our theoretical framework. Sociodemographic measures included gender, age, marital status, education, and language. Structural/environmental factors included years lived in Tijuana, homelessness in past 6 months, hours spent on the street on a typical day, ever traveled to the United States, ever lived in the United States, and ever deported from the United States.

Measures of drug-using behaviors included age at first injection drug use and seeking injection assistance from a hit doctor (i.e., a person who provides assistance with injections, typically in exchange for money, drugs, or sex) in the past 6 months. Participants specified drugs or combinations of drugs they injected in the past 6 months (i.e., heroin, cocaine, heroin and cocaine together, methamphetamine, methamphetamine and heroin together). Regarding economic factors, participants were asked whether their main source of income in the past year was from informal work or odd jobs and were asked their monthly income in pesos (no income; <1,000 pesos; 1,000–1,499 pesos; 1,500–1,999 pesos; 2,000–2,499 pesos; 2,500–2,999 pesos; 3,000–3,499 pesos; or ≥3,500 pesos). Because of the low number of responses in categories, we dichotomized this variable as <3,500 or ≥3,500 pesos (an estimated value of ∼$225 U.S. at press time).

Our binary dependent variable “recent (i.e., past-6-month) police victimization” was constructed by combining affirmative responses to the following eight variables related to problematic policing practices in the past 6 months: (a) was asked for a bribe, (b) had money or valuables confiscated, (c) legal identification documents were confiscated, (d) syringes were confiscated, (e) was physically beaten (hit, punched, kicked), (f) belongings were burned/destroyed, (g) was forced to leave a place of residency, and (h) was asked for a sexual favor to avoid arrest. These variables were chosen based on previous findings linking these problematic policing practices and HIV risk in Tijuana and other international settings (Beletsky et al., 2013; Philbin et al., 2008; Pollini et al., 2008; Robertson et al., 2010; Strathdee et al., 2008b, 2010; Volkmann et al., 2011).

Analysis

We first used descriptive statistics to examine the relationship between recent police victimization and variables within the four domains of our theoretical framework. Variables were tested for association using the Pearson chi-square (binary variables) and Wilcoxon rank sum (continuous variables) tests. In building our final multivariable logistic regression model to identify factors independently associated with recent police victimization, we considered all variables that attained a statistical significance at the p ≤ .10 level in descriptive analyses and variables that were theoretically significant. A correlation matrix was produced to determine collinearity between variables. Independent variables that were highly correlated with our dependent variable (using a threshold of ≥0.5) were excluded from the analysis. For highly correlated independent variables (e.g., types of drugs injected in the past 6 months), we conducted sensitivity analyses by running each variable separately in our final model; no substantive differences were observed. We tested for possible interactions between drug use variables. No collinearity or interaction between variables in the final model was found.

Results

Sample characteristics

Among 733 PWID, 414 (56%) reported experiencing recent police victimization in Tijuana. The majority of participants were male (62%) with an average age of 37.4 years (SD = 8.9) and had lived in the city for an average of 23 years (SD = 14). Most (70%) had previously traveled to the United States, and more than half (59%) had previously lived in the United States; ∼41% of PWID had been deported from the United States. On average, participants began injecting drugs at age 14 and had been injecting drugs for more than 20 years. Past-6-month injection drug use included heroin (95%), cocaine (46%), methamphetamine (56%), and meth-amphetamine and heroin combined (70%).

Factors independently associated with recent experiences of police victimization

In our final multivariable logistic regression model, we identified factors independently associated with recent police victimization in Tijuana (Table 1). Within the sociodemographics domain, men were significantly more vulnerable than women to having experienced recent police victimization (adjusted odds ratio [AOR] = 1.86, 95% CI [1.30, 2.67]). Within the structural/environmental domain, those who had lived more years in Tijuana were less likely to be victimized by police (AOR = 0.98 per year, 95% CI [0.97, 0.99]). Within the drug-using behaviors domain, PWID who initiated injection drug use at an earlier age (AOR = 0.96 per year, 95% CI [0.92, 0.99]), injected methamphetamine in the past 6 months (AOR = 1.62, 95% CI [1.18, 2.21]), and sought injection assistance from a hit doctor in the past 6 months (AOR = 1.56, 95% CI [1.03, 2.36]) had greater odds of experiencing recent victimization by police in Tijuana. Economic factors independently associated with increased odds of police victimization included having an average monthly income of 3,500 pesos or more (AOR = 2.21, 95% CI [1.52, 3.22]).

Table 1.

Factors independently associated with police victimization in the past 6 months among persons who inject drugs in Tijuana, Mexico (N = 733, 2013)

Variable	Adjusted odds ratio	95% CI
Sociodemographics Gender (male)	1.86	[1.30, 2.67]
Structural/environmental factors Years lived in Tijuana	0.98	[0.97, 0.99]
Hours spent on the street	1.02	[1.00, 1.05]
Ever deported from the United States	1.02	[0.73, 1.42]
Drug-using behaviors Age at first injection drug use	0.96	[0.92, 0.99]
Past-6-month injection drug use Methamphetamine	1.62	[1.18, 2.21]
Received injection assistance by hit doctor in the past 6 months	1.56	[1.03, 2.36]
Economic factors
Average monthly income ≥3,500 pesos	2.21	[1.52, 3.22]

Note: CI = confidence interval.

Discussion

We found that more than half of PWID in our sample (56%) had experienced police victimization in the past 6 months. Consistent with our hypothesis, we identified structural/environmental factors and drug-using behaviors independently associated with increased odds of experiencing recent police victimization that may serve as important visual markers to quickly distinguish and target PWID in Tijuana. Below, we examine the relevance of these factors within the context of our study framework.

Within the structural/environmental domain, PWID with fewer years lived in Tijuana, an indicator of being a recent migrant, were significantly more likely to experience recent victimization by police. Tijuana’s proximity to the United States and its economic opportunities have attracted internal migrants from poorer regions of Mexico and Central America; more than half (∼52%) of Tijuana’s residents are migrants (i.e., born outside the state of Baja California) (Fussell, 2004; Instituto Nacional de Migracion, 2013). Tijuana also has a prominent population of migrants, primarily men, who have been deported from the United States and have resettled in Tijuana (Pinedo et al., 2014b, 2015); the city received ∼316,000 deported migrants between 2010 and 2013 (Instituto Nacional de Migracion, 2011, 2012, 2013). Migrants are highly stigmatized in Tijuana, where they are perceived by the community as the cause of prevailing social problems, including increasing rates of crime and drug use (Infante et al., 2012; Pinedo et al., 2014a). Sociocultural and visual differences (e.g., clothes, speech) may quickly distinguish migrants from locals (Infante et al., 2012; Pinedo et al., 2014a). Recent migrants who are PWID may be less acclimated to the local culture and environment, thus quickly identifiable and vulnerable to police victimization.

We identified three factors related to drug-using behaviors that were positively associated with experiencing recent police victimization: younger age at first injection drug use, recent (i.e., past-6-month) injection of methamphetamine, and recently seeking injection assistance from a hit doctor. Injecting drugs at an earlier age suggests a longer history of drug injection use and greater risk exposure to infections. Some participants in our study had been injecting drugs for more than 20 years. Repeated injection of drugs can cause the skin to scar, become discolored, and produce lesions (“track marks”) along the length of the vein (Cagle et al., 2002). Past studies have linked track marks with increased arrests and adverse experiences with local police (Miller et al., 2008; Pollini et al., 2010; Robertson et al., 2010; Strathdee et al., 2005, 2008b). Recent injection of methamphetamine significantly increased PWID’s odds of experiencing police victimization. Rates of methamphetamine use have been increasing along the U.S. border region, and use of this drug is especially pervasive in Tijuana compared with other border cities (Brouwer et al., 2009; Case et al., 2008; Volkmann, et al., 2011). The use of methamphetamine can have drastic physical effects on individuals, such as extreme weight loss, tooth decay (i.e., “meth mouth”), and skin sores/lesions (National Institute on Drug Abuse, 2014; Shaner et al., 2006). Mental (i.e., paranoia, anxiety, hallucinations) and behavioral (i.e., aggressiveness) symptoms are also common (National Institute on Drug Abuse, 2014). Physical outcomes of injecting drugs and methamphetamine use can serve as visual indicators for police to rapidly identify and target drug users and can contribute to increased susceptibility to maltreatment.

Seeking injection assistance from a hit doctor in the past 6 months was positively associated with police victimization. Hit doctors often operate out of shooting galleries (i.e., abandoned buildings, alleyways, hidden rooms) (Ball et al., 1998; Ouellet et al., 1991; Robertson et al., 2010; Strathdee et al., 2005). Among PWID in Tijuana, an independent association between using a hit doctor and being previously arrested for carrying unused syringes has been documented (Robertson et al., 2010). PWID may choose to seek hit doctors and inject in shooting galleries to avoid police (Philbin et al., 2008; Robertson et al., 2010; Volkmann et al., 2011). However, emerging research suggests that police in Tijuana routinely target local public spaces known to be areas where drug users live or congregate (Gaines et al., 2014; Volkmann et al., 2011). Drugs users in Tijuana are highly vulnerable and easy targets for local police seeking to meet arrest quotas. Law enforcement officers make frequent visits to the Tijuana River Canal, the Red Light District, and needle exchange programs, where drug users can easily be found (Gaines et al., 2014; Philbin et al., 2008; Pollini et al., 2008; Volkmann et al., 2011). More than 200 shooting galleries are estimated to be operating in Tijuana (Case et al., 2008; Strathdee et al., 2005). Close monitoring by police of shooting galleries throughout the city may occur. PWID seeking hit doctors for assistance with injections may increase their likelihood of coming in contact with local law enforcement and ensuing victimization.

Some limitations should be considered when evaluating our findings. Participants were recruited via nonrandom sampling strategies; findings may not be generalizable to all PWID in Tijuana or elsewhere. Cross-sectional data do not allow us to make causal inferences between findings. Given that information regarding experiences with police and drug use behaviors can be sensitive, experiences and some activities may have been underreported. Recall bias may be present in our analyses because data relied on selfreported measures that asked participants to recall behaviors and experiences with police. Participants received a financial incentive for their participation, which may have influenced their responses. Important efforts to reduce bias were used, such as using highly trained interviewers who are comfortable interacting with participants in a nonjudgmental manner to increase rapport (Robertson et al., 2014). Measures for police victimization referred to a past-6-month timeframe; shorter timeframes may produce more accurate and reliable responses (Napper et al., 2010). Lastly, we lacked contextual data on police victimization events and PWID’s participation in crime-related activities, which would have enhanced our understanding of police victimization in Tijuana.

This study yields new information regarding factors associated with increased vulnerability to police victimization among a high-risk population of PWID in Tijuana, Mexico. Understanding factors that increase susceptibility to police victimization is imperative to designing effective interventions aimed at mitigating HIV risk in this region. Problematic policing practices are an important environmental driver that may significantly increase the risk of acquiring HIV among PWID in Tijuana (Beletsky et al., 2013; Pollini et al., 2008; Robertson et al., 2010; Volkmann et al., 2011). Independent associations between increased numbers of arrests and being HIV positive have been documented in Tijuana (Strathdee et al., 2008a, 2008b). Monitoring police and reduction of abusive law enforcement policing practices may help reduce future HIV incidence among PWID (Strathdee et al., 2010). Such interventions should consider crucial physical markers that may increase susceptibility to police targeting and victimization, as well as migration status, drugusing behaviors, and types of drugs. Increased partnerships and collaboration between public health agencies and police may help align law enforcement with HIV prevention efforts (Beletsky et al., 2011a; DeBeck et al., 2008; Kerr et al., 2005). Police education programs that incorporate topics of harm reduction and HIV prevention are currently underway in Tijuana (Scholl & Nicholson, 2010). Additional evaluation of Mexican/Tijuana’s police training programs is needed to determine whether this is an effective strategy to mitigate HIV risk among PWID in Tijuana.