Abstract
Physical and psychological violence that is structurally mediated by the system of law enforcement results in deaths, injuries, trauma, and stress that disproportionately affect marginalized populations (e.g., people of color; immigrants; individuals experiencing houselessness; people with disabilities; the lesbian, gay, bisexual, transgender, and queer [LGBTQ] community; individuals with mental illness; people who use drugs; and sex workers). Among other factors, misuse of policies intended to protect law enforcement agencies has enabled limited accountability for these harms. Furthermore, certain regulations (e.g., anti-immigrant legislation, policies associated with the war on drugs, and criminalization of sex work and activities associated with houselessness) have promoted and intensified violence by law enforcement toward marginalized populations. While interventions for improving policing quality to reduce violence (e.g., community-oriented policing, training, body/dashboard-mounted cameras, and conducted electrical weapons) have been implemented, empirical evidence suggests notable limitations. Importantly, these approaches also lack an upstream, primary prevention public health frame. A public health strategy that centers community safety and prevents law enforcement violence should favor community-built and community-based solutions. APHA recommends the following actions by federal, state, tribal, and local authorities: (1) eliminate policies and practices that facilitate disproportionate violence against specific populations (including laws criminalizing these populations), (2) institute robust law enforcement accountability measures, (3) increase investment in promoting racial and economic equity to address social determinants of health, (4) implement community-based alternatives to addressing harms and preventing trauma, and (5) work with public health officials to comprehensively document law enforcement contact, violence, and injuries.
Relationship to Existing APHA Policy Statements
Problem Statement
Prevalence, impacts, and inequities: Law enforcement violence is a critical public health issue. Consistent with domains of violence defined by the World Health Organization (WHO), law enforcement violence has been conceptualized to include physical, psychological, and sexual violence as well as neglect (i.e., failure to aid).[1–3] While all forms of violence are important to consider and have been shown to correlate with poor mental health outcomes in at least one study,[1] this statement focuses on physical and psychological violence.
According to The Counted (a United Kingdom–based Web site that operated from 2015 to 2016 and was the most timely, comprehensive source of U.S. data at the time),[4–6] at least 1,091 individuals were killed by law enforcement officers in the United States in 2016.[7] These deaths amounted to 54,754 years of life lost.[8] Based on data from the Centers for Disease Control and Prevention (CDC), there were 76,440 nonfatal injuries due to legal intervention in 2016.[9] At least 28 serious injuries were inflicted on students between 2010 and 2015 by school-based law enforcement officers.[10] The CDC estimates that the overall cost of fatal and nonfatal injuries by law enforcement reported in 2010, including medical costs and work lost, was $1.8 billion.[11] Legal scholars describe a clear connection between increased exposure to stops and an elevated risk of death or physical harm by law enforcement officers.[12]
Inappropriate stops by law enforcement are one form of psychological violence with serious implications for public health.[1,2] Even in the absence of physical violence, several studies have shown that stops perceived as unfair, discriminatory, or intrusive are associated with adverse mental health outcomes, including symptoms of anxiety, depression, and posttraumatic stress disorder.[1,13,14] In addition, one study revealed that neighborhood-level frisks and use of force were linked to elevated levels of psychological distress among men living in these neighborhoods.[15] In two large surveys, Black individuals were more likely than White individuals to report stress as a result of encounters with police[13,14]—a concern given evidence of an association between stress due to perceived racial discrimination and risk factors for chronic disease and early mortality.[16] A nationally representative study showed an association between deaths among Black individuals due to legal intervention and subsequent poor mental health among Black adults living in the same state.[17]
The impacts of physical violence likewise extend beyond injuries and death, affecting individuals’ and communities’ ability to achieve positive health outcomes in the short and long term and compounding extant health inequities. For example, one study revealed that residents of neighborhoods with high rates of law enforcement use of force were at increased risk for diabetes and obesity.[18] Among youths, exposure to violence from school-based law enforcement officers has been linked to “denial of educational and social growth”[19]—both key determinants of health [20]—and ethnographic research indicates that current policing practices alter key developmental processes among Black male adolescents.[21] In summary, aggressive policing is “a threat to physical and mental health” that may be exacerbated among marginalized populations.[13]
Marginalized populations are inequitably affected by law enforcement action and violence. People of color accounted for more than 50% of years of life lost due to legal intervention in 2016 but account for just under 40% of the U.S. population.[8] Native Americans have been killed by law enforcement at a higher per capita rate than any other group in the United States (3.5 times higher than White Americans), with these mortality data likely to be an undercount.[7,22] In 2016, Black and Native American individuals were more than two and three times (respectively) as likely to be killed by law enforcement as White individuals.[7] Stratification by gender and age showed that male Blacks and Native Americans 15 to 34 years of age were nine and six times (respectively) more likely to be killed than other Americans in their age group.[7] Similarly, Black women are disproportionately represented among women killed by police.[23] Black and Latino individuals are more likely to be stopped and arrested and to experience nonfatal violence by law enforcement.[1,24–27] Of the 4,400 individuals shot by officers from the 50 largest police departments from 2010 to 2016, 55% were Black, more than double the proportion of the Black population in these departments’ jurisdictions.[28] In 2012, Black and Native American individuals were admitted to emergency departments for injuries due to legal intervention at proportions three and six times (respectively) their representation in the general population,[29] and in a nationally representative sample of emergency departments during 2001 to 2014, Black individuals 15 to 34 years old were treated for legal intervention injuries at almost five times the rate of their White counterparts [30]. Students most at risk for violence by school-based law enforcement officers include children with disabilities, students of color, and poor students.[31]
Other marginalized populations also experience inequitable exposure to law enforcement violence. Among recorded U.S. deaths attributed to law enforcement in 2015, an estimated 27% involved individuals with mental illness.[32] Other groups highly affected by law enforcement violence include people who identify as transgender, lesbian, gay, and/or bisexual[1,33]; individuals experiencing houselessness[34]; low-income individuals[1,35]; sex workers[36,37]; and people who use drugs.[2] Women also experience sexual violence by police officers, particularly women of color. In a 2003 study in New York City, 38% of Black women, 39% of Latina women, and 13% of Asian or Pacific Islander women reported being sexually harassed by police officers.[38] Immigrant communities are subject to policing from local, state, and federal immigration authorities such as the Department of Homeland Security’s Bureau of Immigration and Customs Enforcement. Immigration raids result in “immigration enforcement stress” and fear of interacting with government agents and informal social networks.[26] Policies that increase law enforcement contact or fear of contact create barriers to health care and other health-supportive services (e.g., Medicaid, harm reduction programs, and domestic violence services) for undocumented individuals and their U.S. citizen family members.[39–43] The disproportionate impact of policing on these communities has been documented since at least the 1960s.[44]<
Insufficient monitoring and surveillance of law enforcement violence: The data presented above likely underestimate the magnitude of law enforcement violence given that comprehensive information on deaths, mental and physical injuries, and frequency of encounters is limited (e.g., there are no systematic public health data on sexual assaults committed by police).[45] While the Federal Bureau of Investigation’s Uniform Crime Reporting System and the CDC’s National Violent Death Report System (NVDRS) generate some data on injuries and fatalities by law enforcement, they neglect indicators vital to understanding the magnitude and scope of the issue, such as type of injury, deaths on federal property (e.g., federal prisons, tribal lands, military bases), and types of law enforcement officers involved.[46–48] Most concerning, reporting occurs on a voluntary basis. As a result, even the NVDRS—the most reliable of the official reporting systems[48]—notably underestimates deaths by law enforcement.[6] The U.S. National Vital Statistics System failed to capture 55% of such deaths in 2015 due to misclassification.[4] The magnitude of disparities in violence committed by school-based law enforcement officers is likely underestimated as well, given communication challenges and unreliable mechanisms for reporting abuse .[31] Given this situation, public health practitioners and researchers must rely on nongovernmental, Web-based social media data sources such as The Counted, which captured 93% of deaths by law enforcement in 2015.[4,5] Yet, it is feasible to gather reliable, real-time data on law enforcement-related deaths via existing public health reporting mechanisms.[6]
Policing as a mechanism of social control that exacerbates social inequity: The ecosocial theory of disease distribution holds that to meaningfully analyze and interpret the population distribution of a health exposure, a grounding in the historical context from which the exposure emerged is necessary.[49] Namely, U.S. policing was historically deployed for the social control of communities deemed socially marginal (i.e., in the 19th century, it evolved from ruling-class efforts to control the immigrant working class in the North and slave patrols in the South).[50]
Policies and practices continue to implement and sustain this historical intent. For example, the war on drugs assigned drug use intervention to law enforcement in lieu of formulating a public health approach. Scholars suggest that the associated “tough on crime” rhetoric was a racially coded appeal to White populations across class lines aimed at legitimating targeted policing in communities of color.[51,52] By encouraging drug arrests with cash incentives, loosening restrictions on searches, and creating a culture that encouraged law enforcement to repeatedly stop and search people of color without reasonable cause, the federal government disproportionately subjected marginalized communities to increased contact with the law enforcement system.[51] Data-driven policing is another example of a structural and targeted policing practice that links crime to place and race and facilitates increased contact with law enforcement among marginalized communities.[53,54]<
Policies and practices that facilitate a system of discriminatory policing are particularly problematic given the weakening of the Posse Comitatus Act, the enactment of the National Defense Authorization Act, and the 1033 program, which distribute surplus military equipment to local and state law enforcement agencies.[24,55,56] Delivery of military equipment to law enforcement agencies precipitates military-style training, allows military weapons to become the tools of law enforcement, and increases the use of special weapons and tactics (SWAT), resulting in increasing rates of use of force and extrajudicial murders by law enforcement—disproportionately among marginalized communities.[24,57] The observed militarization and extensive purview of domestic law enforcement are facilitated by mounting investments of federal funds in police departments and financial enticements.[51]
Research on predictors of police force size has indicated that the system of law enforcement upholds existing racial and class hierarchies by targeting socially marginalized groups, often low-income communities of color. Key predictors maintaining an association with police force size after control for crime rates include the size and growth of populations of color, racial economic inequality, and poverty.[58,59] Such findings suggest that these populations are perceived as a threat to the social order and that policing is used as a mechanism of control.[58,60] Upholding social hierarchies perpetuates and exacerbates adverse health outcomes among those who are already disproportionately affected by inequities in key social determinants of health, or those underlying factors that “affect a wide range of health, functioning, and quality-of-life outcomes and risks” and are widely understood in the field of public health to be the primary contributors to persistent health inequities.[61] These factors include access to education and economic opportunities, perceptions of public safety and exposure to violence, quality of housing and transportation, social norms and attitudes (e.g., discrimination, racism, and distrust of government), and availability of community-based resources.[20,61]
Ineffective response to social problems: The concentration of policing in socially marginalized communities—and the associated public health threats—stems from a framework that crime originates from inherently “bad” individuals and communities, or a “thin blue line” ideology.[44,50,60,62] Yet, the social determinants of health framework indicates that efforts to promote physical, mental, economic, and social well-being are more effective if premised on an assessment of the social conditions underlying the behaviors that are typically addressed through the criminal justice system. With this framework, the range of interpersonal harms and behaviors deemed “criminal” can be understood from a social determinants of health perspective as emerging from social inequities. Theft, as just one example, can be understood as a behavior to meet material survival needs in the context of poverty due to long-standing, systematic economic disinvestment from low-income communities of color, and intra-community violence has been shown to be linked to the chronic stress of poverty.[61]
Criminalization of houselessness, sex work, and drug abuse exemplifies how law enforcement is deployed to rectify social inequities.[34] However, laws that criminalize houselessness (e.g., local and state laws prohibiting loitering and sleeping in public spaces) are costly to enforce, perpetuate houselessness, and violate basic human rights, among other harms to public health.[63,64] According to the National Law Center on Homelessness and Poverty, criminalizing behaviors associated with houselessness violates the United Nations Convention Against Torture, and it recommends that federal agencies take active steps toward decriminalization while funding constructive alternatives.[65] Police officers have also indicated that criminalization of houselessness is an ineffective response to the root cause and that responsibility for addressing houselessness should reside outside of law enforcement’s purview.[66] Criminalization of sex work likewise results in high rates of law enforcement violence toward sex workers and those assumed to be sex workers, such as transgender women of color.[33] Similarly, punitive strategies of addressing drug abuse show little evidence of reducing substance abuse and have proven harmful to working-class communities of color.[67]
Although the need to invest in addressing the social determinants of health is clear, government spending on social services such as housing assistance and education has decreased since the 1980s. The Center for Budget and Policy Priorities documents a median budget reduction of 26% among 11 of the 13 largest health, housing, and social service block-grant programs between their inception in the 1980s and 2016 and a $13 billion reduction in these funding streams between 2000 and 2016.[68] Yet, spending on policing increased 445% between 1982 and 2007, including a 729% increase in federal funding[34] The Center for Population Democracy found that, in nine of the 10 cities it examined, more than one quarter of general funds were committed to local police departments. For instance, in Oakland, California, 41% of the general fund went to the police department, which had a 19% budget increase between 2013 and 2017 while total city expenditures increased by just under 8%.[3]
Barriers to accountability and reform: Between 2005 and 2011, only 47 police officers across the United States were charged by prosecutors with a crime for their involvement in civilian deaths, with 11 of these 47 individuals convicted.[69] Multiple barriers impede accountability and obstruct meaningful reform. Cultural barriers such as efforts to “protect one’s own” can manifest in a “code of silence,” or a norm of not reporting other officers’ misconduct and protecting them during investigations.[26,70,71]
Laws and policies such as state-based police bills of rights (generally referred to as law enforcement officers’ bills of rights, or LEOBORs) and police union contracts provide law enforcement officers accused of excessive use of force or murder with protections from investigation and disciplinary action, known as “super due process.”[72,73] including suppression of law enforcement data related to deaths.[74] LEOBORs are found in 14 states and first emerged in the 1970s, when law enforcement officers pursued unionizing efforts in reaction to grassroots mobilizations demanding democratic accountability and transparency over police (e.g., civilian review boards) given experiences of officer misconduct, corruption, and brutality.[75,76] LEOBOR provisions can generally be broken into two categories: those that should be eliminated due to their ability to hinder efforts to hold law enforcement officers accountable (e.g., investigative delays)[76,77] and some protections that should be extended to everyone, including civilians suspected of a crime (e.g., limits on the duration of interrogations).[78] Rights and protections present in some LEOBORs that protect law enforcement officers from merited accountability include the following: unreasonable limitations on reporting time that disqualify civilian complaints, restriction of interrogation of officers to other sworn officers, preventing civilian investigators from interviewing or investigating officers, and restrictions of public access to disciplinary records.[76] In addition, investigative delays, coupled with notifications of who will interrogate an officer and unrestricted access to all of the evidence brought against an officer, allow officers to prepare the most exculpatory and/or least inculpatory narrative.[75–77]
Structural racism embedded within “legal, social, and political systems…enable[s] police officers to disproportionately stop people of color, often without cause…with greater use of force [and] without any repercussions.”[79] Protective laws and policies, obstruction from oversight, and cultural norms inhibit accountability, confound reform, and lead to harm, especially among marginalized communities.
Evidence-Based Strategies to Address the Problem
Improving surveillance and reporting of law enforcement violence: Improvements to existing public health monitoring systems, such as expanding the NVDRS to include all states and moving to more timely processing and release of data at the local level—not just the state level—could prove highly effective.[6,48,80] To leverage the success of The Counted in capturing and classifying deaths by law enforcement, state and local public health agencies could collect additional data beyond what are typically reported by using validated, existing social media sources. In addition to these data already being publicly available, they capture real-time reports that include data on age, gender, race/ethnicity, and census tracts of residence and death, and they serve to correct misclassification in vital statistics.[4,5] With regard to reporting, transparency can help identify appropriate policy and programmatic interventions; evidence indicates the success of transparency measures such as making health inequities visible by presenting data stratified in relation to categories of race/ethnicity, nativity, gender identity, sexual identity, and socioeconomic position; including housing tenure (as a proxy for houselessness); and including type of law enforcement official, mechanism of death (e.g., firearm, Taser, chokehold), and locale of death (e.g., on the street, in the decedent’s home, at a school, at a border crossing).[3,6,81] Furthermore, a mechanism for state and local public health agencies to share data with various entities can encourage appropriate prevention and intervention measures, such as sharing with state attorney generals for further investigation.[82]
Decriminalization of activities shaped by the experience of marginalization: As criminal justice scholars have argued, mass criminalization is a key mechanism through which communities of color experience heightened rates of law enforcement violence.[12] Others have concluded that disparities in contact with law enforcement may be a root cause of differential exposure to physical violence by law enforcement and that “reducing inequality in police stops can simultaneously reduce inequalities in exposure to violence.”[1] Therefore, a critical step in reducing structurally mediated physiological and psychological violence by law enforcement is to repeal laws that promote or justify increased scrutiny of specific populations. Such laws include those relating to drug use or possession, sex work, houselessness, and immigration. By removing justification for law enforcement intervention, this will reduce encounters between law enforcement officers and individuals whose activities are presently criminalized. Crimes should not simply be downgraded to lower-level offenses; for example, research shows that marijuana-related arrest rates remained stable or increased when possession was reclassified as a lesser offense but was still considered against the law.[83] By contrast, in Massachusetts courts ruled to limit police enforcement of marijuana possession, and arrests fell by 86%.[83] Not only can drug decriminalization reduce arrests and incarceration, it also has the public health benefit of increasing uptake of drug treatment, with cost savings due to redirecting resources from criminal justice to the health system.[84] Regarding sex work, one meta-synthesis of qualitative studies concluded that New Zealand’s full decriminalization of sex work was associated with reductions in law enforcement contact and improvements in HIV prevention among sex workers.[85] These findings may be generalizable to the U.S. context and serve as a model for structural intervention. Decriminalization is consistent with the WHO recommendations for structural interventions that address social determinants of health for marginalized groups.[61]
Under certain legislation, criminalization extends to protesting and mass mobilizations, which are vital means by which marginalized communities voice concerns. In 2017, several states passed anti-protest legislation; among them were North Dakota and South Dakota, where, in 2016, protestors against the Dakota Access Pipeline at the Standing Rock Indian Reservation—including many Native Americans—were met with violent force by local law enforcement and the North Dakota National Guard, leading the United Nations to declare human rights violations.[86,87] Advocating against such laws is critical to protect free speech and human rights and to reduce unnecessary contact with law enforcement.
Reallocation of funds from policing to the social determinants of health: As described above, policing reproduces inequitable social and economic conditions that precipitate intervention by law enforcement. This places both law enforcement officers and marginalized community members at risk of injury, death, and adverse health outcomes. By contrast, a public health approach targets the structural inequities that manifest in criminalized behaviors by addressing the social determinants of health[88,89] This type of approach includes increasing access to housing, expanding educational and employment opportunities, increasing access to mental health and substance use treatment, and restoring a sense of safety by addressing interpersonal and institutional factors contributing to perceptions of safety and experiences of discrimination.[61] The social determinants of health approach is associated with reduced community trauma and interpersonal harm and improved community health and safety,[88] and it is the basis of the CDC’s recommendations for data-driven, community-level, prevention-focused interventions.[90] This approach is a key element of the Movement for Black Lives platform, a policy agenda that calls for “reallocation of funds at the federal, state, and local level from policing and incarceration…to long-term safety strategies such as education, local restorative justice services, and employment programs.”[91]
Evidence demonstrates the benefits of shifting from criminalization to a framework grounded in social determinants and primary prevention. For example, there is a well-established link between improving educational attainment and positive employment and socioeconomic outcomes and subsequent positive short- and long-term health outcomes.[92] More evidence is found in houselessness services. The U.S. Interagency Council on Homelessness recommends providing permanent housing as a proven approach to improve health among those experiencing houselessness, as such efforts have been associated with higher housing retention rates, reductions in use of crisis services and institutions, and improvements in health and social outcomes[93] and have been cost effective.[94,95] Similarly, because exposure to violence is a critical determinant of health and can lead to further violence by trauma survivors and later contact with law enforcement,[96] “trauma-informed” approaches to care and policy are recommended across sectors.[97] Reinvestment in community resources can also occur in tertiary prevention by using a health model for crisis response. For example, health workers in Oakland are training community members to respond to mental health crises and suspected overdoses in ways that minimize law enforcement involvement.[87]<
The above evidence, combined with decreasing crime rates,[34] suggests that funds disproportionately allocated to policing could be more effectively invested in social services to improve health, particularly in communities where historically rooted endemic disinvestment has negatively contributed to health disparities.
Strategies to ensure community safety without reliance on armed law enforcement: Although greater social and economic equity is likely to lead to higher quality of life for marginalized communities, interpersonal harm will still exist, and strategies to ensure community safety will still be necessary. Alternative approaches can improve public safety without the harms associated with the system of policing. For instance, community-based violence intervention programs that detect and interrupt potentially violent conflicts, identify and address high-risk situations, and mobilize the community to change norms have significantly reduced homicides and nonfatal shootings in urban neighborhoods with the highest numbers of incidents.[98] These programs have had success employing violence interrupters and culturally appropriate unarmed street outreach workers; these interrupters have been able to defuse potentially harmful or violent situations with no, or minimal, intervention by police.[98]
Similarly, restorative justice is a nonpunitive approach to resolving interpersonal harm through dialogue among perpetrators, victims, and others affected without reliance on law enforcement. Its implementation in school settings has been associated with reduced suspensions, expulsions, and referrals to law enforcement.[99] Future programs might increase efficacy by ensuring that the populations most affected by law enforcement violence lead program design and implementation, which is a widely acknowledged best practice.[100]
Opposing Arguments and Evidence
Arguments against reducing law enforcement presence and ensuring accountability as mechanisms to address law enforcement violence assert that these strategies will increase crime, decrease public safety, and harm public health. Others propose to address law enforcement violence through tactics such as community-oriented policing, use of body-mounted cameras and Tasers, and increases in officer training. This section presents these arguments along with research that suggests the former strategies are aligned with a public health approach and have a negligible impact on increasing crime or decreasing public safety, while the latter tactics do not address the structural predictors of law enforcement violence or its health implications..
Decriminalization harms the public’s health: Proposals to decriminalize drug possession and sex work are often met with concern that doing so will negatively affect the public’s health. For example, opponents suggest that decriminalization of drugs leads to an increase in drug use and higher rates of traffic accidents. Initial research on decriminalization has yielded mixed findings.[101,102] and studies show that the legitimate concern about negative health effects of drug use is better addressed with health service approaches. Data from Portugal, which decriminalized all drug use in 2001, as compared with Spain and Italy—which maintained criminal penalties for drug use—showed increased uptake of drug treatment, reductions in opiate-related deaths and infectious diseases, and increases in the quantity of drugs seized by the authorities due to shifting law enforcement resources from minor possession crimes to a focus on traffickers.[103] Many organizations support drug decriminalization to improve human rights and public health, such as the Office of the United Nations
