Volume 38, Issue 1 p. 6-23
Open Access

Naturalizing unnatural death in Los Angeles County jails

Nicholas Shapiro

Corresponding Author

Nicholas Shapiro

Carceral Ecologies, Institute for Society and Genetics, UCLA, Los Angeles, California, USA


Nicholas Shapiro, Carceral Ecologies, Institute for Society and Genetics, UCLA, Los Angeles, California, USA.

Email: [email protected]

Terence Keel, BioCritical Studies Lab, Department of African American Studies, Institute for Society and Genetics, UCLA, Los Angeles, California, USA.

Email: [email protected]

Search for more papers by this author
Terence Keel

Corresponding Author

Terence Keel

BioCritical Studies Lab, Department of African American Studies, Institute for Society and Genetics, UCLA, Los Angeles, California, USA


Nicholas Shapiro, Carceral Ecologies, Institute for Society and Genetics, UCLA, Los Angeles, California, USA.

Email: [email protected]

Terence Keel, BioCritical Studies Lab, Department of African American Studies, Institute for Society and Genetics, UCLA, Los Angeles, California, USA.

Email: [email protected]

Search for more papers by this author
First published: 18 October 2023
Citations: 2


In this paper we use quantitative and qualitative methods to examine how death investigations in Los Angeles County jails disproportionately naturalize death among Black and Latino incarcerated people. Our study is based on an assessment of 58 autopsies, coroner investigator narratives, and toxicology reports produced between 2009 and 2018. We found that the Medical Examiner frequently arrived at natural or undetermined death determinations that minimized the culpability of carceral staff for loss of life that occurred within county jail. In our dataset, Black people were disproportionately classified as natural. Undetermined deaths were almost exclusively Latino. More than 75% of the cases in our study were deaths that occurred before standing trial. Our findings reveal how biomedical knowledge about incarcerated Black and Latino people is used to erase the life-diminishing effects of punishment, neglect, and maltreatment that are central to the project of mass incarceration.


In the early 1980s, Los Angeles Police Department Chief Darryl Gates attempted to naturalize the killings of young Black men by police using chokeholds. Gates claimed that “We may be finding in some blacks that when [the chokehold] is applied, the veins or arteries do not open up as fast as on normal people” (Woo & Malnic, 2014). This difference, he believed, made Black people more susceptible to death at the hands of police than other races (Wynter, 1994). Gates was attempting to recruit medicine and science to provide a biological explanation for the violent, racist, and unlawful actions of police. Some 40 years later, the deaths of Black men that occur in custody of law enforcement in Los Angeles continue to be naturalized, albeit through more insidious forms of medical racism and structural violence.

The Los Angeles County Jail system is the largest in the United States and leads the nation in in-custody deaths (Smith, 2020). Without public witnesses, the circumstances of death for wards of the state are almost exclusively narrated by their captors. Medical examiners and coroners are the primary parties charged with the task of determining the manner and causes of deaths in custody. This paper assesses the production of death and subsequent bio-legal naturalization of lives lost inside the Los Angeles County Jail. We argue that the Los Angeles County Department of Medical Examiner-Coroner (DME-C) has produced death determinations that naturalize the structural and racial violence advanced by law enforcement and carceral health services. Our study is based on an assessment of 58 autopsies, coroner investigator narratives, and toxicology reports written between 2009 and 2018.

In some instances, we documented evidence of physical violence on the bodies of individuals who died from purported natural causes. In other cases, deputies and carceral staff appear to have neglected the medical or psychiatric needs of incarcerated people. Deputies also placed the incarcerated in psychological and spatial contexts that exacerbated pre-existing mental health conditions. Black and Latina/o people died from purportedly natural or undetermined causes at the highest rates for all races/ethnicities (see Table 1). The predominance of Black death is particularly disconcerting, because while only 8% of the county is Black, Black people make up 29% of the jailed population (Vera, 2022) and 39% of the autopsies supplied to us by the County. Black and Latina/o people also suffered most from physical violence and potential medical neglect. In the cases we reviewed, the medical examiner-coroner deployed a series of biomedical inferences that minimized or ignored the lethal actions of deputies and the punitive culture that shapes the carceral ecosystem. The end result is the naturalization of unnatural death, where the health histories of the deceased are not consulted, and racialized biomedical imaginaries supplement official accounts of death, making the incarcerated responsible for the ends of their own lives.

TABLE 1. Manner of death, demographics, custody status, and race population by race/ethnicity
Black Latino/a White Asian Total Percentage of total reviewed Deaths
Manner of death Count + percent of reviewed deaths for each attribute
Natural 16 64% 6 24% 2 8% 1 4% 25 43%
Undetermined 3 38% 5 63% 0 0% 0 0% 8 14%
Suicide 1 8% 4 33% 6 50% 1 8% 12 21%
Accident 0 0% 4 50% 4 50% 0 0% 8 14%
Homicide 2 40% 0 0% 2 40% 1 20% 5 9%
Male 19 37% 18 35% 12 23% 3 6% 52 90%
Female 3 50% 1 17% 2 33% 0 0% 6 10%
Age (Mean) 42 mean -1yr 40 mean -3yr 45 mean + 2yr 58 mean + 15yr 43
Pre-trial 18 42% 13 30% 9 21% 3 7% 43 74%
County Population (Percent) 9% 49% 25% 16.0% 10,014,009
Jail Population (2020-01-02) 4,872 29% 8,703 52% 2,540 15% 75 <1% 16,791
Total 22 38% 19 33% 14 24% 3 5% 58
  • Note: Population data (Vera, 2022) are from 2020 due to access and formatting changes in jail records.

We situate our paper within the scholarly literature that examines how state institutions govern the lives and deaths of civilians. Within this literature, science and biomedicine have been associated with biopolitical governance seeking to regulate bodies and populations to optimize human-produced value for industry and the state. The scientific authority to control what counts as natural or abnormal (Georges Canguilhem, 1991 [1966]) is a power that has been found to medicalize social problems, reduce the freedom of individuals to manage their own health, and pathologize the bodies of marginal groups (Conrad, 1992). This medical power of naturalization is also associated with the biopolitical calculus of “making live” or “letting die” (Foucault, 2004 [1974-1975]). Scholars working in Black studies and/or with a critical race perspective have explained that racism has always been integral to biopolitical governance, with medicine and science as trusted tools (Braun, 2014; Eneanya et al., 2019; Gravlee, 2009; Jackson, 2020; Mbembe, 2019; Obasogie et al., 2017; Warren, 2018; Weheliye, 2014).

Our study shows that the carceral environment creates a life-denying venue—or what Mbembe calls “death-worlds” (Mbembe, 2019)—for multiple projects of violence, death, and population control to coexist with one another. The coexistence of many traditions of racial and economic violence in carceral spaces makes it difficult to hold any single socio-political entity responsible. Biomedical alibis—in the form of autopsies—camouflage the death-driving relations of carceral enmity, making necropower appear merely biopolitical. In our assessment, jail provides a “death world” where race-specific beliefs and technologies of violence historically formed against one group can be used, often very effectively, against any population. Violence in carceral spaces, one might say, is quintessentially liberal in the sense that its distribution is universal but its practice carries racial histories often beyond the awareness of its perpetrators.

Relatedly, our study sheds light on the role of law enforcement in the deaths of civilians and wards of the state. Police, volunteer militias, border patrol, and sanctioned gangs have been used throughout the history of Los Angeles to control who lives and dies throughout the county. Indeed, Los Angeles County jails were born out of anti-Indigenous and anti-Mexican campaigns of violence, genocide, and forced assimilation that were integral to the founding of Los Angeles, first under Spanish rule and later under Anglo-American governance (Bauer, 2016; Deverell, 2004; Lytle Hernandez, 2017; Madley, 2017). Consistent with California's history of using power beyond the letter-of-the-law to govern and terrorize those on its territory (Gonzales-Day, 2006; Lytle Hernandez, 2010; Carrigan & Webb, 2017), Sheriff gangs continue to exert control within these facilities (Baird et al., 2012; Castle, 2021). Deputy gangs inside Los Angeles County jails and the nearly unrestricted power of the County Sheriff seek “legitimation by blurring the relations between violence, murder, and the law,” producing “topographies of cruelty” and racialized terror (Mbembe, 2019; Rouse, 2021).

Our study also demonstrates how medicine participates in this topography of cruelty by producing biological justifications for deaths produced inside one of the most lethal jail systems in the world. As Lesly Sharp has underlined in assessing the inescapable losses that characterize life behind bars, carceral deaths are not simply deaths in jail or prison, but deaths by incarceration (2022). We read this insight, together with local histories of jailing (Lytle Hernandez, 2017), to illustrate how the racialized production of death in jail is a feature and not a bug in the carceral project of surveilling, policing, warehousing, and eliminating racial outsiders. Like police documentation that “enables and is part of policing” (Yu & Monas, 2020, 45), jail death investigation records written by medical examiners advance the goal of minimizing jailer liability. The medical reports that serve as the empirical basis of this article enable and are part of carceral power. In a setting of liberal violence, dying in jail becomes a natural process because the socio-political forces enabling these deaths are erased by medical examiners or obviated from their biomedical gaze. Jail death autopsies complete centuries of anti-Mexican, anti-Black, and anti-poor campaigns launched in Los Angeles and California more broadly. The omissions of the jail autopsy are part of a longer project of naturalizing death by law enforcement in the so-called “progressive” state of California.

We find, however, that the autopsy is a precarious form of biomedical knowledge capable of working against the state. In the hands of impacted communities and their allies, death investigation records can be used to resist necropolitics as they capture some of the effects of violence and medical neglect that incriminate deputies and jail staff. Such details, when read with a critical eye, allow communities to attribute responsibility to law enforcement and the county for lives taken in captivity. Families that read the state sanctioned autopsy looking for what has been omitted, overstated, or conjured can elicit from the bodies of their loved-ones alternative biomedical knowledge that can expose unlawful death and injustice. Too often, the academy is skeptical of community-based knowledge, seeing it as either biased, irrational, or mere anecdotes to be subsumed under the more authoritative interpretation of the scholar trained to believe in the authority of science (Keel, 2023, 42). This paper, however, would not be possible without the resilience and clarity of the impacted communities in Los Angeles who began documenting the violence of the county jails long before we were aware of this crisis.

Thus, we root our analysis in an abolitionist anthropology that seeks to work “within, against, and beyond the state in the service of collective liberation” (Shange, 2019, 10). We grapple with government-issued autopsy reports, working within the politics that govern death investigation and tethering our analysis to the perspectives of impacted communities who are working against weaponized death records. Our research thinks beyond the current criminal punishment system by refusing the naturalization of death by incarceration and reading counter to “the normalization of violent difference-making hierarchies” (Sojoyner, 2023, 21) within county records. At the same time, our study moves beyond traditional ethnographic methods and includes the voices of those directly impacted by violence and death inside Los Angeles County jails. Rather than merely quote their stories, we integrated the vision, interpretive insights, and abolitionist strivings of impacted communities into the analytic analysis used in this paper. Their direct experiences with state violence and tacit knowledge constitute the condition of possibility for this paper and our understanding of how law enforcement power appears in county-issued death investigation documents.

Death investigation and the US jail system

The most recent data released by the Bureau of Justice Statistics (BJS) reported the highest death rate for incarcerated people in US jails since the federal government began tracking these figures in 2000 (Carson, 2021). These data paint a troubling picture of pre-trial detention and reflect increased struggles with mental illness, addiction, and chronic health conditions. Although these figures are alarming, the accuracy of these data is compromised by a number of limitations. In 2017, a panel of prison and jail administrators, researchers, and healthcare professionals, convened by the National Institute of Justice, noted the important, yet inconsistent, role played by the nation's medical examiners in investigating and documenting deaths within the US carceral system (Russo et al., 2017, 21). Other studies of law enforcement-related deaths outside of carceral institutions have similarly pointed to the key role played by medical examiners in arrest-related death data inaccuracies (Feldman et al., 2017; GBD, 2019, 2021).

Currently, a patchwork of Coroner and/or Medical Examiners Offices serve as the primary accountability mechanism for determining if a death in jail or prison was natural or otherwise. Medical examiner and coroner offices across the nation are biomedical institutions made up of physicians, forensic pathologists, lab assistants, and crime scene specialists. Most offices are governed by a hierarchical social system with a chief medical examiner (ME) or coroner who oversees a team whose medical background and training varies according to state and county law. Ideally, ME departments and coroners are designed to be nonpartisan institutions distinct from prosecutorial authority and law enforcement. Their jurisdiction is determined by state laws and local regulations that specify the types of deaths medical examiner and coroners are required to examine, if a medical degree is needed to perform their work, whether they are elected or appointed, how long they can hold office, and the processes for reporting the results of their death investigations.

In 40 states, autopsies are required for deaths in jail, and 17 states mandate autopsies for deaths in state custody, with additional states requiring investigation into violent, sudden, or unusual deaths (CDC, 2015). Unlike deaths that occur on the street, deaths that occur in carceral institutions are without public witnesses, leaving the autopsy report to serve as a monolithic account of death. The outsized importance of the autopsy in the context of jail deaths is compounded by the conflicts of interest involving law enforcement and medical staff who oversee inmates. In 48 of the 58 counties in California, the office of the Sheriff, which operates county jails, is also the coroner charged with investigating deaths within the jails it administers (California State Association of Counties, 2014). Sheriff's deputies, doctors, and detectives are often present during the autopsies of deaths that occur in the facilities they manage. Forensic pathologists have been found to exhibit cognitive bias in both official determinations of manner of death and in experimental settings when presented with irrelevant non-medical information (Dror et al., 2020). Increased contact with law enforcement during key moments of biomedical discovery and analysis is ripe for bias in favor of reduced liability for the Sheriff, its contractors, and the county, which supervises both departments (LASD and DME-C) and bears ultimate liability. Additionally, deaths within Los Angeles County jail are investigated by detectives in the same Sheriff department that operates the jail. This additional conflict of interest adds further obstacles to impartial oversight.

The clear connection between the DME-C and LASD complicates access to autopsies of those that pass in jail. Autopsies are considered public records under the California Public Records Act (CPRA) and can be accessed by members of the public for a fee unless they are subject to a specific exemption. In the case of deputy-involved deaths, the Sheriff often invokes a law enforcement records exemption (§ 6254[f]) to enact a “security hold.” A security hold keeps the autopsy not only out of the hands of those that might provide oversight and public scrutiny but also to the next of kin seeking answers, closure, and justice.


In 2019, senior campaign leads Helen Jones and Michele Infante, both members of the non-profit organization Dignity and Power Now, requested autopsies for all deaths that occurred between 2009 and 2018 in four county jails under the jurisdiction of the Sheriff from the Los Angeles County DME-C. The four jails were Twin Towers Correctional facility, Men's Central Jail, Century Regional Detention Center, and Pitchess Detention Center. In late June of 2019, the Los Angeles County DME-C released 56 autopsies.

To better understand what was not included in the county-supplied documents, we analyzed all available proceedings of the LA County Claims Board, where settlements for wrongful death lawsuits against the county are ratified. We found 17 total cases in which the city paid settlements to families who lost a loved one in LA county jails. Only three of those cases were in the responsive documents of our partners’ records request. We discovered and acquired the additional report that was on the LA County Claims Board website and in the LA County DME-C database (i.e., without security hold). We additionally acquired a second autopsy report from the family of a decedent that obtained their security held autopsy through litigation. We analyzed a total of 58 autopsies, representing the maximum possible sample size given structural limitations. The earliest case we analyzed occurred on March 30, 2009, and the latest occurred on November 28, 2018.

Autopsies in LA county are variable in length (from ∼15 to >40 pages), and include a set of both standardized (microscopy, specimen collection, checklist of anatomy for internal and external examination, external skin diagram(s), toxicology lab reports, collected medical evidence) and open-ended forms (field investigator's narrative, autopsy findings). Larger sections of text are often typed, and most fields requiring a sentence or less are hand-written, as are diagrams (Figure 1). The autopsies we reviewed amount to 22% of the 260 deaths inside LA county jails that were reported to BJS between 2009 and 2018 (Smith, 2020). We believe that we have a non-random sample. Given the LASD's history of strategic and liberal use of security holds on autopsies that may bear liabilities for the county, this sample likely biases towards cases that bear less liability to the Sheriff. Despite these structured exclusions, the forms of violence and potential medical mistreatment observed in our study do not represent the LASD or the DME-C in a favorable light.

Details are in the caption following the image
External examination diagram (excluding head and neck) from the autopsy of John Horton III. Documentation of tattoos and trauma are not separated, visually minimizing trauma.

Most states in the United States use the following death designations, which appear in our case study: “accident,” “homicide,” “natural,” “suicide,” and “undetermined” (Hanzlick et al., 2002). Each manner of death classification is not a legal conclusion of culpability or intent. Rather, it is important for determining who or what institution is legally responsible for the loss of life. We focus on deaths reported to be “natural” and, to a lesser extent, “undetermined.” These designations ascribe the least potential liability to law enforcement and play a crucial role in the multilayered social processes that naturalize unnatural death.

We developed a protocol for systematically analyzing these autopsies utilizing peer review literature on forensic pathology and the conceptual insights of community activists Jones and Infante, who, for the last decade have navigated the medico-legal bureaucracy surrounding the production of death investigations in the Los Angeles County jail system. Through Jones and Infante, the perspectives and experiences of dozens of other impacted families informed the protocol we used to critically examine these state records. The protocol aims to systematically and critically read these official documents and open up community-insights and archival analysis to undergraduate students. Damien Sojoyner has recently argued that carcerality is the dominant mode of governance in the contemporary state, a state whose “main imperative […] is to make life legible through an archival process” (2023, 24). In addition to abstracting medical phenomena that reveal the drivers of death, our approach attempts to assess if and how county officials have overdetermined deaths in custody, attempting to make deaths legible along the lines of racial imaginaries that minimize liability.

Over 30 months, the authors, and a team of 19 student researchers from UCLA created a 198-page protocol (Bedewyi et al., 2023a) for analyzing these documents across 84 variables that were designed with direct input from our community partners. Each autopsy analysis was led by an individual analyst and then checked by a partner. Every analysis was validated by a third analyst, and columns were validated by teams of three to ensure that the data gathered for each variable bore a homogeneous interpretation across all cases. Patterns identified through analysis of this data (Bedewyi et al., 2023b) were then subject to individual case reviews to assess for potential mitigating information not captured by our variables.

Our focus on the largest jail system in the world bears both study design benefits and limitations. Although the size of the jail system enables a larger sample size, larger jails on average have lower mortality rates than smaller jails (<50 incarcerated people), and rural jails continue to have the highest mortality rates (Carson, 2021). Jailed women consistently have higher rates of mental illness and death than men (Bronson & Berzofsky, 2017; Carson, 2021).1 Proportionate to the total female population of the Los Angeles jail system, only 10% of the cases we analyzed are female. Despite the gravity of the issues documented herein, our analysis is likely not fully capturing the risks faced by the growing number of incarcerated women and those in smaller and rural facilities. We were also unable to determine the gender identity or sexual orientation of any of the decedents.


The naturalization of unnatural deaths in jail involves a series of actions and non-actions that constrict what is known about lives lost behind bars. Under delimited conditions of knowing, racial bias and biomedical prejudice are allowed to influence death determinations. Our analysis began with the observation that the deaths of jailed Black people, both men and women, are disproportionately labeled as “natural.” Black people were 64% of natural deaths, while only 38% of the sample population and only 8% of the county population. All Black women in our sample supposedly died from natural causes (n = 3, 50% of all female deaths).2 We also observed that 8 of the total 58 reviewed cases were classified as undetermined (13.79%), more than four and a half times higher than the average percentage of undetermined deaths among residents of Los Angeles County between 2007 and 2017 (2.98%) (LA DME-C 2018). All the undetermined deaths in our sample were of men (Latina/o, n = 5, Black n = 3) and predominantly young. Undetermined deaths averaged an age of 24.5 years old (range = 18–35) at time of death, compared to an overall average of 42 for all cases and 45.5 for deaths deemed natural (See supplementary materials for additional quantitative results).

To understand both the racialized production of unnatural death and subsequent racialized naturalization (or non-determination) our results are structured as follows: (1) pre-trial detention (2) mental illness (3) use of force (4) medical care (5) conflicts of interest (Figure 2).

Details are in the caption following the image
The above figure represents an overview of our findings. On the left we depict our focus on pre-trial detention as the dominant precondition for jailed death, followed by three deeply interconnected correlates of jailed death. We then focus on the predominant determination of death (“natural”), and another determination (“undetermined”) that is overrepresented in our sample compared with LA averages. On the right we depict the informational and institutional context that defines our access to death determination documents. The total number of deaths during this period is based on BJS data supplied to a news outlet (Smith, 2020). Image by Amisha Gadani and Nicholas Shapiro. [This figure appears in color in the online issue]

Pre-trial detention

More than three quarters (78%) of the individuals whose autopsies we reviewed were being held prior to trial, meaning that they were incarcerated while legally innocent and may never have been convicted of a crime.3 Financial inequality, disproportionate policing and criminalization of racialized communities, and widespread use of monetary bail has propelled the United State to maintain the world's highest pretrial detention rate (Coyle et al., 2016). In addition to life and earnings disruptions, pre-trial detention has been found to bear substantial negative impacts on jurisprudence, including increasing the likelihood of guilty pleas (Dobbie et al., 2018) and increasing sentence length compared to released populations (Oleson et al., 2016) without decreasing failure to appear rates (Tafoya, 2015) or reducing future crime (Dobbie et al., 2018). Research has demonstrated that bail judges are racially biased against Black defendants (Arnold et al., 2018). Our observations in this paper contribute to an understanding of death as an ultimate cost of racialized pre-trial detention practices.

Mental health

Carceral institutions are increasingly being used to detain people with mental health issues in place of providing substantive treatment, leading some to refer to prisons and jails as “the new asylums” (Fuller et al., 2016). Between 2009 and 2018, the percentage of people incarcerated in Los Angeles County jails that were placed in mental health units or on psychotropic medications more than doubled, from 14% to 30% (Holiday et al., 2020). This increase is also reflected by a 350% increase in the number of people found mentally incompetent to stand trial between 2010 and 2015 (Katz, 2016). National data from 2011−2012 similarly indicate that 44% of those incarcerated within the nation's jails have been told in the past that they had a mental disorder (Bronson & Berzofsky, 2017). Those with serious mental illnesses are not only incarcerated at higher rates but are also disproportionately met with violence and death in encounters with law enforcement (Fuller et al., 2015; Mueller et al., 2019; Perry & Carter-Long, 2016).

We found mental health issues to be an overwhelming correlate of death in Los Angeles County jails during the period of study. Nearly 3 in 4 (70%) of the autopsies we analyzed note a history of mental illness for the decedent, whether substantiated or not. This variable also tracked across other factors. Of those that had been designated as suffering from a mental illness, 67% were pre-trial. The relationship between mental illness and healthcare is complex and is discussed at length in the healthcare sub-section below. From the documents reviewed, it appears that those with mental health issues faced an elevated likelihood of death potentially stemming from medications and poorly treated chronic or acute health issues, as well as failures to meet their basic needs. We also found that 6 out of 8 undetermined deaths also had substantiated (n = 4) or unsubstantiated (n = 2) histories of mental illness.

The naturalization of deaths of incarcerated individuals begins at the normalization of incarcerating those facing mental health crises and concludes with the determination of these deaths as natural by the medical examiner-coroner. Deaths of those with professed or documented mental illness were deemed “natural” at a rate that was more than double and nearly triple, respectively, the deaths of those that had no noted history of mental health. Indicative of how mental illness is not only exacerbated but potentially initiated by jailing, four of the cases deemed suicides were of individuals without a noted history of mental illness in their death investigation documents.

The case of one decedent, John Horton III, who was the son of our community partner Helen Jones, is particularly important for illustrating how jailers use claims of mental illness, drug abuse, and limited documentation to hide possible causes of death and in turn influence the conclusions of the medical examiner-coroners. According to the coroner's investigator's narrative, on March 30, 2009, Horton, a 22-year-old Black artist from Watts, was found at 3:45 a.m. unresponsive in a single occupancy cell inside Los Angeles Men's Central Jail with “a possible ligature around his neck.” The deputy medical examiner who wrote Horton's autopsy declared his death a suicide. According to the forensic investigator's narrative, there was no evidence indicating that the investigator or the deputy medical examiner made use of jail medical records to substantiate whether Horton had been diagnosed with a mental illness.

The forensic investigator notes at the beginning of their report that Horton had a confrontation with jail deputies “on or about 3/18/09” and was claimed “to be under the influence of PCP” nearly 1 month into his incarceration. These claims work to establish Horton as a drug addict, even though the toxicology results after his apparent suicide revealed no controlled substances in his system. The autopsy documents several injuries on Horton's body that do not square with the classification of his death as suicide by hanging. These injuries include a quarter-inch abrasion on the bridge of his nose, hematoma on the right side of his forehead, abrasions on his right shoulder, a 1-in. hemorrhage on the lower right side of his back, and soft tissue hemorrhages within his liver and kidneys. Photos of the scene of death taken by the forensic investigator revealed fresh blood on the bridge of his nose, in his mouth, and on his jail uniform, indicating the temporal proximity of these injuries to his death. Horton's mother questioned the disconnect between the death determination and the multiple documented traumas after a doctor told her that her son would not have been able to walk after sustaining those injuries, let alone shred a thick blanket and weave it into a ligature. Some 3 months later, the official manner of death was changed to undetermined. Horton's reclassification raises the question of how many cases in which Sheriff-advanced narratives have been bio-legally codified into fact via medical examiner documents. And how far should an unsubstantiated claim of mental illness advanced by the Sheriff be used to frame the official manner of death of lives lost behind bars?

Use of force

In none of the analyzed cases did the medical examiner determine that deputy use of force directly contributed to the cause of death. Also, use of force records maintained by the jail do not appear to have been consulted in the writing of any of the post-mortem reports. As a result, we do not have systematic data across these reports on deputy use of force. It has been well established that LASD has significantly under-reported deputy-involved conflicts (Baird et al., 2012, 41). To unpack how use of force can become naturalized, we turn to a case in which an acutely mentally ill man was moved from specialty housing to the general population and was subject to violence both by his cell mate and by deputies.

Juan Manuel Correa Jr was a 31-year-old Latino man who died in Men's Central Jail on September 26, 2017. He was bipolar and for reasons unknown was transferred out of the Twin Towers psychiatric wing designed to support those with mental health issues (Correa, Sr. et al. v Los Angeles County et al., 2018) and into the Men's Central Jail general population. Given his mental illness, this bureaucratic shift into the general population increased his likelihood of being the recipient of violence. Correa notified a deputy that he was not feeling well and allegedly noted that he wanted to kill himself. After 15 min, the deputy returned, and Correa repeated his plea and shook the bars to which his cellmate told him to shut up and threatened him with violence. The two men started fighting, and three deputies pepper sprayed them. Although deputy accounts differ, the District Attorney's office noted that there were a minimum of four individual dispatches of pepper spray (Justice System Integrity Division, 2019). Within 2 min of the first spray, Correa was handcuffed and led to a hallway where he was seated on a bench. Within 10 min, he had slid off the bench and was laying on the floor, noting that he was not feeling well. While decontaminating in the shower, Correa again slumped to the floor, around 23 min after the first pepper spraying and 37 min after his first plea for help. When deputies approached him, he was unconscious and without a pulse. Paramedics determined him to be dead approximately a half-hour later. The LA County DME-C concluded that the manner of death was natural (Figure 3).

Details are in the caption following the image
A security camera still of Juan Manuel Correa Jr. having slid off of the bench after having been pepper sprayed and engaged in an altercation with his cellmate. Source: Justice System Integrity Division, 2019. [This figure appears in color in the online issue]

The incident report form for the events leading up to Correa's death does not mention any use of force by deputies. This is one small demonstration of how deputy violence can be bureaucratically erased prior to oversight by third parties (Figure 4).

Details are in the caption following the image
An expert from the 2017 incident report form for Correa's death, missing is an x next to “EMPLOYEE ASSAULT–GASSING”. Source: Correa, Sr. et al. v. Los Angeles County et al., 2018.

In an interview with the office of District Attorney Jackie Lacey, the deputy medical examiner that presided over his autopsy “reasoned there was no known mechanism in OC spray [Oleoresin Capsicum, i.e., pepper spray] that could cause death in people. In his experience, he has never seen this” and further that “if the OC spray was the cause of Correas's [sic] death, Correa would have died immediately after the spray, and the 20-min gap in time proves otherwise” (Justice System Integrity Division, 2019, 7). In this explanation, the medical examiner deploys three lines of dismissal. The first and last are related, as he avers that, in essence, pepper spray is not known to cause cardiomyopathy and that if it could cause death, it would be immediate. Scientific literature has in fact documented cases of pepper spray-related cardiac distress or death in the United States, Canada, and Turkey (Cil et al., 2012; Pollanen et al., 1998). For example, a 21-year-old police trainee with no past medical history completed a pepper spray training course and decontaminated in the shower without any signs of distress. Immediately after, he collapsed, and his heart stopped beating (Arora, 2019). He was revived after 35 min of cardiopulmonary resuscitation (Correa received 15 min of resuscitation, and approximately 33 min elapsed between his collapse and the arrival of paramedics). As early as 1995, the American Journal of Forensic Medicine and Pathology published scientific studies demonstrating the lethal effects of pepper spray exposure (Steffee et al., 1995). Hypothesized mechanisms for pepper spray impairment of cardiac function range from vasoconstriction-mediated deterioration of cardiac function (Lechner et al., 2021) to pulmonary edema, which was noted in Correa's autopsy, and which can affect gas exchange and lead to arrhythmia via this excess stress on the myocardium (Arora, 2019). The deputy medical examiner's claims—that there is no history or mechanism of OC spray-related death and that if there were OC-related deaths they would occur immediately—are not consistent with the scientific literature.

Lastly, the deputy medical examiner uses his own clinical experience to make claims about the non-lethal nature of OC exposure. This tactic has been identified in defense expert witnesses in torts as a means of legitimizing the overly constricted epistemological fields that serve as the foundation of their expert testimony (Shapiro, 2014). The District Attorney's medical assessment of Correa's death ends with an assertion by the Deputy Medical Examiner of a statistical counterfactual: that Correa was just as likely to die from gardening as being pepper sprayed due to his enlarged heart (Justice System Integrity Division, 2019, 7). Ascribing Correa's death to his heightened arrhythmia susceptibility due to his weight, at the expense of the immediate context of his death, aligns with what a dozen doctors have called “structural gaslighting,” or the “weaponization of medical language” to individualize responsibility for cause of death (Crawford-Roberts et al., 2020).


The 1976 Supreme Court decision in Estelle v. Gamble established the legal obligation of jailers to attend to the serious medical needs of the people in their custody in order to uphold the Eighth Amendment's prohibition of cruel and unusual punishment. In addition to high rates of mental illness, jailed populations bear high burdens of chronic diseases such as asthma, diabetes, and hypertension (Davis et al., 2011). Issues of healthcare and healthcare access appeared to be present in the autopsies assessed in this study.

We found inconsistent documentation about the medical history of the deceased in the autopsy reports and coroner investigator narratives we studied. In only 7% of cases did the medical examiner review the medical history of the deceased prior to autopsy. The coroner investigator, who visits sites of death, reviewed the medical records of only 34% of the cases in our study. Field investigators working for Los Angeles County are not licensed medical doctors, but they do collect statements, samples, and documentation and establish the narrative frame for the medical examiner. The task of performing the autopsy and declaring manner of death falls to a medical examiner. Medical examiners documented reviewing the medical history of the deceased in only 6% of the cases in our study. Relaxed standards for reviewing the deceased's medical history before autopsy is possibly a contributing factor to misclassification in the cause and manner of death. This oversight, when coupled with the absolute control of deputies and jail staff over the lives of the incarcerated, creates the opportunity for biased death determinations that reduce the culpability of jailers for the death of an inmate.

In some cases, multiple medical requests in the weeks and months leading up to death were seemingly ignored or mired in a backlog. Such is the case of Clyde B. Harris, a 48-year-old Black male, whose gradual decline was documented by his social worker, who filed two medical referrals in the immediate lead-up to his death without any apparent action. The medical examiner determined his death to be natural and the primary cause of his death to be atherosclerotic heart disease. In other cases, medical examinations appear to have missed dramatic health conditions. Malik Jevon Slims, a 25-year-old Black male, complained of chest pains followed by pain in his left flank and was prescribed ibuprofen, acetaminophen (generic for Tylenol), and methocarbamol (a muscle relaxant) by the Men's Central Jail infirmary. Later that day, three days before Christmas, he collapsed and died from acute pneumonia with empyema (accumulation of pus within the cavities of the lungs) from a multiple-drug-resistant bacterial infection. A pulmonary infection at this advanced stage should have been easily diagnosed through chest auscultation with a stethoscope, yet this death was deemed to be natural by the DME-C.

In addition to issues related to a lack of medical intervention, multiple deaths may be due to pharmaceutical interventions. Jorge Rosales was an 18-year-old Latino male, incarcerated at the Twin Towers facility in psychiatric housing. After being arrested on July 27, 2011, Rosales was later beaten by deputies on October 4, 2011, when reportedly running away from a deputy escorting him to the shower. His assault resulted in multiple head injuries including a black eye.4 The death investigation documents indicate that later in the evening he yelled in pain and disturbed the sleep of others on the ward. Rosales was seen on his feet by deputies in the first minutes of October 6, 2011, only to be found unresponsive on the floor by 12:42 a.m. and then pronounced dead within an hour. The medical examiner declared that he died due to acute hemorrhagic pancreatitis, an illness rarely seen in men under 35 (Weiss et al., 2019). As the medical examiner noted, acute hemorrhagic pancreatitis in men of this age is either due to blunt force trauma, potentially resulting from the use-of-force incident a day-and-a-half earlier, or due to his dual prescription of both Olanzapine, an antipsychotic used to subdue symptoms of bipolar disorder and schizophrenia, and the antibiotic Trimethoprim-sulfamethoxazole. Both are known to cause drug-induced acute pancreatitis (Simons-Linares et al., 2019). Although the medical examiner did not document visible abdominal bruising, Rosales’ inferior abdomen displayed faint green discoloration, which can be a rare sign of acute pancreatitis, one that often clinically presents 3–4 days into the course of the disease (Guldner & Magee, 2021). This suggests that the drug-induced etiology is more likely, although a combination of both blunt force trauma and pathological pharmaceutical intervention is possible. Moreover, from the medical examiner's report, it is unclear when the trimethoprim-sulfamethoxazole was prescribed, and for what ailment, but all three causal routes trace back to actions by the LASD. The manner of death was left undetermined.5

Prescription of antipsychotic medication was relatively common in the reviewed autopsies. Five people were prescribed Risperidone, and four people prescribed Olanzapine (one individual was prescribed both). Risperidone is potentially implicated in three deaths. Two people on Risperidone died of pulmonary thromboembolism, a relatively rare lung artery blood clot that is known to be caused by the drug (Borras et al., 2008). Steven Douglas Holmes (white male, 38) died of a pulmonary embolism and deep leg vein thrombosis. Milton Earl Travis (Black male, 50) similarly died of a pulmonary thromboembolism and probable deep vein thrombosis while taking Risperidone. Both deaths were deemed natural, and the medical examiner did not mention the potential connection to their antipsychotic medication. In a third death, an apparent fatal seizure of Cesar Raul Franco (Latino, 26), the medical examiner notes “Risperidone use” as a primary cause of death. Confusingly, Risperidone use is not associated with seizures (Holzhausen et al., 2007) as the medical examiner also later notes in the autopsy. An “old” contusion to the right inferior frontal cortex was found in autopsy and, as the medical examiner speculated, may have been the site in the brain from which the seizure originated. There was no apparent attempt by the deputy medical examiner to date the contusion that likely lies at the center of the cause of death. Additionally, records of deputy use of force do not appear accessed in this (or any) medico-legal report. An attempt to date the contusion in conjunction with a review of use of force or incident report records could have helped elucidate or refute potential deputy involvement in this death. His death was deemed undetermined.

Quanell Holton, a 43-year-old Black male, was housed at the acute mental ward of the Twin Towers jail in a padded single person cell and restrained in a jacket from which he had escaped shortly before passing away due to diabetic ketoacidosis. The deputy medical examiner deemed the manner of death to be natural. Yet the report neither includes the word “insulin” nor makes any mention of care for his diabetes, other than noting that it was part of his known history. The closest mention of diabetes care is the declared absence of prescribed insulin: “The decedent was not taking any prescription medications, since he was not being issued any prescription medications.” Medical neglect is not entertained within the report. Dr Nickolay Teophilov, Chief Physician at the LASD, is listed as present at Holton's autopsy.

At times, incarcerated peoples’ basic needs of food and water appear not to have been met. Sergio Silva, a 33-year-old Latino male, dropped 61 pounds (30% of his body weight) between May 4, 2016 and his death on July 29 of that year of dehydration. Silva's death was classified as natural. In another case, having recently turned 18, Jimmy Cornejo was transferred from a juvenile facility to Twin Towers on October 7, 2010. In less than a month, Cornejo, a Latino male, was dead from starvation. At the end of his time in juvenile detention, he was in an altercation with a custody officer that his family told forensic investigators was self-defense. From family statements detailed in the forensic investigator's report, it appears that Cornejo remained a ward of the state because he could not afford bail and that he refused to eat in order to prompt his release. Cornejo's manner of death was classified as undetermined. The deaths of Silva and Cornejo could have been prevented.

Conflicts of interest

To understand how biases might be systematically introduced in the assessment of deaths in custody to allow for the naturalization of unnatural deaths, we analyzed autopsies based upon the National Association of Medical Examiners (NAME) recommendations for reporting deaths in custody (Mitchell et al., 2017). We assessed who documented reviewing medical records and which of the records defined as relevant (namely jail health records and Emergency Medical System [EMS] run sheets) were reviewed. None of the cases in our study documented that both the forensic investigator that completed the field report and the deputy medical examiner (DME) that wrote the autopsy report consulted the decedent's pertinent health records as NAME recommends. For only one natural death (and three across all manners of death) was it documented that both the forensic investigator and the DME directly consulted any of the decedent's medical records (jail records or EMS run sheets).

Coming to an understanding of the medical needs of the deceased and the quality of the medical care provided to them through the reports of the DME-C proved difficult because of this insufficient documentation. Without direct consultation of pertinent medical records, assertions by Sheriff deputies and detectives with deep conflicts of interest serve as the sources of medical and mental health information and limit the credibility of DME-C findings. Relaxed standards for reviewing the deceased's medical history before autopsy creates the opportunity for biased death determinations that reduce the culpability of jailers for the deaths of incarcerated people.

Issues of LASD influence extend into the dissection room. We found that law enforcement was present during 51 out of the 58 autopsies we analyzed. Among those deaths classified as “natural,” law enforcement was present in all but two cases. Additionally, Dr Nickolay Teophilov, the Chief Physician at the LASD, who has supervisory responsibilities over medical care in LA jails, attended 14 of the autopsies (24%) across all manners of death (including 8 deaths deemed natural, or 33%), presenting a potential conflict of interest that may minimize any liabilities of the medical treatment or lack of treatment provided by his staff. The presence of Sheriff staff during crucial periods of physical analysis and the formation of the death narrative appears to be another avenue for LASD influence on medicolegal determinations.


In this paper, we used a mixed methods approach to assess how the autopsy serves as an instrument of necropolitical power that can serve the interest of the state by naturalizing the disproportionate death among Black and Latino/a wards of the state in Los Angeles County. Insufficient due diligence and conflicts of interest enable the medical examiner-coroner to transform deaths by incarceration—produced through physical and psychological punishment, medical neglect, and incompetence—into natural processes within racialized bodies. This shields law enforcement, jail staff, and local government from being held accountable for in-custody deaths. Jail deaths in Los Angeles sanitized by medical examiners and coroners offer clear evidence of a necropolitics aligning law enforcement and medicine. It is important to note that the Los Angeles County DME-C is arguably one of the nation's best models for post-mortem death investigation. Our findings indicate that the politics of a purportedly progressive state and the application of science by a top ME office do not sever the violence and lack of accountability of anti-Black and Latino jailing practices.

Yet by working with Black and Brown communities impacted by police violence and re-evaluating the claims of the county autopsy through a method forged with those directly impacted, we demonstrate in this paper how necropolitics leaves in its wake forms of biomedical knowledge that can work against the macabre interests of the carceral state and police power. Access to autopsies that serve the public interest is therefore essential for holding the state accountable for lives lost under its watch. Jail diversion programs and prison abolition efforts are the clear-cut way to prevent people from dying while incarcerated. Forging a path towards these progressive reforms requires access to records that prove jail to be a lethal place. Without these files or the courage of communities contesting the truth claims of compromised autopsies, the means for limiting the unchecked power of the carceral state and police power are lost. This loss comes at the expense of our ability to comprehend the social and political relationships that produce unnatural death in racialized people held in state captivity.


We would like to recognize the people who have died in LA County jail during this study period and beyond and the impacts of those deaths as they ripple disproportionately through Black and Latino communities. We would like to recognize the struggles of impacted families and the years of work by organizers, particularly Helen Jones and Michele Infante, for making this work possible. We would like to thank Dignity and Power now for their leadership and their profound care for impacted families. We owe deep thanks to the many students that made this work possible by reviewing autopsies or working on protocols to review autopsies: Raven Bedewyi, Arezoo Kalan, Daniela Vieira, Gabriela Ramirez-Vides, Rachel Smith, Alina Giapis, Lilia Diaz, Rachel Chau, Alyssa S. Moreno, Armando Davalos Jr., Tasnia Haider, Oscar De La Rosa, Destiny Ryales, Khalil J. Jacobs-Flier, Gianna Gunier, Najma Ali, Bianka Aguirre, Katherine Manganaro, Cora Miller, Vivica Rush, Bella Hung, Kevin Shaffman, Armond Corshawn Lee, Hussein Saleh, Yoshiko (Yoshi) Kohlwes, Jasmin Argueta, Cienna Henry Milton, and Mattie Hollenbach. We would like to thank Kelly Lytle Hernández for introducing us to our community partners and the many students in both Carceral Ecologies and BioCritical Studies that worked adjacent to this work for their camaraderie and solidarity. The content is solely the responsibility of the authors and does not necessarily represent the official views of the funders that provided support for this research. This study was supported by National Institutes of Health—National Library of Medicine (grant no. 1G13LM013930-01); Robert Wood Johnson Foundation (grant nos. 108655, 79694); Russell Sage Foundation (grant no. 2111-34931).


    • 1 Data on non-binary and trans rates were not available. Similarly, gender identity was not available in any of the autopsies we reviewed.
    • 2 No white, Asian, and Latina female deaths were represented in our sample of natural death cases.
    • 3 We were unable to determine custody status (pretrial vs convicted) for three individuals.
    • 4 The accounting of events on the forensic investigator's report directly contradicts the summary of events posted by the county government in a settlement over the beating. https://file.lacounty.gov/SDSInter/bos/supdocs/116070.pdf
    • 5 Reginald Harper (a 35-year-old black male) also died of Idiopathic pancreatitis in 2014 also while prescribed Risperidone. His death was deemed undetermined.