ARTICLES
STRATEGIC SEGREGATION IN THE MODERN PRISON

Sharon Dolovich*
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
I. PRISON RAPE, HYPERMASCULINITY, AND THE FEMINIZATION OF
VICTIMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
II. THE ORIGINS, MECHANICS AND EFFECTS OF L.A. COUNTY’S K6G
UNIT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A. Origins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
B. Institutional Structure: Managing the Segregation. . . . . . . . .
C. Classification: Deciding Who Is In and Who Is Out . . . . . . . .
1. The Initial Classification . . . . . . . . . . . . . . . . . . . . . . . .
2. The Second Step: Classification as Detective Work . . . . .
D. Is K6G Safer? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Direct Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2. Corroborating Circumstantial Evidence . . . . . . . . . . . . .
III. K6G OR NOT K6G: THREE CRITICAL PERSPECTIVES . . . . . . . . . . . . .
A. “Demoralizing and Dangerous”: The Antisegregationist
Objection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2
11
19
19
23
25
27
30
44
44
48
54
55

* Professor of Law, UCLA School of Law. Thanks are due to Joe Doherty, Director of the UCLA School of
Law’s Empirical Research Group, who helped me design the research project on which this Article is based and
who guided me through a sixteen-month IRB process; Craig Haney and Jeannie Suk for their generous
willingness to respond to this Article and their thoughtful commentary; Asli Bali, Michael Boucai, Bennett
Capers, Heidi Feldman, Cheryl Harris, Allegra McLeod, Lara Stemple, Chad Trulson, Robin West, Jordan Woods,
Noah Zatz, and workshop participants at Georgetown University Law Center and the Williams Institute at the
UCLA School of Law for helpful comments and conversation; Scott Dewey, Rebecca Johns, Max Kamer, Anne
Marie Morris, Michael Reiss, and the reference librarians of the Hugh and Hazel Darling Law Library for their
terrific research assistance; Meredith Garagiola, Megan Wolf, and the editors of the American Criminal Law
Review for their hard work on this article; the UCLA Academic Senate, the UCLA Dean’s Office, Georgetown
University Law Center, and Harvard Law School for supporting this research; and especially Jody Freeman for
her encouragement, support and advice over the life of this endeavor. I also owe a debt of gratitude to several
officers of the L.A. County Sheriff’s Department: to Chief Alex Yim for his openness to research and for
generously granting me access to the Jail; to Director Karen Dalton and Lieutenant Robbie Ibelle for all they did
to facilitate my research; and, above all, to Senior Deputy Randy Bell and Deputy Bart Lanni for graciously
welcoming me into their domain and helping me to understand it. Their dedication and professionalism are second
to none. Unless otherwise stated, descriptions of the structure and functioning of the L.A. County Jail offered in
this Article are drawn from my own observations during the course of my research. All the views and conclusions
presented here are based soley on my own observations and the research I conducted in the Jail, and do not
represent the views of the L.A. County Sheriff’s Department. © 2011, Sharon Dolovich.

1

Electronic copy available at: http://ssrn.com/abstract=1894819

2

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

B. Drawing the Line: The Underinclusivity Objection . . . . . . . .
C. Unconstitutional: The Equal Protection Objection. . . . . . . . .
IV. CONCLUSION: THE PROSPECTS FOR REPLICATION . . . . . . . . . . . . . . . .
METHODOLOGICAL APPENDIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
APPENDIX B: QUESTIONNAIRE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

64
82
87
92
100

INTRODUCTION
In corrections circles, it is well recognized that people who are gay or
transgender1 face heightened vulnerability to sexual victimization behind bars.2
Although accurate statistics on prison rape are notoriously difficult to generate,3
recent research confirms this dynamic. A 2007 study conducted in the California
prison system found that “67 percent of inmates who identified as LGBTQ
reported having been sexually assaulted by another inmate during their incarceration, a rate that was 15 times higher than the inmate population overall.”4 Recent
Bureau of Justice Statistics findings suggest similarly disproportionate rates of
assault for LGBTQ detainees in juvenile facilities, with “[y]outh with a sexual
orientation other than heterosexual” reporting sexual victimization at a rate almost
ten times higher (12.5%) than that reported by heterosexual youth (1.3%).5 The
title of a recent publication by the advocacy group Just Detention International
(formerly Stop Prisoner Rape) crisply captures the point: LGBTQ Detainees Chief
Targets for Sexual Abuse in Detention.6

1. The term “transgender” denotes people whose gender identity does not match their birth sex. Throughout
this Article, I use the term “trans women” to refer to people who were born biologically male but who self-identify
and self-present as women.
2. As Human Rights Watch observed in its 2001 report on male rape in American prisons, “gay inmates are
much more likely than other inmates to be victimized in prison . . . .” HUMAN RIGHTS WATCH, NO ESCAPE: MALE
RAPE IN U.S. PRISONS 52 (2001), available at http://www.hrw.org/legacy/reports/2001/prison/. The report goes on
to note that “many gay inmates—even those who are openly gay outside of prison—carefully hide their sexual
identities while incarcerated . . . because inmates who are perceived as gay by other inmates face a very high risk
of sexual abuse.” Id. at 57. On the particular vulnerability of transgenders, see infra note 4.
3. See, e.g., GERALD G. GAES & ANDREW L. GOLDBERG, NAT’L INST. OF JUSTICE, PRISON RAPE: A CRITICAL
REVIEW OF THE LITERATURE 1–2 (2004), available at http://www.ncjrs.gov/pdffiles1/nij/grants/213365.pdf (discussing the various difficulties associated with studying the problem of prison sexual victimization, including the
“stigmatization associated with sexual victimization”).
4. LGTBQ Detainees Chief Targets for Sexual Abuse in Detention, JUST DETENTION INTERNATIONAL, Feb. 2009,
at 1 (citing Valerie Jenness et al., Center for Evidence-Based Corrections, Violence in California Correctional
Facilities: An Empirical Examination of Sexual Assault (2007)). This same team of researchers found that “59%
of [California’s] transgender population reported sexual victimization as compared to 4% of the general prison
population.” Valerie Jenness, The Victimization of Transgender Inmates (2006), available at ucicorrections.seweb.
uci.edu/. . ./Victimization%20of%20Transgender%20Inmates.ppt, slide 14.
5. BUREAU OF JUSTICE STATISTICS, SEXUAL VICTIMIZATION IN JUVENILE FACILITIES REPORTED BY YOUTH,
2008–09, Jan. 2010, at 1. The increased rate of victimization experienced by LGBT detainees reflected in these
figures is consistent with what is known about the culture of juvenile facilities. Anecdotal evidence suggests that
these findings actually underreport the overall incidence of sexual assault in juvenile detention.
6. See supra note 4.

Electronic copy available at: http://ssrn.com/abstract=1894819

2011]

STRATEGIC SEGREGATION

3

Gay men and trans women7 are not the only people vulnerable to sexual
victimization in men’s prisons and jails.8 But their assigned place in the prison
sexual hierarchy makes them almost automatic targets for such abuse.9 For this
reason, many carceral facilities around the country routinely house gay men and
trans women separately from the general population (GP). Often, this segregation
takes the form of protective custody, a classification that typically involves
isolation in “a tiny cell for twenty-one to twenty-four hours a day[,]”10 the loss of
access to any kind of programming (school, drug treatment, etc.),11 and even
deprivation of basics like “phone calls, showers, group religious worship, and

7. See supra note 1. At present, most prisons and jails in the United States follow the practice of classifying
detainees according to their genitalia, which means that preoperative trans women are housed with men. See
Farmer v. Brennan, 511 U.S. 825, 829 (1994). For strong criticism of this policy, see Christine Peek, Breaking Out
of the Prison Hierarchy: Transgender Prisoners, Rape, and the Eighth Amendment, 44 SANTA CLARA L. REV.
1211, 1219, 1247–48 (2004); Darren Rosenblum, “Trapped” in Sing Sing: Transgendered Prisoners Caught in
the Gender Binarism, 6 MICH. J. GENDER & L. 499, 522–24 (2000); STOP PRISONER RAPE & ACLU NATIONAL
PRISON PROJECT, STILL IN DANGER: THE ONGOING THREAT OF SEXUAL VIOLENCE AGAINST TRANSGENDER PRISONERS
7 (2005), available at http://www.justdetention.org/pdf/stillindanger.pdf (last visited Mar. 6, 2011). The U.S.
Department of Justice has proposed a national standard on which, “[i]n deciding whether to assign a transgender
or intersex inmate to a facility for male or female inmates, and in making other housing and programming
assignments, the agency shall consider on a case-by-case basis whether a placement would ensure the inmate’s
health and safety, and whether the placement would present management or security problems.” See National
Standards to Prevent, Detect, and Respond to Prison Rape, 76 Fed. Reg. 6248, 6281 (Feb. 3, 2011) (to be codified
at 28 C.F.R. pt. 115) (Att’y Gen. regulations proposed pursuant to PREA) (§ 115.42 (c)) [hereinafter National
Standards].
8. I return to this important point below. See infra Part III.B. Among the criteria known to increase the
vulnerability of male inmates are “mental or physical disability, young age, slight build, first incarceration in
prison or jail, nonviolent history, prior convictions for sex offenses against an adult or child, sexual orientation of
gay or bisexual, gender nonconformance (e.g., transgender or intersex identity), [and] prior sexual victimization . . . .” NATIONAL PRISON RAPE ELIMINATION COMMISSION REPORT 217 (June 2009) (Appendix B: NPREC
Standards—Adult Prisons and Jails: SC-1) [hereinafter COMMISSION REPORT]; see also Christopher D. Man &
John P. Cronan, Forecasting Sexual Abuse in Prison: The Prison Subculture of Masculinity as a Backdrop for
“Deliberate Indifference,” 92 J. CRIM. L. & CRIMINOLOGY 127, 164–75 (2001) (identifying a range of factors
known to correlate with vulnerability to sexual victimization in men’s carceral facilities, including age, projecting
“feminine characteristics,” physical size, and being a new inmate who is “unfamiliar with the game”).
9. For more on the prison sexual culture, and on the place of gay men and trans women in that culture, see infra
Part I.
10. Gabriel Arkles, Safety and Solidarity Across Gender Lines: Rethinking Segregation of Transgender People
in Detention, 18 TEMP. POL. & CIV. RTS. L. REV. 515, 538 (2009).
11. As Arkles explains, lack of access to such programming can carry with it serious hardships for people
behind bars:
For one, it is one of the only ways to interrupt hours of deadening boredom with some sort of
activity. For another, certain programs can help build skills to increase the chance of success once
released. For another, while typically prisoners are only paid pennies an hour for their labor, these
programs are virtually the only way to earn money to use in the commissary to buy such luxuries as
shampoo, toothpaste, cigarettes, or stamps. Finally, participation in certain programs can be
mandatory to make parole or an early release date. Even when not mandatory, successful
participation in prison programs is generally regarded as a very favorable factor in parole
decisions.
Id. at 542.

Electronic copy available at: http://ssrn.com/abstract=1894819

4

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

visitation . . . .”12 Such conditions, even if increasing a person’s protection from
sexual assault—a proposition some commentators challenge13—force vulnerable
prisoners into the cruel position of having to choose between personal safety and
the satisfaction of other basic and urgent human needs, above all, those of
community and fellow human contact.
There is, however, a notable exception to this national trend. In the Los Angeles
County Jail—the biggest jail system in the country—officials have found a way to
increase the personal security of gay men and trans women detainees without
forcing them to choose between safety and community. For more than two
decades, the L.A. County Sheriff’s Department (the Department), which runs the
County’s jail system, has been systematically separating out the gay men and trans
women admitted to the L.A. County Jail (the Jail) and housing them wholly apart
from GP.14 As a consequence of this segregated unit—long known as “K11” but
recently officially rechristened “K6G”15—gay men and trans women detained in
the Jail are relatively free from the sexual harassment and forced or coerced sexual
conduct that can be the daily lot of sexual minorities in other men’s carceral
facilities.16
In the summer of 2007, following a lengthy negotiation with both the UCLA
IRB and the Jail’s command staff, I spent over seven weeks conducting research in
the Jail.17 During that time, I observed the operation of K6G and the Jail more
generally,18 sat in on K6G classification interviews, spent countless hours in the
officer’s booth overlooking the K6G dorms, and had many informal conversations

12. Id. at 541 (explaining that “people who find themselves in segregation . . . typically find their rights and
‘privileges’ dramatically restricted”). As one federal court put it, summing up the situation, protective custody
brings the loss of “adequate ‘recreation, living space, educational and occupational rehabilitation opportunities,
and associational rights . . . .” Rosenblum, supra note 7, at 530 (quoting Meriwether v. Faulkner, 821 F.2d 408,
416 (7th Cir. 1987)).
13. See generally Arkles, supra note 10.
14. In corrections, prisons and jails serve distinct purposes. Prisons provide long-term housing, typically for
sentenced offenders serving terms of longer than one year, although the precise cut-off can vary by state. Jails only
hold sentenced prisoners serving short terms, typically less than one year. In addition, jails house individuals
awaiting trial but denied bail, convicted offenders awaiting sentencing, and prisoners sent from state or federal
prison to serve as witnesses in trials, whether their own or those of others. See Margo Schlanger, Inmate
Litigation, 116 HARV. L. REV. 1555, 1579, n.76 (2003). The role of jails in providing housing for detainees with
court dates explains why jails are typically situated adjacent to courthouses, although L.A. County is so large that
most Jail inmates with court dates have to be bused from the Jail to their respective courthouses.
15. This name change was apparently necessitated by a new computer system that was only able to count up to
“K10.” To avoid confusion, in this Article, I refer to the unit by its current designation of “K6G.”
16. Those sexual minorities housed in women’s prisons also face a heightened risk of sexual abuse in custody.
See COMMISSION REPORT, supra note 8, at 74. In this Article, I focus on the vulnerability of sexual minorities in
men’s prisons.
17. For a description of the research design, see Methodological Appendix, infra.
18. This enterprise was made possible by Chief Alex Yim, who generously allowed me open access to all parts
of the facility.

2011]

STRATEGIC SEGREGATION

5

with unit residents, custody officers, and other staff.19 I also conducted one-on-one
interviews, structured around a 176-question instrument,20 with a random sample
of K6G residents.21 In all, I interviewed thirty-two residents,22 almost ten percent
of the unit’s population at the time,23 a process yielding fifty-one hours’ worth of
audio recordings.24
This Article draws on that original research25 to provide an in-depth account of
the K6G unit. The aim is both descriptive and evaluative—to describe the
mechanics of the program and its implications for residents, and to assess the
weight of possible objections to the program’s design and to the undertaking as a
whole. As I show, L.A. County has managed to create a surprisingly safe space for
the high-risk populations K6G serves. That it has done so in a carceral system that
is severely overcrowded and notoriously volatile makes the success of the program
even more remarkable.
There is, however, no getting around it: with K6G, L.A. County is engaged in a
process of state-sponsored, identity-based segregation. Although this program

19. I took lengthy field notes each day and then dictated the notes at night, when what I had seen was still fresh
in my mind.
20. I developed this instrument with the help of my colleague, Joe Doherty. It is attached as Appendix B.
21. In the course of identifying my sample, I met with a number of K6G residents who, for reasons of mental
illness, developmental disability, or other psychological incapacities, were not able to engage in the interview
process. These individuals were excluded from my sample. In addition, because I do not speak Spanish and
because the substance of the interviews was sensitive enough that introducing a translator would have risked
complicating and even compromising the interview process, my sample also did not include dorm residents who
did not speak English. Despite this latter constraint, which might have seemed to threaten the racial diversity of
my sample, the racial profile of my interviewees wound up matching fairly closely the racial profile of K6G itself,
as indicated in the table below.
My sample

Make-up of K6G

Black

39.4%

36.5%

White

27.3

29.3

Latino

30.3

31.3

other

2.89

3.03

22. I consented and commenced the interview process with 33 subjects, but one subject proved to be
developmentally disabled, an incapacity that had not been evident during the consent process. In that case, I
terminated the interview without completing the questionnaire.
23. Interviewees were assigned random interview numbers. The interviews were recorded and later transcribed. Most interviews encompassed multiple audio-files, which were saved—and therefore transcribed—
alphabetically, with the sequence restarting each day. Citations to these interview transcripts will be referenced
hereinafter in the following manner: Int. # (Interviewee number), at file # (i.e. A–G) page # (transcript page
reference); e.g., Int. 46, at C3.
24. These recordings were subsequently transcribed. I thank the UCLA Academic Senate, the UCLA Dean’s
Office, Harvard Law School, and Georgetown University Law Center for their generous support of this costly
enterprise.
25. UCLA IRB # G07-01-106-03.

6

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

would most likely survive a constitutional challenge,26 it nonetheless puts government officials in the business of intruding into the most private and intimate details
of detainees’ lives in order to determine whether they meet the Department’s
definition of “homosexual.”27 Worse still, it engages state officers in a process of
openly labeling certain individuals as sexual minorities—with color-coded uniforms, no less.28
These concerns are serious ones, and point to admittedly troubling aspects of the
K6G program.29 They are, however, insufficient grounds to reject the enterprise.
Given the current state of the American carceral system—overcrowded,30 understaffed, volatile and often violent,31 frequently controlled from the inside by prison
gangs and other powerful prisoners32—there is at present no prospect for risk-free
reform. If K6G provides gay men and trans women in the L.A. County Jail with
safer and more humane conditions of confinement, the question we should be
asking is not whether the program ought to be allowed, but what it would take to

26. For further discussion of this point, see infra Part III.C.
27. The consent decree that created K6G explicitly stipulated “homosexual inmates” as the population to be
served. For further discussion on the issues this label raises, see infra Part II.C.2, III.B. Because those who are
transgender are typically readily apparent, the need to identify trans women does not carry the same unappealing
prospect of state intrusion into detainees’ private lives.
28. See discussion infra Parts II.B & III.A.
29. For further discussion, see id.
30. Over the past four decades, the population of America’s prisons and jails has soared from approximately
360,000 to over 2.3 million people. See THEODORE L. DORPAT, CRIMES OF PUNISHMENT: AMERICA’S CULTURE OF
VIOLENCE 55 (2007) (“In 1970, there were about 200,000 Americans in prison.”); BUREAU OF JUSTICE STATISTICS,
U.S. DEP’T OF JUSTICE, REPORT TO THE NATION ON CRIME AND JUSTICE 104 (2d ed. 1988) (reporting that the number
of jail inmates reached 160,863 in 1970); THE PEW CTR. ON STATES, ONE IN 100: BEHIND BARS IN AMERICA 2008, at
5 (2008) (“With 1,596,127 in state or federal prison custody, and another 723,131 in local jails, the total adult
inmate count at the beginning of 2008 stood at 2,319,258.”). As of January 1996, thirty-six states and the District
of Columbia were under court order to reduce overcrowding in some or all of their prisons. See NATIONAL PRISON
PROJECT, STATUS REPORT: STATE PRISONS AND THE COURTS 1 (1996). By the end of 2000, 310 prisons nationwide
were operating under such a court order, and “courts had placed population caps on forty-four prisons.” LYNN S.
BRANHAM, CASES AND MATERIALS ON THE LAW OF SENTENCING, CORRECTIONS, AND PRISONERS’ RIGHTS 621 (6th ed.
2002). And by the end of 2001, 33 states, the District of Columbia, and the federal system were housing prisoners
in jails and other facilities because of overcrowding. See PAIGE M. HARRISON & ALLEN J. BECK, U.S. DEP’T OF
JUSTICE, PRISONERS IN 2002, at 1 (2003), available at http://bjs.ojp.usdoj.gov/content/pub/pdf/p02.pdf. In
California, as of 2006, over 15,000 incarcerated people were sleeping in prison common areas “such as prison
gymnasiums, dayrooms, and program rooms.” Arnold Schwarzenegger, Governor, State of Cal., Prison Overcrowding State of Emergency Proclamation (Oct. 14, 2006). In the fall of 2006, California Governor Arnold
Schwarzenegger declared a “Prison Overcrowding State of Emergency,” noting that overcrowding creates an
“increased, substantial risk of violence, and greater difficulty controlling large inmate populations.” Id.
31. See, e.g., Sharon Bernstein & Richard Winton, Sheriff to Move Violent Inmates, L.A. TIMES, Feb. 8, 2006,
at A1; Megan Garvey & David Pierson, More Rioting Erupts at Jails, L.A. TIMES, Feb. 9, 2006, at A1; Stuart
Pfeifer, Inmate’s Death Is Ruled a Homicide, L.A. TIMES, Dec. 12, 2006, at B1; Stuart Pfeifer, Crowding Problems
Grow as Jail Sits Vacant, L.A. TIMES, Oct. 16, 2007, at B3 (reporting 15 people killed in Los Angeles County jails
since 2000).
32. See Jonathan A. Willens, Structure, Content and the Exigencies of War: American Prison Law After
Twenty-Five Years 1962–1987, 37 AM. U. L. REV. 41, 55–61 (1987) (describing the ways gangs dominate prison
life); Peter M. Carlson, Prison Interventions: Evolving Strategies to Control Security Threat Groups, 5 CORR.
MGMT. Q., Winter 2001, at 10 (stating that gangs present an enormous problem for inmate management).

2011]

STRATEGIC SEGREGATION

7

maintain the protection it provides while minimizing the dangers posed whenever
the state authorizes differential treatment on the basis of identity.
The more vexing question is whether the K6G program is one that other
jurisdictions ought to seek to emulate. As I argue in what follows, given its
success, this model should be available as a tool in the toolkits of officials seeking
to reduce the incidence of victimization in their facilities. Whether prison administrators elsewhere will find it an appropriate model for their jurisdictions is an open
question. Carceral facilities differ widely as to the profiles of their populations (in
size, racial and ethnic mix, gang culture, etc.), mission (whether jail, prison,
juvenile detention, etc.), institutional culture, and physical structure and design.
These differences among facilities mean that, as with any program goal, one size
will not fit all.33 Moreover, given the risks of a segregated approach like K6G,34
many jurisdictions may conclude that the possible harms outweigh the benefits.
Yet one thing is clear: even where the K6G model seems, as in L.A. County, to
meet the needs of a given institution, this approach can never be sufficient.
Although K6G succeeds in keeping its residents relatively safe, its admission
criteria are sorely underinclusive, excluding even people who, although neither
gay nor trans, are nonetheless liable to victimization in GP.35 Plainly, all detainees
known to face a risk of abuse in custody must be protected.36 The key policy
question is whether there may be grounds for dividing K6G’s target populations
even from other vulnerable groups. The National Prison Rape Elimination Commission (the Commission), created by Congress through the Prison Rape Elimination
Act of 2003 (PREA),37 made recommendations in its final report suggesting a

33. Cf. CHAD R. TRULSON & JAMES W. MARQUART, FIRST AVAILABLE CELL: DESEGREGATION OF THE TEXAS
PRISON SYSTEM 215 (2009) (noting that “[n]o two prison systems are alike” and, therefore, that racial
desegregation “must be a system-specific process” taking into account “the total environment in which a
particular prison system is situated”).
34. See infra, Part III.A.
35. See supra note 7.
36. The state’s legal obligation to protect vulnerable prisoners from sexual assault stems both from the Eighth
Amendment to the United States Constitution, see Farmer v. Brennan, 511 U.S. 825 (1994), and from federal law,
see the Prison Rape Elimination Act (PREA) of 2003, 42 U.S.C. §§ 15601–09 (2003); National Standards, supra
note 7. See also Sharon Dolovich, Cruelty, Prison Conditions, and the Eighth Amendment, 84 N.Y.U. L. REV. 881,
910–23 (2009) (arguing that the state has a moral obligation to protect prisoners from serious physical and
psychological harm).
37. See 42 U.S.C. § 15606. Among other things, Congress directed the Commission to “carry out a
comprehensive legal and factual study of the penalogical [sic], physical, mental, medical, social, and economic
impacts of prison rape in the United States,” 42 U.S.C. § 15606 (d)(1); hold public hearings on these issues, 42
U.S.C. § 15606 (g); and submit a report to the Attorney General and the Secretary of Health and Human Services
containing “recommended national standards for enhancing the detection, prevention, reduction, and punishment
of prison rape.” 42 U.S.C § 15606 (e)(1). The Commission submitted its final report in June 2009. See
COMMISSION REPORT, supra note 8. PREA also directed the Attorney General to promulgate a “final rule adopting
national standards for the detection, prevention, reduction, and punishment of prison rape.” 42 U.S.C. § 15607
(a)(1). In January 2011, the Attorney General issued a Notice of Proposed Rulemaking and called for notice and
comment. See National Standards, supra note 7, at 6248. The comment period closed in April 2011, see id., and
the final rule is expected in 2012.

8

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

negative answer to this question.38 Concerned about the “demoralizing and
dangerous” effects of the L.A. County model,39 the Commission advanced an
approach that did not distinguish among at-risk groups.40 This unified strategy has
much to commend it, not least that it mitigates many of the troubling aspects of
state-sponsored identity-based segregation. For this and other reasons,41 the
Commission’s approach will in most cases be preferable. Still, there may be reason
to regret the widespread adoption of a unified model, which could come at the cost
of some of the more humane and appealing aspects of life in K6G and result in a
direct loss of some of its benefits.42
Certainly, no single strategy will be without its dangers and drawbacks. Prisons
are an ugly business, and the problems they pose—including prison rape—admit
of no easy fix. Indeed, to await such a fix would be to consign some of the most
vulnerable people behind bars to the worst forms of suffering and abuse. K6G
merits attention not because it is a perfect program, but because with it, L.A.
County has created a relatively safe space for people who would otherwise be at
great risk of victimization. Understanding how the program works day-to-day
helps to explain its remarkable success. Equally important, it sheds light on the
causes of prison sexual violence in general, as well as what, given the current
realities of the American carceral system, is required to guard against it.
K6G also bears a close look for a further, unexpected reason: there is a lot to
learn from this unit about incarceration more broadly. Although K6G shares many
of the features of any custody situation, life in the K6G dorms differs in notable
ways from life in GP. Through this contrast, a study of K6G exposes some
important and troubling aspects of the carceral enterprise, including the role of the
hypermasculinity imperative in constructing the social order in men’s prisons; the
role of gangs in defining the carceral experience; the relationship between racial
tensions and the fear of victimization in custody; the social services function of the
American carceral system; and even what humane prison conditions might look
like. It turns out, in short, that people who care about understanding prisons and
making them safer and more humane may have much to gain from studying a
small, unorthodox, and entirely unrepresentative unit in a massive, overcrowded,

38. See COMMISSION REPORT, supra note 8, at 80.
39. Id.
40. The Commission’s proposed standards provided that all inmates be screened on arrival (and at “all
subsequent classification reviews”) to assess their risk of being victimized or a victimizer. Id. at 217. To ensure
accurate assessments, these determinations were to be individualized. For the criteria stipulated for consideration
when “screen[ing] male inmates for risk of victimization,” see note 9 above. By contrast to the long list of criteria
provided in the case of male inmates, the Commission included just two criteria to consider in “screen[ing] female
inmates for risk of sexual victimization: prior sexual victimization and the inmate’s own perception of
vulnerability.” Id. DOJ’s proposed standards do not distinguish between men and women, and add just one further
consideration to the list: “[w]hether the inmate is detained solely on civil immigration charges.” National
Standards, supra note 7, at 6280–81 (§ 1154.41 (c)(10)).
41. See discussion infra Part III.B.
42. See, e.g., infra Part II.D.2 (discussing the gang politics in GP and the absence of such politics in K6G).

2011]

STRATEGIC SEGREGATION

9

decaying, dangerous, and otherwise wholly traditional carceral facility.
This Article proceeds as follows. Part I explains the special vulnerability of gay
men and trans women to sexual victimization in men’s carceral facilities, focusing
in particular on the prison culture of hypermasculinity in which men struggle to
prove their manhood, often by sexually victimizing those culturally defined as
female. Part II examines L.A. County’s K6G program. It describes the mechanisms of its operation, including the policies by which K6Gs are kept separate
from GPs and the classification process by which gay men and trans women are
identified for placement in the unit. This Part closes by making the case for the
program’s success in keeping its target populations relatively safe. In the course of
doing so, it describes the gang politics that govern life in L.A. County’s GP dorms
and the role these politics play in prompting even some men who know themselves
not to satisfy K6G’s classification criteria to pretend to be gay to try to get access to
the unit. Part III then addresses three objections that might be raised to the K6G
enterprise. Part III.A considers the anti-segregationist objection, which would
condemn as “demoralizing and dangerous” any official policy of separating groups
along identity lines.43 Although this section finds this objection insufficient to
justify abandoning the K6G project, it identifies a number of concerns on which
the anti-segregationist view sheds light, which anyone committed to the safe
housing of K6G’s target populations must take seriously. Part III.B turns to a
cluster of objections that challenge K6G’s admissions criteria as underinclusive.
Some of these objections are shown to be unpersuasive, but at least one—the
concern that K6G still leaves unprotected many vulnerable individuals who are
neither gay nor trans—points to the limits of the K6G model in protecting all
at-risk prisoners from sexual victimization. The most salient question proves to be
whether institutions ought to have a single unit for all at-risk prisoners, or whether
it may sometimes make sense to separate K6G’s target populations even from
other vulnerable groups. Part III.B ultimately argues that a unified approach such
as that recommended by the Commission represents the better default option,
although it emphasizes the dangers that would still attend such a strategy. At the
same time, Part III.B cautions against any efforts to dismantle K6G and maintains
that wide differences among carceral facilities mean that under some circumstances, two separate segregation units may represent the wiser option. Part III.C
addresses a final objection: that, even should prison administrators in L.A. County
or elsewhere wish to follow a K6G model, they would be precluded from doing so
on constitutional grounds. As Part III.C shows, this notion is based on a misunderstanding of governing law, specifically the 2005 case of Johnson v. California.44
Part IV concludes by identifying the factors that account for the relative success of

43. COMMISSION REPORT, supra note 8, at 80.
44. 543 U.S. 499 (2005).

10

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

K6G, and considers the extent to which other jurisdictions might be able to
replicate that success by implementing K6G-type programs of their own.
The aim of this project is in part ethnographic. In this Article, I explore the
strategy developed in L.A. County for keeping gay men and trans women separate
from GP. In a companion piece, I describe daily life in K6G and contrast it with the
experience of life in GP.45 Especially given the subject matter, there is an inherent
danger in such an undertaking—that of seeming to endorse and thus to legitimate
what is being described. Certainly, there is much to disturb in what I have to report.
I describe a prison sexual culture in which the strong prey on the weak and gain
status and power through the domination and abuse of fellow human beings. I
describe a set of racial dynamics the governing principles of which are totally
antithetical to prevailing socio-political norms of racial equality and mutual
respect. And I describe state practices that enable and even collude in the
perpetuation of this culture and these dynamics. Even the official steps taken to
alleviate the worst effects of these conditions may strike some readers as both
sorely insufficient and objectionable in themselves. To make matters worse, I offer
validation—albeit equivocal—of existing strategies in the Jail, which may seem
only further to legitimate the cultural and institutional dynamics that necessitate
those strategies. To the extent that my own motives and commitments are
genuinely in question, all I can do is state up front that I too regard with abhorrence
much of what I describe and fervently wish things were otherwise. As I see it,
however, no widespread change will occur unless the reality of what goes on in
detention is brought to light and understood in all its complexity and ugliness.
But the charge of legitimation has a political dimension to which assertions of
good faith are an insufficient answer. Indeed, from this political perspective,
assertions of good faith reflect a naı¨vete´ that can be as dangerous as open collusion.
The objection here is that by engaging in a conversation about reform, one risks
normalizing the cultural dynamics—in this case, the prison culture of hypermasculinity—that policy initiatives like K6G are designed to address, thereby only
further entrenching a penal system that is illegitimate at its core.46
There may well be something to this concern. It is arguable, for example, that
once decades of judicial pressure forced states like Alabama, Louisiana, and Texas
45. See Sharon Dolovich, Two Theories of the Prison: Accidental Humanity and Hypermasculinity in the L.A.
County Jail (draft copy on file with the author).
46. At its most extreme, this objection leads into absurdity, calling to mind arguments made by hard-line
Bolsheviks in the early twentieth century against worker protections such as a minimum wage, a forty-hour
work-week, and improvements to workplace safety on the grounds that any move to improve the daily lives of
paid laborers would only dull their appetite for revolution. See, e.g., VLADIMIR ILYICH LENIN, IMPERIALISM, THE
HIGHEST STAGE OF CAPITALISM 22 (Resistance Books 1999 (1917)) (arguing that imperialism provides capitalists
with “superprofits” from the cheap labor and raw materials available in colonized countries, allowing them to
“bribe” workers at home with higher wages, a shortened work-week, etc., thereby forestalling revolution in the
most developed economies). But one could also see this challenge in a less extreme light, as a wholly reasonable
worry that to tinker around the edges of a fundamentally corrupt system, thereby mitigating its worst offenses,
risks only strengthening the seeming legitimacy of the whole carceral enterprise.

2011]

STRATEGIC SEGREGATION

11

to address conditions of the sort that led an Arkansas district court to characterize
that state’s prison system as “a dark and evil world completely alien to the free
world,”47 it became more difficult for claims as to the harms incarceration inflicts
to gain traction.48 This is a risk of reform efforts in any context, and everyone must
make their own calculations as to the right course. For myself, the alleviation of
immediate suffering is the greater imperative—hence the instant undertaking, and
my endorsement in what follows of imperfect half-measures.
I. PRISON RAPE, HYPERMASCULINITY, AND THE FEMINIZATION OF VICTIMS
To understand why gay men and trans women are at a heightened risk of
victimization behind bars, it is first necessary to understand the phenomenon of
prison rape more generally. The notion of “prison rape” typically summons images
of the violent gang rape of a lone defenseless prisoner. And indeed, this horror is
often enough the experience of vulnerable people behind bars. But such a rape, or
the mere threat of it, also functions as a disciplining mechanism,49 the means by
which weaker individuals are forced into a relationship in which they occupy a
subordinate and even abject position vis-a`-vis stronger prisoners, who effectively
“own” them.50 Among other indignities, this slave-like status routinely demands
that weaker prisoners accede to sexual penetration by their owners or anyone else
to whom they are directed. Even when sexual access is not physically forced, it is

47. Hutto v. Finney, 437 U.S. 678, 681 (1978) (quoting Holt v. Sarver, 309 F. Supp. 362, 381 (E.D. Ark. 1970)).
48. This effect is arguably evident in Rhodes v. Chapman, a 1981 Eighth Amendment case challenging the
practice of double-celling (ie, housing two men in tiny cells (here, 55 square feet) designed for a single person) in
the Southern Ohio Correctional Facility (SOCF). See Rhodes v. Chapman, 452 U.S. 337 (1981). In Rhodes, the
Supreme Court rejected this constitutional claim, despite the unanimous conclusion of expert witnesses that “a
long-term inmate must have to himself, at the very least, 50 square feet of floor space . . . in order to avoid serious
mental, emotional, and physical deterioration.” Id. at 371 (Marshall, J., dissenting). Judging from the majority
opinion, what most moved the Court was the trial court’s finding that SOCF, less than ten years old, was
“unquestionably a top-flight, first-class facility.” Id. at 341. Even Justice Brennan, whose concurrence emphasized
the continued obligation of the federal courts to enforce the Eighth Amendment in the prison conditions context,
extolled, among other things, the relative newness of the facility; the adequacy of the furnishings, plumbing,
lighting, and food; the availability of contact visits; and the “modern, well-stocked library.” Id. at 365 (Brennan,
J., concurring). SOCF was arguably among the fruits of the federal district court-led prison reform movement of
the late 1960s and 1970s, which mitigated many of the worst features of American prisons, not only in the South,
but around the country. See MALCOLM M. FEELEY & EDWARD L. RUBIN, JUDICIAL POLICY MAKING AND THE
MODERN STATE 51–143 (2000) (describing in detail federal court cases that led to extensive prison reforms in a
number of jurisidictions, including Arkansas and Texas). Given SOCF’s relative humanity compared with the
prisons that went before it, it would be no wonder if the Court were to hesitate to condemn conditions in that
facility, even were they certain to cause some inmates “serious mental, emotional and physical deterioration.”
Rhodes, 452 U.S. at 371 (Marshall, J., dissenting). Something of the same effect may be feared in the prison rape
context; if this worst of all penal abuses is mitigated, will it blunt the political will to condemn the carceral project
for its many other, albeit less obviously horrific, harms?
49. Craig Haney, The Perversions of Prison: On the Origins of Hypermasculinity and Sexual Violence in
Confinement, 48 AM. CRIM. L. REV. 121, 129 (2011) (reporting that “countless prisoners [have described to him]
the way they can ‘feel’ the threat of rape ‘in the air’ around them, or have heard frightening accounts of it taking
place, even if they have not seen it themselves or been directly victimized”).
50. See infra note 55.

12

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

nonetheless often experienced by the victim as a profound violation and thus must
also be included along with other forms of sexual victimization under the umbrella
term “prison rape.”51
As in society in general, sexual assault in men’s prisons can take many forms,52
from verbal harassment and unwanted touching to coerced sexual contact and
forcible rape.53 In the prison context, however, verbal sexual harassment is often
used in a strategic way, to alert the target that he has been singled out for more
serious sexual victimization54 and may soon face a forcible rape. In many cases,
that rape comes quickly, marking the victim—now labeled a “punk”—as the
property of their attacker, to be kept in a state of what amounts to sexual slavery for
the use of the assailant and his associates, or sold to other prisoners for this
purpose.55 In other cases, the harassment serves as advance warning of an attack to
come, giving targets a short time to take defensive action. Prisoners who are put on
notice in this way—the relatively lucky ones—have two options. Those able to
defend themselves can choose to fight back against the attack when it comes. A
person who successfully repels efforts at forcible rape—especially if the attackers
are seriously injured in the attempt—may succeed in deterring others from any

51. As one incarcerated person, quoted in a Human Rights Watch report, put it, “[a] prisoner that is engaging in
sexual acts, not by force, is still a victim of rape because I know that deep inside this prisoner do not want to do the
things that he is doing but he thinks that it is the only way that he can survive.” HUMAN RIGHTS WATCH, supra note
2, at 52; see also Stephen “Donny” Donaldson, A Million Jockers, Punks, and Queens, in PRISON MASCULINITIES
118, 125 (Don Sabo, Terry A. Kupers & Willie London eds., 2001) (“From the typical punk’s point of view, none
of his passive sexual activities is truly voluntary, since, if he had his own way he would not need to engage in
them.”). The Prison Rape Elimination Act (PREA) has been criticized for focusing on violent rape at the expense
of the coerced sex that more typically occurs in the context of protective pairings. See, e.g., Alice Ristroph, Sexual
Punishments, 15 COLUM. J. GENDER & L. 139, 146, 175–76 (2006). What these critiques overlook is that reducing
the incidence of violent assault is crucial to protecting vulnerable prisoners from all unwanted sex, since without
the threat of violent rape, the pressure to enter into a protective pairing would necessarily be reduced.
52. The dynamics described in what follows do not exist to the same extent in all prisons and jails. But they are
sufficiently common and recognizable to affect life to some degree in most if not all men’s carceral facilities, and
thus to allow for a generalized description.
53. See James E. Robertson, Cruel and Unusual Punishment in United States Prisons: Sexual Harassment
Among Male Inmates, 36 AM. CRIM. L. REV. 1, 9–11 (1999) (explaining that inmate-on-inmate sexual harassment
in prison ranges from verbal harassment to sexual propositions, sexual coercion—which “occurs when an inmate
pressures another inmate to pay a debt via a sexual act”—and physical harassment, which can include “[k]issing,
[t]ouching, or [f]ondling [i]ntimate [b]ody [a]reas”).
54. Id. at 15 (“Sexual harassment of another inmate . . . communicates aggressive intentions.”).
55. See Donaldson, supra note 51, at 119 (explaining that punks are, “for all practical purposes, slaves [who]
can be sold, traded, and rented or loaned out at the whim of their ‘Daddy’”); see also JAMES GILLIGAN, VIOLENCE:
REFLECTIONS ON A NATIONAL EPIDEMIC 180 (1997) (explaining that after a prisoner is raped, he “becomes a slave in
the fullest sense of the term.”); Adam Liptak, Ex-Inmate’s Suit Offers View Into Sexual Slavery in Prisons, N.Y.
TIMES, Oct. 16, 2004, at A1 (describing one prisoner’s experience in a Texas prison, where for eighteen months, as
a sex slave to a gang, he was “‘forced into oral sex and anal sex on a daily basis,’” “bought and sold[,]” and
“rented . . . out” for sex for the benefit of the gang); HUMAN RIGHTS WATCH, supra note 2, at 71–72 (“Victims of
prison rape, in the most extreme cases, are literally the slaves of the perpetrators. . . . They are frequently ‘rented
out’ for sex, sold, or even auctioned off to other inmates . . . .”).

2011]

STRATEGIC SEGREGATION

13

subsequent efforts to “punk” him. Sometimes the target may have to fight more
than once, at risk of great physical harm.56
Those on notice of an attack who lack the capacity to protect themselves may
instead choose to “hook up” with a more powerful prisoner in a protective pairing.
As already noted, this protection is not free: the cost to the weaker inmate (or
“catcher”57) includes providing his protector (or “Daddy”58) with regular sexual
access as well as other “wifely” services, “such as doing laundry, making the bunk,
keeping the cell clean, and making and serving coffee.”59 Catchers may also be
expected to provide sexual access to the friends or associates of their protectors,
and may be rented out for this purpose—i.e., prostituted—with their protector
keeping the profits. Such pairings can thus consign the weaker person to an
experience of serial sexual violation by his owner or others. To enter into such a
relationship means that a person is no longer his “own man,” but will, for all
intents and purposes, belong to someone else for the duration of his incarceration.60 In the worst cases, the subordinate party will be consigned to sexual
slavery.61 Given these stakes, vulnerable prisoners may try to resist the pressure to
enter a protective pairing. But typically, any such resistance does not last long. As
Stephen Donaldson chillingly put it, “[u]sually, a rape or two is sufficient to
persuade an unattached catcher to pair off as soon as possible.”62
Whatever method is used to force the victim to submit to sexual penetration,
once this aim is accomplished, the victim is redefined in the prison culture as a
“punk.”63 This experience of being “punked” will signify only the beginning of his

56. Those who fight back under such circumstances also risk disciplinary sanction and further criminal
charges. For an apt critique of this catch-22, see James E. Robertson, “Fight or F . . .” and Constitutional Liberty:
An Inmate’s Right To Self-Defense When Targeted By Aggressors, 29 IND. L. REV. 339 (1995).
57. Donaldson, supra note 51, at 119–20.
58. Id. at 118–19. Other ethnographic accounts label the men in this position as “wolves” or “jockers.” See,
e.g., Peek, supra note 7, at 1227.
59. Donaldson, supra note 51, at 120.
60. It bears noting that, notwithstanding the often coerced nature of these relationships, genuine feelings of
affection and even love may arise between the parties. See T.J. PARSELL, FISH: A MEMOIR OF A BOY IN A MAN’S
PRISON 112–13 (2006) (recounting an episode during his incarceration in a Michigan prison, in which “Slide
Step,” the stronger inmate with whom he had entered into a protective pairing relationship, told Parsell of his
“having [developed] feelings for him”).
61. See Wilbert Rideau, The Sexual Jungle, in LIFE SENTENCES 73, 75 (Wilbert Rideau & Ron Wikberg eds.,
1992) (explaining that “[t]he act [of rape] redefines [the victim] as a ‘female’ in this perverse subculture, and he
must assume that role as the ‘property’ of his conqueror or whoever claimed him and arranged his emasculation.”).
62. Donaldson, supra note 51, at 121. This assumes, of course, that the victim of such a rape retains the power
to choose, and does not automatically come to be regarded as the property of his rapist.
63. See Peek, supra note 7, at 1226. According to some accounts of the prison sexual subculture, those
people—generally trans women or gay men—who voluntarily assume a female identity are known not as punks,
but as “queens,” and enjoy a slightly higher status in the prison pecking order, “probably because [queens] are
considered desirable sexual partners, and because punks are condemned for lacking the courage to defend
themselves and their masculinity.” Id. at 1228. But, as a practical matter, the difference in status means little, as

14

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

sexual victimization64; as even the federal courts have recognized, “[o]nce raped,
an inmate is marked as a victim and is subsequently vulnerable to repeated
violation.”65 “Through the act of rape,” in other words, “the victim is redefined as
an object of sexual abuse,” and will in all likelihood retain that role through the
entirety of his incarceration.66
To outside observers, prison life can seem like a dog-eat-dog state of nature. Yet
the sexual victimization of weaker prisoners by more powerful ones in fact takes
place within a highly organized social system, in which power is allocated and
exercised along surprisingly conventional lines. As with intimate relationships in
society in general, the defining scripts are gendered: in men’s prisons, as in the free
world, men dominate women.67
Given that the state rigidly divides prison populations along gender lines, with
men and women typically housed in entirely separate facilities, the claim that rape
in men’s prisons is a gendered phenomenon may seem an odd one. But “male” and
“female” are not essential truths. They are instead culturally ascribed aspects of
social identity. That in men’s prisons there are no “women” as conventionally
defined does not mean that there are no gendered relationships. It simply means
that for purposes of this particular cultural model, some prisoners must be
designated as female.
Why this pressure towards a gender binary? Here, an observation made by
Gresham Sykes in his classic prison ethnography, Society of Captives, may shed
some light. Sykes observed that “[a] society composed exclusively of men tends to
generate anxieties in its members concerning their masculinity.”68 In such a
context, the “competition” among men that is endemic in American culture reaches
a heightened pitch, so that its typical features become exaggerated. These features,
which Frank Rudy Cooper associates with the “hegemonic model of U.S. masculinity,”69 are both familiar and destructive even in society more generally. As

both punks and queens are at the mercy of their “protectors,” obliged at the risk of physical harm to do their
bidding. See id. at 1228–29.
64. See Ristroph, supra note 51, at 152–53 (2006) (explaining that “[a] rape [in prison] is often ‘the first act in a
lengthy drama of conquest and control.’” (quoting Ian O’Donnell, Prison Rape in Context, 44 BRIT. J.
CRIMINOLOGY 241, 244 (2004))).
65. Schwenk v. Hartford, 204 F.3d 1187, 1203 n.14 (9th Cir. 2000) (noting that “[t]he victims of these attacks
are frequently called female names and terms indicative of gender animus like ‘pussy’ and ‘bitch’ during the
assaults and thereafter”) (emphasis added).
66. HUMAN RIGHTS WATCH, supra note 2, at 54.
67. See CATHERINE MACKINNON, TOWARD A FEMINIST THEORY OF THE STATE 128 (1st ed., 1989). As MacKinnon
succinctly puts it, “Man fucks woman; subject verb object.” Id. at 124. See also Jeannie Suk, Redistributing Rape,
48 AM. CRIM. L. REV. 111 (2011) (exploring the implications of this gendered structure for both K6G and GP).
68. Robertson, supra note 53, at 13 (1999) (quoting GRESHAM SYKES, SOCIETY OF CAPTIVES: A STUDY OF A
MAXIMUM SECURITY PRISON 71 (First Princeton Classic ed. 2007)). As Robertson observes, “[i]mprisonment
represents more than a loss of freedom; it also diminishes you as an adult male.” Id. at 12.
69. See Frank Rudy Cooper, “Who’s the Man?”: Masculinities Studies, Terry Stops, and Police Training, 18
COLUM. J. GENDER & L. 671, 687 (2009); see also MICHAEL S. KIMMEL, MANHOOD IN AMERICA: A CULTURAL
HISTORY (1996).

2011]

STRATEGIC SEGREGATION

15

Cooper describes it, the American man on this hegemonic model has four distinct
features. First, he is “concerned with how other men rate him” as to his own
masculinity level.70 Second, he is “chronically insecure that he has not sufficiently
proved that he is as masculine as he should be.”71 Third, he is driven to compete
with other men “to outdo [them] in collecting indicia of manhood.”72 Fourth and
finally, men in this competition must “[repudiate] that model’s contrast figures,”
among them “women [and] gays.”73 As Cooper explains, the hegemony of this
model manifests itself in a compulsion to “hypermasculinity” on the part of those
who are “denied the stature of the normative man . . . .”74 Displays of hypermasculinity—“the exaggerated demonstration of the features associated with maleness”—
compensate for a failure to “meet the masculine cultural ideal.”75 Cooper offers
“weight-lifting, bragging about sexual exploits, and homophobic jokes” as examples of “hypermasculine behaviors” in society at large.76
As those familiar with prison life will readily recognize, this account translates
directly into the prison context,77 where the imperative exists among incarcerated
men to “be hard and tough, and [not] show weakness.”78 The archetype of the
stoic, weightlifting, muscle-bound prisoner has its origins in this dynamic.79 In
prison, however, displays of strength and toughness alone are not sufficient proof
of masculinity for men anxious about others’ perceptions of their gender identity.
As in society in general, the construction of identity in prison is relational: claims
to masculinity are “only meaningful in relation to constructions of femininity.”80
Would-be men must therefore struggle against and ultimately vanquish the
seemingly feminine in themselves and—to clearly make the point—in others as
well.81 In the absence of other socially productive means to prove their manhood

70. Cooper, supra note 69, at 687.
71. Id. at 688 (“Manhood is a relentless test of how close you are to the ideal. . . . [M]en are constantly
suffering from anxiety that other men will unmask them as insufficiently manly.”).
72. Id.
73. Id. at 689. Cooper observes that “racial minorities” also function as contrast figures for a hegemonic
masculinity for which the standard is not only middle-class, “early middle-aged,” and heterosexual, but also
white. Id.
74. Id. at 691.
75. See id.
76. Id. at 692. Cooper also notes that, “[b]ecause homoerotic desire is depicted as feminine desire, the
repudiation of homosexual men is a necessary component of hegemonic masculinity.” Id. at 690.
77. This same exaggerated masculinity imperative also defines life among the street gangs of South Central
L.A. See CRIPS AND BLOODS: MADE IN AMERICA (PBS television broadcast May 12, 2009) (interviewing gang
members attesting to the need to purge their affect of any signs of feeling or sensitivity in order to protect
themselves from being judged as weak and consequently vulnerable to victimization by other gang members).
78. Derrick Corley, Prison Friendships, in PRISON MASCULINITIES 106 (Don Sabo et al. eds., 2001).
79. Don Sabo, Doing Time, Doing Masculinity: Sports and Prison, in PRISON MASCULINITIES 61, 65 (Don Sabo
et al. eds., 2001). Indeed, in men’s prisons, muscles are arguably “the sign of masculinity.” Id. (quoting BARRY
GLASSNER, BODIES: WHY WE LOOK THE WAY WE DO (AND HOW WE FEEL ABOUT IT) 192 (1988)).
80. SUE LEES, RULING PASSIONS: SEXUAL VIOLENCE, REPUTATION AND THE LAW 105 (1997).
81. See Cooper, supra note 69, at 690 (“Most of all, masculinity is the repudiation of femininity.”).

16

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

(business, politics, family, “cars and the like”82), the sexual domination of women
becomes the method of choice.83 In society in general, the hypermasculinity
imperative to conquer and repudiate the feminine frequently motivates rape,
sexual harassment, domestic violence, and other forms of violence against women.84
In the prison, those men seeking to prove their masculinity vie for possession of
weaker inmates—the “women” in this social system—whose utter subordination
to them, known to include ongoing sexual access, stands as public proof of their
masculine power. In this culture, the performance of rape—the sexual penetration
of another inmate defined as female—is a way to shore up the rapist’s own claim to
maleness and, thus, his status and power in the prison hierarchy. As one prison
official put it:
[s]ex and power go hand-in-hand in prison . . . . Deprived of the normal
avenues, there are very few ways in prison for a man to show how powerful he
is—and the best way to do so is for [him] to have a [sex] slave, another who is
in total submission to him.85

In this social system, anyone who can be perceived as at all feminine is assigned
the subordinate “woman’s” role. As Christopher Man and John Cronan note, “[i]n
many ways, an inmate’s aura of femininity can overlap with his age and size, as
younger and smaller inmates are often perceived as more feminine than older
inmates.”86 The same is true of anyone who comes across as weak and defenseless;

82. James E. Robertson, A Clean Heart and an Empty Head: The Supreme Court and Sexual Terrorism in
Prison, 81 N.C. L. REV. 433, 441 (2003).
83. As feminist theorists have argued, in society in general, rape or the threat of rape is a central mechanism by
which men—especially those men insecure about their own relative social position—reaffirm their masculine
self-image. See Man & Cronan, supra note 8, at 148 (explaining that rape can symbolize “the act of putting
women ‘back in their place,’” and that “rapists are often men who feel threatened by the fear that women or a
particular woman may achieve equality or superiority over them”). In a setting in which prisoners lack other
effective means for this reaffirmation, rape or the threat thereof becomes the “premier” means for demonstrating
one’s “masculinity and strength.” Robertson, supra note 53, at 14.
84. See supra note 83.
85. Rideau, supra note 61, at 75 (quoting C. Paul Phelps, then-secretary of the Louisiana Department of
Corrections). That the sexual penetration of one man by another could work to reinforce the masculine image of
one of the participants may seem strange given that, generally speaking, men who have sex with other men are
regarded in mainstream culture as homosexual themselves—and thus equally perceived as not “real men.” But
this puzzle is resolved once it is recognized that in prison, those who take the passive sexual position or who
perform oral sex on other men are regarded in the prison not as men at all. See Man & Cronan, supra note 8, at
167–68 (“By raping someone perceived as feminine, an inmate can assert his dominance without thinking of
himself as a homosexual and, thereby, securing his male identity. Thus, the inmate redefines seemingly
homosexual activity as a heterosexual activity.”). As was explained by Human Rights Watch in its 2001 report
about male rape in U.S. prisons, most prison rapists continue to view themselves as heterosexual since the victim
is “substituting” for a woman. HUMAN RIGHTS WATCH, supra note 2, at 70 (“[T]he crucial point is not that the
[aggressor is] having sex with a man; instead it is that they are the aggressor, as opposed to the victim—the person
doing the penetration, as opposed to being the one penetrated. Indeed, if they see anyone as gay, it is the victim
(even where the victim’s clear sexual preference is for heterosexual activity).”).
86. Man & Cronan, supra note 8, at 166.

2011]

STRATEGIC SEGREGATION

17

hence the vulnerability of the shy,87 the physically or mentally disabled, or anyone
who is a first-time or non-violent offender and, thus, unlikely to be savvy to the
ways of the prison or to have friends or allies on the inside to protect him. Prisoners
who are weak and defenseless are vulnerable to rape and other forms of sexual
victimization, not merely because of their weakness, but because they are regarded
as available for emasculation—or, in prison parlance, for being “turned out”—and,
thus, transformed into women for purposes of life in the prison.88 No wonder men
in prison work so strenuously to “be hard and tough, and [not] show weakness.”89
Viewed in this light, the felt imperative to perform a gendered ideal of “hypermasculinity” is a rational response. Desperate not to be victimized themselves, male
prisoners do all they can to avoid any behaviors that might suggest qualities
associated with femininity: passivity, expressing emotion, sensitivity, kindness,
etc. In short, fear of rape motivates displays of hypermasculinity among prisoners
wishing to avoid being “turned out” themselves.90
This brings us to the special vulnerability of gay men and trans women in this

87. “[I]nmates who look scared, shy, or nervous face immediate danger because they exude signs of weakness.
Similarly, inmates who talk too much, thinking that it is the only way to fit in, may be perceived as nervous and
become targets of rape.” Id.
88. This effect can be accomplished through rape, which may be violent but which need not be, “as it is easy
enough for several prisoners to overcome a single victim simply by holding him in place.” HUMAN RIGHTS WATCH,
supra note 2, at 86. Alternatively, aggressors seeking to force the target into the passive sexual role may resort to a
method referred to by some—one hopes ironically—as “seduction.” See Helen M. Eigenberg, Correctional
Officers and Their Perceptions of Homosexuality, Rape, and Prostitution in Male Prisons, 80 PRISON J. 415, 420
(2000). Here, the aggressor seeks out a possible target—someone who appears alone and nervous or scared, and
who therefore may be too trusting of anyone who shows him kindness. The aggressor approaches the target,
offering cigarettes, food, drugs, alcohol, or anything the target may be induced to accept. The aggressor will then
demand payment for the proffered goods from the target, despite the fact that the initial offer was made without a
hint of the need for repayment. When no payment is forthcoming, the aggressor and any co-conspirators in the
plot will threaten the target with physical violence unless he allows them sexual access. See id. at 420; see also
HUMAN RIGHTS WATCH, supra note 2, at 88 (noting that this process is “best called coercion” and explaining that
“the victim is usually tricked into owing a favor”).
89. Corley, supra note 78, at 106.
90. As Wilbert Rideau puts it in a chilling passage:
The act of rape in the ultramasculine world of prison constitutes the ultimate humiliation visited
upon a male, the forcing of him to assume the role of a woman. It is not sexual and not really
regarded as “rape” in the same sense that society regards the term. In fact, it isn’t even referred to
as “rape.” In the Louisiana penal system, both prisoners and personnel generally refer to the act as
“turning out,” a nonsexual description that reveals the nonsexual ritualistic nature of what is really
an act of conquest and emasculation, stripping the male victim of his status as a “man.” The act
redefines him as a “female” in this perverse subculture, and he must assume that role as the
“property” of his conqueror or whoever claimed him and arranged his emasculation.
Rideau, supra note 61, at 75.
This set of cultural norms is reinforced not only by prisoners, but also by guards. Prisoners at risk of rape who
seek protection from correctional officers (COs) often report receiving the age-old advice of “fuck or fight”—that
is, to fight their aggressors or suffer the consequences. Robertson, supra note 53, at 33. When, for example,
Roderick Johnson sought protection from COs from the repeated rapes he suffered as the “sexual slave” of prison
gangs, he was “repeatedly told . . . that he either had to fight off his attackers or submit to being used for sex.”
Johnson v. Johnson, 385 F.3d 503, 513 (5th Cir. 2004) (reporting that classification officers allegedly made

18

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

system.91 In the twisted logic of the prison sexual hierarchy, the act of rape—of
forcing a weaker prisoner into the sexually passive role—is itself the moment of
gender redefinition, the point at which that person is “turned out” and becomes
newly classified as female. But in the case of gay men and trans women, there is no
need for any moment of gender redefinition, since in this hypermasculine culture,
gay men and trans women are regarded as female by definition and are thus
automatic targets for sexual assault. It is as though they have already been “turned
out” by virtue of their sexual orientation/gender identity. As James Robertson puts
it, the prison sexual code defines gay men and trans women as “fair game.”92 This
is true even for those gay men who take the dominant sexual role in their
consensual sexual relationships.93
In this cultural system, trans women, who self-identify and self-present as
women, become obvious targets.94 The same is true of gay men with “stereotypically feminine characteristics.”95 As for gay men who are placed in GP, they must
do everything possible to conceal their sexual orientation toward other men, since
to be outed is to become an immediate mark for sexual victimization. Again, gay
men and trans women are not the only victims of this “perverse subculture.”96 But,
for the reasons just explored, they are its most ready victims, which explains why

statements to the effect that “[prisoners] need to get down there and fight or get you a man,” and “[t]here’s no
reason why Black punks can’t fight and survive in general population if they don’t want to f***” Id.).
91. See, e.g., Call for Change: Protecting the Rights of LGBTQ Detainees, STOP PRISON RAPE 1, 7 (Aug. 2006),
http://www.champnetwork.org/media/callchange.pdf [hereinafter Call for Change] (stating that LGBTQ detainees continue to be “dramatically overrepresented” among the victims of sexual assault behind bars); HUMAN
RIGHTS WATCH, supra note 2, at 70 (“[I]nmates who are perceived as gay by other inmates face a very high risk of
sexual abuse.”); see also WAYNE S. WOODEN & JAY PARKER, MEN BEHIND BARS: SEXUAL EXPLOITATION IN PRISON
18 (1982) (presenting data suggesting that forty-one percent of homosexual inmates reported having been
sexually victimized behind bars); Terry A. Kupers, The Role of Misogyny and Homophobia in Prison Sexual
Abuse, 18 UCLA WOMEN’S L.J. 107, 116 (2010) (discussing a study that found that fifty-nine percent of trans
women in California prisons experienced some form of sexual assault in the previous year); SEXUAL VICTIMIZATION IN JUVENILE FACILITIES REPORTED BY YOUTH, supra note 5, at 11 (reporting that “[y]outh with a sexual
orientation other than heterosexual reported significantly higher rates of sexual victimization by another youth
(12.5%) compared to heterosexual youth (1.3%).”).
92. Robertson, A Clean Heart and an Empty Head, supra note 82, at 464; see also Man & Cronan, supra note
8, at 166 (explaining that in prison, “an inmate who is openly gay or who is a transvestite or a pre-operational
transsexual . . . is a clear target of sexual aggression”).
93. Donaldson, supra note 51, at 120.
There is no niche in the prisoner structure for a sexually reciprocal or masculine-identified gay
man such as we see in our androphilic communities. In a rural jail or minimum-security prison, he
may succeed in fending off [the pressure to take the passive sexual role], but in any other
confinement environment, the entire institution would be against him and he would have to survive
repeated combat.
Id.
94. See Peek, supra note 7, at 1229 (“[B]ecause of the nature of the prison hierarchy, an entering transgender
inmate can either choose to act submissive or be beaten into submission, and more often than not, sex is coercive
for them on some level.”).
95. Man & Cronan, supra note 8, at 166.
96. Rideau, supra note 61, at 75.

2011]

STRATEGIC SEGREGATION

19

members of these groups are “dramatically overrepresented” among the victims of
sexual assault behind bars.97
This environment is a hellish one for anyone tagged as weak and therefore open
to definition as female. For this reason, all individuals who are subject to ongoing
sexual victimization in prison must be protected. The issue is how best to do so.
The L.A. County approach is to separate gay men and trans women from all other
at-risk populations.98 In what follows, I examine L.A. County’s K6G program in
detail, describing the mechanics of the unit’s operation and the day-to-day
experience of its residents. I also report data regarding the relative safety of K6G,
and describe the gang culture that dominates the Jail’s GP unit, the absence of
which in K6G in part explains that unit’s appeal. This discussion will be confined
as much as possible to description, deferring until Part III consideration of likely
objections to the K6G approach.
II. THE ORIGINS, MECHANICS AND EFFECTS OF L.A COUNTY’S K6G UNIT
A. Origins
L.A. County is home to the biggest Jail system in the country.99 On any given
day, it houses more than 19,000 people100 in 8 facilities, with 166,000 people on
average cycling through each year.101 The biggest facility in this system is Men’s
Central, a decaying, old-style jail with an average daily count of almost 5,000
prisoners, making its population bigger than that of most prisons and even many
prison systems.102 Men’s Central is a mix of dormitories and cellblocks designed
in the linear fashion familiar from prison movies. There are also scores of
infirmary cells in the medical wing, and rows of single-man cells used for

97. Call for Change, supra note 91, at 7.
98. Unfortunately, in Los Angeles County, protection of the remaining groups of likely victims—infelicitously
labeled “softs”—is at present sorely inadequate. But there is nothing to prevent the Department from committing
to also providing meaningful protection for those falling within this designation. With K6G, the Department has
shown itself capable of creating safe spaces for the most vulnerable prisoners, and there is no reason that it could
not make good on its obligation to protect all vulnerable prisoners, whether in K6G or in some other designed unit.
For more on this issue as regards L.A. County, see infra Part III.B.
99. Unless otherwise noted, the content of this and the following sections is based on information learned or
observations made during the course of my research.
100. The average daily count for the L.A. County Jail system was 19,373 in 2007, 19,570 in 2008 and 19,008
in 2009. Interview with Sgt. Sandra Petrocelli, L.A. Sheriff’s Department, in L.A., Cal. (Jan. 20, 2010).
101. See Personal Communication, Sgt. Steve Suzuki, L.A. Sheriff’s Department, in L.A., Cal. (April 15,
2011) (on file with the author) (providing data indicating that between 2001 and 2010, the average annual
admissions rate in the L.A. County Jail was approximately 166,000, and that in 2005, the year with the decade’s
highest number of admissions, 182,471 people were admitted to the Jail).
102. See supra note 14 (explaining the differences between prisons and jails).

20

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

disciplinary segregation, for protective custody, and to house the “K10s,” those
detainees whose profiles suggest they cannot be housed safely with others.
Most of the housing, whether minimum, medium, or maximum security,103 is
designated as “general population.” Once classified as GP, detainees can be sent to
any unit in the Jail system corresponding to their security level.104 But as with all
carceral facilities, L.A. County also has units set aside for groups of prisoners in
need of specialized housing. There are designated units for veterans, for people
with serious medical needs or physical disabilities,105 and for those who are deaf,
developmentally disabled, or seriously mentally ill.106 And there is also, occupying
three dormitories in Men’s Central, a segregation unit that houses all gay men and
trans women prisoners. That L.A. County should take particular steps to protect
these last two populations may not be surprising, given its relatively high numbers
of gay and transgender residents as compared with other counties nationwide.107
But there are other factors that make the existence of such a program in the Jail
unlikely, to say the least. The L.A. County Jail system is not only enormous; it is
also a notoriously volatile and even dangerous institution in which severe overcrowding,108 coupled with racial divisions imposed and rigidly policed by the
prisoners themselves,109 frequently leads to riots or other forms of violence.110 The
management challenges posed by this set of conditions are considerable, and one
might reasonably expect to find little by way of official protection of gay male and
trans women prisoners, who are often the first to be victimized in crowded and

103. These classifications are intended to indicate the level of danger prisoners are thought to pose and thus the
level of restrictions thought necessary to be placed on their movement and living conditions during their
detention.
104. There is, of course, a prior classification based on gender. In the L.A. County Jail system, women are
presently held at the Central Regional Detention Facility (CRDF). All other facilities house men only.
105. Members of this group are housed in the medical wing.
106. Those detainees who are seriously mentally ill are housed in the Twin Towers Correctional Facility,
directly adjacent to Men’s Central Jail. Twin Towers is the newest facility in the L.A. County Jail system.
107. Although several metropolitan areas, including the San Francisco Bay Area, outrank the Los Angeles
metro area by the estimated percentage of LGBT inhabitants, only the New York City metro area has a larger
estimated gay, lesbian, and bisexual population (568,903) than that of greater Los Angeles (442,211). See GARY J.
GATES, THE WILLIAMS INSTITUTE, SAME-SEX COUPLES AND THE GAY, LESBIAN, BISEXUAL POPULATION: NEW
ESTIMATES FROM THE AMERICAN COMMUNITY SURVEY 1–2, 7 (Oct. 2006), available at http://escholarship.org/uc/
item/8h08t0zf.
108. In 2006, a federal district judge held that overcrowding in the L.A. County Jail system had led to a variety
of unconstitutionally inhuman conditions of confinement. See Rutherford v. Baca, No. CV 75-04111 DDP, 2006
WL 3065781, at *1 (C.D. Cal. Oct. 27, 2006). In 2007, that same judge found that overcrowding had led to 24,000
instances of “floor-sleeping,” whereby detainees were forced to sleep on the floor of the jail due to insufficient bed
space. See Thomas v. Baca, 514 F. Supp. 2d 1201, 1215 (C.D. Cal. 2007).
109. For more on these racial divisions and their impact on Jail conditions, see infra Part II.D.2.
110. See, e.g., Richard Winton & Sharon Bernstein, More Violence Erupts at Pitchess: Black and Latino
Inmates Clash at the North County Jail, Leaving 13 Injured, L.A. TIMES, Mar. 1, 2006, at B1. There is an argument
to be made that the high degree of control exercised over prisoners by the leadership of the various racial
groupings also ensures a level of order and stability that could not otherwise be maintained in such an
overcrowded, understaffed facility. Indeed, this may be the reason that officials work with this culture rather than
trying to disrupt it.

2011]

STRATEGIC SEGREGATION

21

chaotic facilities. Yet for more than two and a half decades, L.A. County has
operated a separate housing unit designed to segregate these groups from GP.
This unit, long known as “K11” and now called “K6G,”111 has been operating in
its current form since 1985. Prior to 1985, there was some effort to keep
homosexual prisoners segregated from the general population, with one housing
module in the Jail set aside for their exclusive use.112 However, these efforts were
partial at best. Those housed in this unit were denied access to many of the basic
privileges GPs enjoyed as a matter of course, including vocational and educational
programming, visitation, medical and mental health care,113 and access to the law
library. Those detainees wanting access to these basic entitlements were thus
forced to forgo housing in the “homosexual” module and take their chances in
GP.114 And, still more troubling, the haphazard nature of the program created many
opportunities for GPs to gain access to the unit, making its residents vulnerable to
attack from the inside.
This early program thus suffered from profound design flaws. On the one hand,
it lacked controls for ensuring that only homosexuals115 were admitted to the
unit.116 As a consequence, all a would-be predator needed to do to gain access to

111. See supra note 15 (explaining the name change).
112. This module was the 3500 unit of Men’s Central Jail. See First Amended Complaint for Declaratory and
Injunctive Relief at 3, Robertson v. Block, No. 82 1442 WPG (Px) (C.D. Cal. Apr. 12, 1982) [hereinafter
Complaint].
113. See id. at 9.
114. This was apparently what happened in the case of James Rumph, one of the named plaintiffs in the lawsuit
brought against the Jail by the ACLU of Southern California on behalf of all gay male inmates. See infra.
According to the Plaintiffs’ First Amended Complaint, Rumph, who “was housed for more than five months in the
[homosexual] module for his own protection,” was transferred to GP without “a notice or hearing of any kind.”
Complaint, supra note 112, at 8. In their answer to the Complaint, defendants maintained that Rumph was not
transferred to module 2500 (apparently a GP unit) against his will, but that 2500 was the designated housing unit
for all “inmates who are defending their criminal action in propria persona and who have [a] court order requiring
that they be given access to the privileges and materials available to inmates of module 2500.” Answer of
Defendants to First Amended Complaint at 2–3, Robertson v. Block, No. 82 1442 WPG (Px) (C.D. Cal. May 11,
1982) [hereinafter Answer]. According to the defendants, Rumph’s transfer to 2500 was the automatic result of his
“affirmatively [seeking] a court order granting him pro per status which required his housing in module 2500.” Id.
at 3. Reading between the lines, it becomes clear that the Jail made no provision for law library access for those
homosexuals defending themselves pro se apart from housing them with the other pro pers in a GP dorm.
115. I realize that using terms like “homosexual” and “gay” in the way I do risks implying both some truth of
the matter as to who is “really” gay and the possibility of meaningfully separating out those who “are” gay from
those who are not. The formulations employed thus court charges of both essentializing and oversimplifying the
inherently fluid and even mercurial character of same-sex attraction. Even as to those men who self-identify as
gay, there is a danger inherent in any effort to distinguish on the basis of sexual identity: that of equating
characteristics stereotypically associated with a given identity with the identity itself, thereby making invisible
those who, although they do self-identify, lack those characteristics conventionally associated with gay men. See
Russell Robinson, Masculinity as Prison: Race, Sexual Identity, and Incarceration, 99 CAL. L. REV. 101, 215
(forthcoming 2011) (draft on file with the author). I address these concerns in more detail in Part III.B. For now, I
adopt the constructions employed by the Jail, the ACLU lawyers who eventually brought suit challenging the
conditions I describe, and the consent decree discussed in the text.
116. As will be explained more fully, see Part II.C.1-2, that situation contrasts sharply with the present
situation, which is modeled on the procedure created by the consent decree entered into by jail officials and the

22

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

potential victims was to aver his homosexuality on entrance to the Jail.117 On the
other hand, no efforts were made to keep gay detainees separate from GP detainees
when they were outside the dorms. This meant that gay prisoners were still
vulnerable to predation during the admissions process, in the court-line holding
cells, or in transit to and from court, the infirmary, the visiting room, or elsewhere
in the facility.118
In 1982, in response to this situation, the ACLU brought a lawsuit on behalf of
all “homosexual inmates,”119 challenging these conditions.120 The result was a
July 1985 consent decree establishing the terms of the K6G program that exists
today. Among other things, the agreement provided that dorm residents would
have access to exercise,121 telephones,122 visitation and medical care,123 as well as
“regular access to library services” and “equal access to educational and vocational training programs.”124 It stipulated that residents of the segregation unit
would be separated out as soon as possible after their arrival at the Inmate
Reception Center (IRC),125 and, further, that they would be kept separate “on

ACLU of Southern California on behalf of their clients. See Stipulation and Request for Dismissal Order at 4–5,
Robertson v. Block, No. 82 1442 WPG (Px) (C.D. Cal. July 17, 1985) [hereinafter Consent Decree] (establishing a
two-step procedure for classification to the “homosexual” units, by which “[i]nmates who state that they are
homosexual” on entering the Jail “are immediately transferred to segregated housing units for homosexuals[,]” at
which point the classification staff of the unit will make a subsequent determination as to their “suitab[ility] for
such segregated housing units”).
117. Interview with Bart Lanni, Deputy Sheriff, L.A. Sheriff’s Department, in L.A., Cal. (Feb. 11, 2010).
118. See Complaint, supra note 112, at 7–8.
119. The Complaint described the “two separate sub-classes . . . represented by [the named plaintiffs to the
suit]” as “all inmates incarcerated at the Jail and housed in that module of the Jail housing inmates who have filed
statements with the Jail Classifications Unit to the effect that they are homosexual” and “all inmates who are or
will in the future be incarcerated in the Jail and who, although they have filed statements with the Jail
Classifications Unit to the effect that they are homosexual, are denied the right to be housed in [that] module or
who are transferred from said module against their will and without hearing.” Id. at 3.
120. The plaintiffs brought several constitutional claims. They alleged an equal protection violation on the
grounds that plaintiffs “receive less access to vocational training and educational facilities” and “less access to
visitors” than “inmates who have not been classified as homosexual by Jail authorities.” Id. at 10. They alleged
that plaintiffs’ limited “access to visitors, books, ideas and information” violated their “rights of expression,
communication and association” under the First, Fourth, and Fourteenth Amendments. Id. at 11. They brought a
general Eighth Amendment claim, alleging that the regime to which they were subjected constituted cruel and
unusual punishment. Id. at 10. They alleged that restrictions on plaintiffs’ “access to law books and the law
library” violated the plaintiffs’ right of “access to court and to the effective assistance of counsel . . . .” Id. at 12.
And, in perhaps the most creative move, the plaintiffs alleged a violation of their right of access to the courts on
the grounds that the failure to keep them separate and apart from GP inmates in transit to court and in the
courtroom holding cells subjected them “to a mentally exhausting and dangerous routine on court days[,]”
presumably compromising their ability to devote themselves fully to defending their cases. Id. at 12.
121. Consent Decree, supra note 116, at 2.
122. Id.
123. Id. at 4.
124. Id. at 6.
125. Id. at 5.

2011]

STRATEGIC SEGREGATION

23

transportation buses while en route to and from court”126 and in the courthouses
themselves. In addition, in one short paragraph, the consent decree sketched the
two-stage process by which classification decisions into the unit were to be
made—a process that is in place to this day. First, as Jail staff had already been
doing, “[i]nmates entering [the IRC]” were to be asked “if they are homosexual.”127 At this point, those “who state that they are homosexual are immediately transferred to segregated housing units for homosexuals.”128 Under the
previous regime, this was the end of the matter. But the consent decree created a
second step, to be carried out “by classification staff” assigned to the unit.
According to the order, those officers were to make a subsequent determination
whether those inmates who declared themselves homosexual at the IRC stage
“[were] suitable for such segregated housing units . . . .”129 In short, it was to be up
to Jail officials to determine whether those men who claimed to be gay belonged in
the unit reserved exclusively for “homosexual inmates.”
B. Institutional Structure: Managing the Segregation
If this program is to be effective, the boundaries between K6Gs and the general
population must be carefully policed. This ongoing project has two components.
The first involves policies regarding movement of residents through the facility.
The second involves the issue of classification: how to make sure that those who
belong, and only those who belong, are assigned to K6G. In this and the following
section, I describe these two components in turn. Again, as much as possible, the
focus here will be descriptive; possible concerns and objections raised by the
practices identified will be generally deferred to Part III.
The population of K6G typically hovers around 350 residents. To accommodate
this group, three dorms have been designated for the residents’ exclusive use.130
There are also several specially designated cells in the Twin Towers facility to
house K6Gs judged to have serious mental illness, as well as a separate row of
single cells in Men’s Central designated as the unit’s disciplinary wing.131 Until
1996, the trans women were housed separately from the gay men.132 But this

126. Id. at 5 (stipulating that during the drive to and from court, “[w]henever possible, homosexuals are placed
in protective cages”).
127. Id. at 4.
128. Id.
129. Id.
130. On occasion, the population will spike, as it did, for example, in the summer of 2008. To accommodate
the extra people, K6G was temporarily allocated a fourth dorm in the same cluster as the other three dorms.
Telephone Interview with Bart Lanni, Deputy Sheriff, L.A. Sheriff’s Department, in L.A., Cal., (Feb. 5, 2010).
131. As of this writing, two further Twin Towers pods have been set aside to house K6Gs who have been
cleared for (supervised) work in the Jail.
132. The trans women originally lived separately in a unit on the top floor of the old Hall of Justice Jail. In
1993, they were moved to a unit in North County Correctional Facility (NCCF) that became known as the
“witches’ castle.” During this period, the gay men were housed at Wayside, adjacent to NCCF. In 1996, the two

24

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

separation sparked vociferous complaints on the part of the trans women in
particular, and eventually the decision was made to house the two groups
together.133 Although the numbers are constantly shifting, trans women today
typically make up anywhere from 10–20% of the unit’s population.
K6Gs spend most of their time in the dorms, but they often have occasion to be
elsewhere in the Jail. Policies are therefore required to manage the risks posed by
such movement. When in the visiting room, for example, K6Gs are seated in the
first row of booths,134 directly in the sight line of the deputies.135 They are brought
to pill call as a group, one dorm at a time, and monitored by deputies as they wait in
the hallway to see the nurse.136 In the court processing line, K6Gs are kept in a
specifically designated holding cell, and en route to the courthouses they sit in the
front seat of the vans, protected where possible by wire cages.137 And, whenever
K6Gs move through the facility for any reason—to pill call, the classroom,138 the

units were moved downtown to the Twin Towers facility and merged. Telephone Interview with Bart Lanni,
Deputy Sheriff, L.A. Sheriff’s Department, in L.A., Cal., (July 27, 2010).
133. See id. This arrangement creates some problems, although it is arguably still the more humane approach.
For more on this issue, see Dolovich, supra note 45, at 16.
134. In Men’s Central, all visits are non-contact. Detainees sit on stools facing a glass wall and speak by
speakerphone to their visitors who are seated on the other side of the glass. The absence of contact visits was
challenged by the ACLU of Southern California, but the constitutionality of the practice was ultimately upheld by
the Supreme Court. See Block v. Rutherford, 468 U.S. 576 (1984).
135. The L.A. County Jail is administered by the L.A. County Sheriff’s Department (the Department). Most
custodial staff at the Jail are deputy sheriffs, who rotate between staffing the Jail and patrolling the County.
136. During the summer of 2007, the Jail was investigating the possibility of moving pill call directly into the
dorms, which would obviate the need for detainees to leave their dorms to get their meds.
137. Several of these procedures, including segregation in the court line and en route to the courthouses, were
provided for in the original court decree. That decree also provided for the segregation of “homosexual inmates”
while they are at “the court facilities for which the Sheriff is responsible and are visually checked for their well
being as often as court routine permits.” Consent Decree, supra note 116, at 7. Unfortunately, I was not able to
establish whether and to what extent this segregation and regular monitoring is actually effected in the various
courtrooms to which L.A. County detainees may be sent.
138. There is a classroom allocated for the exclusive use of K6G. It is through this classroom that Senior
Deputy Randy Bell and Deputy Bart Lanni, K6G’s classification officers, see infra note 164, run what they call the
SMART program (for Social Mentoring and Academic Rehabilitative Training), which is comprised of an
impressive array of additional programming exclusively for the K6Gs. The roster of classes includes, among
others, GED, Drug Education, Spiritual Growth, Computer Literacy (using computers bought with money
donated by the L.A. LGBT community), Job Skills Training, and New Directions. See S.M.A.R.T. Program, Sept.
23, 2008 (program description on file with the author). That K6Gs have programming of this sort may seem at best
a side note, and at worst a way to paper over the violence of incarceration with empty reforms. But when the
experience of a K6G resident—living in a dorm setting with access to a full roster of available rehabilitative
programming—is compared with that of someone in protective custody (the typical safe housing option for trans
women or gay men without the capacity to defend themselves), the value of programs available to K6Gs is hard to
deny. See supra note 11. In most cases, protective custody means extended lockdown in single cells with no
access to programming of any kind. See supra, text accompanying notes 10–12. These conditions, which mimic
the key features of life in disciplinary segregation, have been routinely criticized by advocates for trans women in
particular, who condemn the way the system has forced trans woman to choose between the arguably
soul-destroying conditions of permanent lockdown on the one hand, and, on the other hand, living in GP at
constant risk of sexual assault. K6G, which offers a clear alternative to protective custody, is still jail, and those
who (rightly) condemn the profligate use of incarceration as a response to social disorder may object to the

2011]

STRATEGIC SEGREGATION

25

infirmary, the visiting room, or the court line—they must be escorted by a deputy.
This last measure in particular may not seem noteworthy. Men’s Central is, after
all, a high-security facility. Surely all prisoners are escorted by custodial staff
whenever they are out of their housing units. But custodial escort for inmate
movement is not the norm in Men’s Central. The Jail routinely operates at well
over capacity, and is so understaffed that there are simply not enough officers to
accompany all prisoners in transit.139 It is thus not unusual to walk through the
halls and to pass lines of unescorted detainees en route from one place to
another.140 There are two exceptions to this general rule. The first is the K10s, the
facility’s highest-security inmates, who are always escorted (in shackles) when out
of their cells. The other exception is the K6Gs.141 To help deputies keep track of
who is where, and to ensure that people are where they are supposed to be,
detainees’ uniforms are color-coded142: brown for medical, white and green for
trusties, orange for K10s. GPs are in dark blue, and K6Gs wear light blue.143
C. Classification: Deciding Who Is In and Who Is Out
If a program like K6G is to succeed, an effective classification process is crucial.
Access must be restricted to members of the groups for which the unit is intended,

existence of this unit as a cynical (or merely misguided) effort to make jail tolerable rather than rejecting the
institution altogether. But however objectionable the prevailing regime of mass incarceration may be, the fact
remains that today, over 2.3 million people are living behind bars. If there are reforms that would make that
experience safer and more endurable, not to mention rehabilitative in even the smallest degree, we ought where
possible to implement them.
139. The combination of overcrowding and understaffing has prompted Jail officials to limit inmate movement as
much as possible. To this end, meals are brought directly into the housing units by inmate workers or “trusties.” “Clothing
exchange,” the trading of dirty clothes and bedding for clean, is likewise done by trusties in the dorms and cells.
140. It bears noting that in all my time in the Jail, I never felt the slightest bit of unease when encountering
unescorted detainees.
141. Officers I spoke with about K6G frequently complained about the policy of requiring escorts for unit
residents. In fact, I found a high level of resentment among many of the deputies in general toward the K6Gs, who
are regarded as receiving special treatment. This resentment seemed in part to arise from a feeling that Jail time
should be hard and that the K6Gs “have it too easy.” Requiring an escort at all times was one frequently mentioned
example of this special treatment. Another was the easier “program” K6Gs enjoy in the dorm; unlike GP dorms,
K6G dorms have large-screen televisions that normally remain on as long as someone wants to watch them. In
addition, K6Gs enjoy unrestricted access to the phones and showers, as well as to the full-time SMART program.
See supra note 138. There is also the galling fact that many K6G residents seem at times to enjoy themselves.
Some evenings—especially on “store night,” when commissary items are delivered—there can be an indisputably
party-like atmosphere in the dorms. I lost count of the number of officers who complained that K6Gs have “too
much fun” for people who are supposed to be being punished. There is much that may be said of this view. For
now, I note only that the fact that most of the unit’s residents have yet to be tried, convicted, or sentenced, and thus
are not being detained as punishment, seems to go unnoticed. For more on this issue, see Dolovich, supra note 45,
at 12–14, 16–21.
142. This system of color-coding may strike some readers as distasteful or even dangerous. See infra Part III.A
for further discussion of this issue.
143. During the period of my research, the Jail also housed a small number of juveniles, who wore red. The
practice of housing juveniles in the Jail has since been discontinued. See Personal Communication from Director
of Inmate Services Karen Dalton, L.A. Sheriff’s Department (Dec. 16, 2010).

26

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

and all others must be excluded. According to the terms of the initial consent
decree, the segregation unit was to hold all “homosexual inmates.” Over time, this
mission was expanded to include trans women. As to this latter group, classification poses few challenges, since in most cases, trans women self-identify and
self-present to some extent as female, thereby advertising both their status as trans
women and their vulnerability to abuse in GP. K6G is well-known throughout the
Jail, and if by chance a trans woman should be assigned to any other unit, it will
likely not be long before some officer notices and brings her to K6G.
The classification challenge thus lies primarily with determining which detainees are being truthful when they say that they are gay.144 Jail policy restricts
admission to K6G to “male homosexuals.”145 In practice, this directive is interpreted to include only those men who “live a homosexual lifestyle” when not
incarcerated.146 The precise meaning of this standard is a somewhat shifting target,
but at base, it reflects a binary, essentialist theory of male sexuality, that one either
is or is not gay and there is no in-between.147 As to each interview, what the
classification officers seek to determine is whether interviewees are “really gay,”
by which is meant that, when they are free, they seek out men and only men for
sexual gratification, for romance, and for emotional intimacy.148 This standard is
deliberately designed to exclude bisexuals,149 and in particular men who might be
regarded as “situational homosexuals,” those who have sex with men while
incarcerated and sex with women when not in custody.150 In what follows, I

144. See infra note 176.
145. L.A. CNTY. JAIL REGULATION 5.02/050.00 (Segregation and Classification of Male Homosexuals).
146. Telephone Interview with Bart Lanni, Deputy Sheriff, L.A. Sheriff’s Department, in L.A., Cal., (Feb. 5,
2010).
147. See supra note 115.
148. From what I have observed, however, being married to a woman does not on its own preclude
classification to K6G. This suggests that this standard allows for the possibility that a person could live a
“homosexual lifestyle” as K6G’s classification officers understand the term and still look to women for emotional
companionship (i.e., as life partners), if not for sexual gratification. These terms are of course fluid. Perhaps the
best way to describe the aim of this process is the identification of those detainees who would be perceived as gay
on the mainline. I return to this issue below in Parts II.C.2 & III.B.
149. For further discussion of this exclusion, see infra Part III.B.
150. Situational, or “emergency” homosexuality is commonly defined as sexual activity with partners of
the same sex that occurs not as part of a gay life style, but because the participants happen to find
themselves in a single-sex environment for a prolonged period. . . . At [the heart of this concept] is
the notion that the participants in same-sex sexual activity would not have done so were it not for
their unusual situation and that they therefore are not really homosexual.
Tina Gianoulis, Situational Homosexuality, GLBTQ: AN ENCYCLOPEDIA OF GAY, LESBIAN, BISEXUAL,
TRANSGENDER, & QUEER CULTURE (Mar. 3, 2004), available at http://www.glbtq.com/social-sciences/situational_
homosexuality.html; see also Christopher Hensley, Richard Tewksbury & Jeremy Wright, Exploring the
Dynamics of Masturbation and Consensual Same-Sex Activity Within a Male Maximum Security Prison, 10 J.
MEN’S STUD. 59, 61 (2001) (explaining that, on the general view of situational homosexuality, “most men
engaged in situational same-sex sexual activities would return to heterosexual sexual activities once removed
from the segregated environment”). In Part III.B, I explore the practical reasons for these exclusions and address
likely objections to them and to this standard more generally.

2011]

STRATEGIC SEGREGATION

27

explain the way this classification system operates in practice.151
1. The Initial Classification
Classification into K6G happens in two stages.152 The first stage occurs in the
Inmate Reception Center (IRC), which is the first place to which all new admits are
brought. The IRC, as one deputy memorably put it, is where “Joe Citizen” is turned
into “Joe Inmate.” It is here that, after a strip search and a shower, new admits
exchange their own clothes and personal property for prison blues and a sack
lunch. They then see a nurse for medical and mental-health triage, give a DNA
sample to the state DNA database, have their fingerprints taken or verified, and are
transferred to their housing units.153
In addition, during intake, each person must speak to a classification official.
The set-up for these conversations—a row of officers sitting behind glass and a
line of men waiting their turn—is reminiscent of a wait for a bank teller. Once at a
window, each new admit is directed to pick up a wall phone and asked a series of
questions to determine his security classification (i.e., minimum, medium, or
maximum) and to identify any “special handles,” that is, those people who require
protective custody or who belong in a specialized unit. This interview occurs at
dazzling speed, a pace necessitated by the number of people who must be classified
into the Jail. On a busy night,154 the L.A. County IRC can take in as many as 1000
people. To facilitate this process, almost all new admits to the Jail are booked by
law enforcement officers at the police station.155 Much of the key information
(name, age, current charge, prior arrests, prior incarcerations, etc.) is thus already

151. See infra Part III for possible criticisms of this process.
152. Unless otherwise noted, the content of this and the following sections is based on information learned or
observations made during the course of my research.
153. For some, this process can take a few hours. For others, most notably those who are over 55 (and who are
therefore required to have a more in-depth medical exam) or who admit to any medical or mental health problem
in their triage interview, it can take days. In the past, those awaiting completion of the admission process were
forced to sleep on the floor in IRC. More recently, a unit in Twin Towers was repurposed to provide beds for those
waiting to see the nurse or the psychiatrist in advance of their transfer to more permanent housing. Notwithstanding this improvement, the desire not to be caught in IRC for any length of time is the reason that some of those
who have been through the process before will sometimes deny having any medical or mental health needs during
the initial triage interviews, even if they in fact need medical or mental health attention. That way, they can be
transferred to their housing units and be assigned a bunk—widely thought preferable to being stuck in IRC—
before making their medical needs known.
154. Mondays and Tuesdays are the busiest evenings, since by then, those who were arrested over the weekend
and held in local lockups have seen a judge and have been officially taken into custody within the constitutionally
mandated period. See CAL. PEN. CODE § 825(a)(1–2) (requiring that arrestees be brought before a judge within 48
hours, but “excluding Sundays and holidays” and rolling over to the next day if the 48 hours expire when the court
is closed); Cnty. of Riverside v. McLaughlin, 500 U.S. 44, 56–57 (1991) (holding that jurisdictions generally must
find probable cause to detain a person arrested without a warrant within 48 hours of the arrest).
155. The L.A. County Jail serves an enormous geographical area and receives people into custody from the
Los Angeles Sheriff’s Department, the California Highway Patrol, and the Los Angeles Police Department, as
well as forty-seven other local police departments. See Municipal Police Departments in Los Angeles County, LOS
ANGELES ALMANAC, http://www.laalmanac.com/crime/cr69.htm (last visited Mar. 5, 2011).

28

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

in the Jail’s computer system.
At the window, once the booking number is provided and the relevant information appears on the classification officer’s screen, the officer then asks a series of
questions. Some of these questions go directly to the issue of security level (have
you ever escaped from a secure facility? was there a weapon involved in the
current offense? do you have three or more prior felony convictions?156). Some are
solely for information-gathering purposes or some other policy purpose (are you
currently employed? are you homeless? have you ever been ordered by a court to
pay child support?). And, finally, some questions are asked to identify special
handles and to determine placement in the Jail. In this vein, new admits are asked:
are you suicidal? did you ever serve in the military?157 and, for purposes of
identifying those who belong in K6G, are you homosexual? This is typically the
first moment the institution is alerted that a given person needs placement in K6G.
With an affirmative answer to this question, new admits are immediately directed
to sit on a bench directly in view of custodial staff. As soon as possible, they are
given light blue uniforms to signal their need for official monitoring and segregation from the general population, and transferred to the K6G holding tank.
It bears noting that the initial process of identifying and segregating those men
who report being gay has arisen in an institutional context in which, every week,
thousands of men move through a mammoth facility that is both overcrowded and
understaffed. This in itself is no small achievement: so effectively and seamlessly
has K6G been absorbed into the Jail’s operation that the identification of K6Gs is
now woven into the daily IRC routine. Indeed, K6G has become so commonplace
that even without any formal institutional mechanism or official prompting,
law-enforcement officers across the county have taken it upon themselves to keep
gay men and trans women separated from other detainees—and thus to provide
desirable protection158—even before arrival at the Jail.159
To this, one might object that jail is a dangerous enough environment for gay
men. Is it really wise to call out members of this group in IRC—or worse, at police
stations—and physically place them so that anyone paying attention would know
of their sexual orientation? Should we not aim to discourage official inquiry into

156. The answers to these questions are generally checked against information already in the computer system.
157. The Jail has a unit specially designated for veterans. For veterans in custody, their period of service to
their country is often their proudest achievement. Administrators have found that, by housing veterans together
and thereby acknowledging this service, detainees are much more likely to behave well and to maintain an orderly
and well-run unit.
158. See GILLIGAN, supra note 55, at 163–87 (describing the dangers prisoners face in lockups, courthouses,
and Jail transfers); see also United States v. Bailey, 444 U.S. 394, 421 (1980) (Blackmun, J., dissenting) (“A
youthful inmate can expect to be subjected to homosexual gang rape his first night in jail, or, it has been said, even
in the van on the way to jail.”).
159. In cases when a law enforcement officer aware of K6G determines in advance that a person in her custody
belongs in K6G, that officer will alert Jail officials of this fact on arrival, and Jail officials will take steps to
separate the individual even before he meets with a classification officer for the initial classification interview.

2011]

STRATEGIC SEGREGATION

29

the sexual orientation of arrestees in close proximity to others—whether officers or
fellow detainees—who may be more likely to react to gay men with hostility than
with tolerance, much less with affirmative concern for their well being?
These concerns are valid ones. The problem is that it is impossible to segregate
those people who are at risk of assault on the basis of their sexual identity without
identifying those detainees who are gay. Certainly, all possible steps should be
taken to increase the security of those who self-identify as gay during the prior
booking process or the IRC classification process.160 Perhaps, for example,
measures might be taken to effect the more immediate removal of self-identified
gay men from the main spaces of IRC to reduce their exposure to possible
predators.161 Yet, however risky it may be to invite official identification of gay
men, whether in police lockups or IRC itself, doing so seems an inescapable
feature of housing members of this group separately from the general population.
That the current mechanism for identifying those individuals eligible for K6G
strikes an appropriate balance is strongly implied by the readiness with which
many men being processed through IRC offer an affirmative answer when asked
by classification officers if they are homosexual. There is no way of knowing how
many gay men answer in the negative for fear that admitting their sexual
orientation may expose them to harm.162 But what came through strongly both in
my interviews with K6G residents and in other informal conversations with people
at the Jail was that those in the know—a category that appears to include many
people who have done more than one stint in L.A. County—are eager to identify
themselves as gay to ensure their transfer to K6G. Some readers may view this fact
merely as evidence of how bad it is for gay men in the general population, so bad
that they are willing to out themselves on first entering the Jail. Yet even were it the
case that only a desire to avoid harm leads men returning to the Jail to disclose
their homosexuality to police officers or Jail officials, this in itself should be reason
enough to regard the initial classification process as on the right track. Given the
routine violence that marks the American carceral experience, it would be not only
ill-advised but also morally indefensible to eschew effective if partial improvements in security on grounds of their imperfection.163

160. None of my respondents raised this issue during my interviews, although this does not mean that further
investigation would not turn up cause for concern.
161. In the case of trans women, the danger that attends such exposure is both well known and unavoidable.
Their very appearance often gives them away, which is why trans women face such tragically high rates of sexual
assault while in prison, and why it is standard practice to put trans women in some form of protective custody
immediately on detention. See supra, text accompanying notes 10–13. It is only because the sexual orientation of
gay men is typically not immediately obvious to observers, and indeed in many cases may be kept entirely hidden,
that the question even arises.
162. See infra Part III.B.
163. As with K6G more generally, if inquiry into sexuality reduces victimization, it would seem that the better
approach is not to prohibit inquiry into sexuality, but to take all possible steps to reduce the dangers that arise from
doing so.

30

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

2. The Second Step: Classification as Detective Work
Those detainees who identify themselves as gay in IRC are not automatically
housed in K6G. Instead, they are transferred to the K6G holding cell to await the
second step of the classification process. During my time in the Jail, this step
was—and as of this writing, still is—conducted by two officers who have long
been in charge of the unit: Senior Deputy Randy Bell and Deputy Bart Lanni.164
Altogether, Bell and Lanni have spent a combined 40 years running the unit. In this
position, they wear many hats,165 but their primary responsibility is classification.
Every morning, anywhere from ten to twenty-five people await them in the K6G
holding cell, having been sent there from IRC the previous night. The task of these
two officers is to determine which of these detainees to classify to K6G and which
to send to GP. In many cases, it will be obvious that K6G is the appropriate
classification. Some of those waiting in the holding cell will be trans women, who
generally can be classified into the unit at a glance. Others will be well known to
Bell and Lanni from previous stays in the Jail. L.A. County has an overall annual
recidivism rate (defined as anyone who is readmitted to the Jail within three years
of release) of somewhere around 65 percent. But Bell and Lanni report that in
K6G, the number is even higher, with a lifetime recidivism rate of 90–95
percent.166 As a consequence, many of those who come through the K6G
164. I offered these officers the option of my using pseudonyms instead of their real names, but they declined.
This choice made sense, since judging from Google, Bell and Lanni are already widely known for their work in
K6G.
165. These two officers run the SMART program, an educational program of their own devising exclusively
for K6Gs. See supra note 138. They manage the many providers who serve the K6G population, including the
L.A. Department of Health, whose staff, permanently assigned to K6G, provide testing for sexually transmitted
infections including syphilis, hepatitis C, and HIV; caseworkers from Tarzana Treatment Center, which provides
reentry services and drug treatment for HIV-positive detainees on their release; an on-site psychiatric technician;
teachers from the Hacienda LaPuente School District, who have in the past had the contract to provide GED
classes in the Jail; and staff from the Center for Health Justice, who provide HIV counseling and preventive
education to K6Gs and who distribute condoms in the dorms once a week. Bell and Lanni also maintain security
in the classroom, holding cell, and hallway outside their office; monitor goings-on in the dorms (although they are
somewhat hampered in doing so by the physical separation between their office and the dorms, which are on
different floors); mediate disputes between dorm residents; and provide general counseling and assistance for
those dorm residents who seek them out. The commitment of these officers to K6G has not gone unnoticed by the
L.A. LGBT community. In 2005, Bell and Lanni received a special commendation from Christopher Street West
Hollywood and rode together under the CSW banner in the 2005 L.A. County Gay Pride Parade. So well
respected are they in the LGBT community that they joke that when they retire they should open a bar on Santa
Monica Boulevard called K11 (K6G’s original and long-standing designation), because it is sure to be a success.
166. Interview with Senior Deputy Randy Bell, and Deputy Bart Lanni, L.A. Sheriff’s Department, in L.A.,
Cal. (July 2007). How to explain this high recidivism rate? To some extent, it is a testament to the number of
residents in the unit without a stable and supportive home to which they may return on release. Many men in GP
have women waiting for them on the outside, whether mothers or wives or girlfriends. These women look out for
them while they are incarcerated—visiting them, writing to them, putting money in their commissary accounts,
accepting their collect calls—and will also provide a home to which they can return once they are released. K6Gs
generally do not have girlfriends and wives—although some do—and sadly, many K6Gs, especially the trans
women, report having been rejected by their families. They thus lack the external support that might help them
stay out of trouble once they are released. An even greater factor in the K6G recidivism rate may be the high

2011]

STRATEGIC SEGREGATION

31

classification office have been there many times before.167 Having already been
classified into K6G in a prior interview, members of this group are immediately
sent down to the dorms with the trans women.
The most challenging part of this second stage involves dealing with those who
remain. These are people Bell and Lanni have never met before. Their task is to
determine which of these men belong in K6G and which should be “declassed”168
and sent to GP or some other unit. Plainly put, Bell and Lanni must determine in
each case whether the person is in fact gay in the dualistic, essentializing way they
deploy the term—or, as they put it, whether he “lives a homosexual lifestyle” when
not behind bars.169
There will inevitably be some underinclusiveness in the K6G classification
process. Some gay men may not know about K6G and may well assume that
identifying themselves as gay to Jail officials would put them at greater risk. Some
will no doubt prefer taking their chances on the mainline to “coming out” to state
officials.170 And some may deliberately try to avoid K6G altogether, going to GP to
get access to “trade” (i.e., masculine men attracted to effeminate men). If the
people who choose to hide their sexual orientation or gender identity in order to go
mainline are able to hold their own physically, this misclassification may not pose
a problem. The worry is that some who make this choice may underestimate the
extent of their own vulnerability and wind up badly hurt. This in itself is cause for
serious concern,171 although it bears noting that under these circumstances, the Jail
would probably avoid legal liability for the harm. In order for prisoners to recover
for an Eighth Amendment claim, they have to show that prison officials actually

percentage of K6Gs with drug addiction. In my interviews, I asked respondents whether they had a problem with
drugs prior to their incarceration, and the responses made me think that if not for crystal meth, many of my
subjects would not have been in Jail. Of the thirty-three interviewees who answered the question, twenty-nine
answered in the affirmative (i.e., that they did have a drug problem prior to their incarceration). Of those
twenty-nine, nineteen spontaneously identified their drug of choice. For fourteen of those nineteen, the drug of
choice was crystal meth, a highly addictive narcotic. Of the remaining five, three named marijuana and two named
crack cocaine. These numbers lead me to suspect that many of the ten who did not name their drug were also
likely meth users.
167. See Nina T. Harawa et al., Sex and Condom Use in a Large Unit for Men Who Have Sex with Men (MSM)
and Male-to-Female Transgenders, 21 J. HEALTH FOR THE POOR AND UNDERSERVED 1071, 1073 (2010) (noting that
“[t]he average length of stay” in Men’s Central “is 42–45 days, but varies widely”). I personally can attest to the
revolving door effect in K6G. Although I spent the bulk of my time in K6G in 2007, over the past several years I
have visited regularly. And every time I go, there are many faces I recognize, as well as several people I know by
name whom I met on previous visits. Given the relatively short stays of many K6Gs, it is likely that many of these
people had come and gone from the Jail multiple times since I first saw or met them.
168. Those who identify themselves in IRC are preliminarily classified to K6G. Their subsequent removal to a
different unit thus requires that they be declassified or “declassed” from K6G and reclassified into some other unit,
whether GP or otherwise.
169. See supra notes 145–50 and accompanying text.
170. See Robinson, supra note 115, at 191–92 (explaining the “profound” consequences of coming out,
especially for men of color).
171. For further discussion of this issue, see infra Part III.B.

32

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

knew of a substantial risk of serious harm and failed to act to avert it.172 Where,
however, the victim has failed to disclose his sexual orientation despite the
institution’s efforts to discover it so as to place him in safer housing, the victim will
be hard-pressed to recover, for two reasons. First, having withheld from the Jail the
crucial fact of his sexual orientation, the victim will be unable to show that Jail
officials actually knew of his vulnerability to assault. Second, courts are unlikely to
be sympathetic to one who, by his own choosing, thwarted the very system
designed to protect him from the harm ultimately suffered.
Issues of liability, however, are far more salient for Jail officials in the opposite
case: when a detainee, despite asserting his homosexuality, is denied access to
K6G and ultimately raped in GP.173 This being so, it might be asked, why not let
any detainee into K6G who avers that he is gay? In Part III below,174 I respond
more fully to this question. For now, it is important to recognize the daily pressure
Bell and Lanni face to get the classification right on its existing terms. On the one
hand, given the vulnerability of K6Gs, it is paramount that sexual predators be
excluded from the unit. Certainly, ensuring that admission is restricted to those
who actually satisfy the admission criteria will not guarantee protection; heterosexuals and bisexuals have no monopoly on sexual predation. But making efforts
to expose and exclude those who seek access to K6G on false pretenses seems a
prudent move. On the other hand, erroneously declassifying a gay man from K6G
could not only expose him to sexual assault in GP, but could also open up the Jail to
serious legal liability for the error.175

172. See Farmer v. Brennan, 511 U.S. 825, 837 (1994). See also infra note 175.
173. The need to avoid this outcome—as well as to protect genuinely vulnerable inmates from harm—is very
present to the minds of Bell and Lanni when they go about their work.
174. See infra Part III.B.
175. The risk of legal exposure arising from classification errors in the K6G context exposes a deeply
counterproductive feature of prevailing Eighth Amendment doctrine. Under existing Eighth Amendment law,
prison officials are only liable for failing to protect those in their custody from rape and other forms of violence if
they actually knew of the risk of harm to the victim, see Farmer, 511 U.S. at 836–37, even if those risks were ones
about which reasonably attentive prison officials would have known, and thus about which defendants should
have known. See Dolovich, supra note 36, at 889–90, 945 (describing the implications of Farmer’s deliberate
indifference standard). Prevailing doctrine thus creates an incentive for prison officials not to take affirmative
steps to identify risks to people in their custody. If affirmatively identifying existing risks might expose them to
liability, why should they try to do so? From a liability perspective, it is better for defendants not to know of
existing risks and thus not to be held accountable for any resulting harm. Understanding the incentive structure
current doctrine creates helps to explain a feature of corrections practice that might otherwise seem puzzling: why,
given what is well known about the heightened vulnerability of gay men in prison to ongoing rape and other forms
of abuse, prisons and jails across the country have not made it standard practice to identify gay men in their
custody in order to segregate them. The California prison system is typical in this regard: at no point do California
prison officials seek to identify which persons in their custody are gay, leaving this information to be ferreted out
by predators in the general population, who will use it to guide their selection of victims. Although this passive
strategy arguably puts gay men at risk, it is nonetheless incentivized under prevailing constitutional doctrine,
which rewards official ignorance of risks of harm to vulnerable prisoners. See id. at 945–47. L.A. County,
however, explicitly agreed to a policy of segregating “homosexual inmates” in 1985, nine years before Farmer
was decided. See Consent Decree, supra note 116. Although Jail officials might try to deny it, the establishment

2011]

STRATEGIC SEGREGATION

33

The challenge Bell and Lanni confront every day is to determine which of the
people in the K6G holding cell, all of whom claim to be gay, are telling the truth.176
Why would anyone lie about being gay in the Jail? Although preventing predators
from getting access to the unit is an ongoing concern, many of those seeking
admission under false pretenses only want a respite from the gang politics and
consequent pressure and volatility that define daily life in GP.177 K6G, well-known
to be the safest place in the Jail, is an attractive prospect for repeat players who just
want to do their stints in peace. Because Bell and Lanni take seriously their
assigned responsibility to restrict access to K6G to authorized groups, much effort
is put into ferreting out those who are falsely claiming to be gay.
In the years since becoming familiar with the workings of the K6G classification
office, I have encountered a number of people who, knowing a little about how
these determinations are made,178 have formed a fundamentally mistaken impression of this process. I have been told, for example, that the officers require
maternal confirmation that a detainee is gay, that they administer a “test” of gay
terms and admit only those who correctly define a minimum number, and that only
those who are familiar with the gay bars and clubs on the Santa Monica strip will
be admitted. To be sure, there are shades of truth in these accounts. The officers

and ongoing implementation of the K6G program represents official acknowledgment of the heightened risk of
harm that gay men and trans women face in the Jail. This does not automatically mean that any gay men and trans
women who are raped in the facility will succeed on a constitutional claim. To the contrary, proof that Jail officials
followed established practice in attempting to house the victim in K6G, and to prevent non-K6Gs from gaining
access to the unit, will provide strong evidence that prison officials, having taken affirmative steps to provide
protection, were unaware of a serious risk to the plaintiff, thus satisfying Farmer.
Still, the existence of K6G, and the official awareness it implies of the dangers of housing gay men and trans
women in GP, does create the potential for legal exposure anytime jail officials fail to take sufficient care in
complying with their own policies. And given the substance of those policies—i.e., that K6G is open only to gay
men and trans women—it is both reasonable and appropriate that institutional efforts would be directed to making
sure that no one else gets into the unit. For critical analysis of Farmer’s recklessness standard and an argument
that, at a minimum, prison officials should be held to a standard of gross negligence, see Dolovich, supra note 36,
at 940–71.
176. At this point, some may object that sexuality is more dynamic and complex than the binary gay/not gay
would allow, and that even men who may not “seem” gay in the conventional sense of the term may experience
same-sex attraction and thus not identify as “straight.” This is no doubt the case. But my assertion in the text that
some men lie to get access to K6G is not based on a failure to credit either the complexity of sexual identity or the
range of ways people might understand and relate to their own sexuality. It is based on the frank admissions of
many men whose classification interviews I observed that their claims of being gay, made in their initial sorting
interview, had in fact been lies. Although one’s stated self-understanding can certainly be complicated by fear of
the implications of connecting with those parts of oneself that are in conflict with prevailing social norms, it would
be a mistake to allow theoretical sophistication to blind us to the possibility that, in many cases, the most accurate
explanation is also the most obvious. Sometimes, in other words, a lie is just a lie. And my experience in K6G
leaves me confident that the phenomenon I describe in the text—of men who seek access to K6G by pretending to
be gay—is a frequent occurrence.
177. For more on the pressures that gang politics create in the Jail’s GP units, see infra Part II.D.2.
178. Over its history, the unit now known as K6G has received a fair amount of media attention as well as
considerable interest from the community. In this way, many people who would not otherwise know about the
Jail’s classification procedures have come to learn about Bell and Lanni’s work in K6G.

34

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

have been known to call the mothers of detainees to ask if their sons are gay. They
do have a list of terms that gay men might be likely to recognize, which aspirants to
K6G may be invited to define. And they often ask interviewees if they go to any
gay nightspots, and, if so, which ones. But to imagine that the decision-making
process is as simple as any of these accounts suggest is to fundamentally
misunderstand what these officers do.179
At its core, K6G classification is detective work, and Bell and Lanni are
detectives. The inquiry in each case is the same—does this person live “a
homosexual lifestyle” on the outside?180—and the officers’ aim in each case is to
determine the fact of the matter as best as they can. Inevitably, in some cases, the
evidence will be, if not exactly in equipoise, at least insufficiently clear to allow for
certainty. In such cases, the officers will make an all-things-considered judgment,
and will usually err on the side of admission. Viewed from a legal perspective, it is
interesting that at these moments, Lanni in particular will conjure an imagined jury
before which he might be called upon in a worst-case scenario (the rape in GP of
someone he declassed) to justify his decision. If in his view, he would be unable to
do so convincingly based on the evidence before him, he will keep the person in
question in K6G.
The two officers deploy a range of tools to get at the facts of each case, but the
main instrument is conversation; they get people talking and see what emerges. To
this end, Bell and Lanni will administer forms requesting basic information (name,
age, emergency contact information, etc.) and personal history (marital status,
number of kids, name of current lover, date of last sexual encounter with a woman,
etc.). The answers supplied provide a starting point for conversation. If the person
is from L.A. County, the officers may ask whether he has been to any gay bars in
the city and, if so, to describe them, or they may use that infamous list of “gay”
terms, all in the service of starting a conversation that may help them develop a
sense of the person they are interviewing. There is never one answer that is
dispositive. Instead, in the course of an interview, a picture develops of someone

179. Just such a misunderstanding appears to inform a forthcoming article on K6G by Russell Robinson. See
Robinson, supra note 115, at 101, 120 (asserting that K6G’s classification officers “determine gay identity largely
by testing inmates on perceived norms of gay culture” and that “those who flunk the test” are “[sent] back to the
general population”). Robinson’s mistaken impression of the K6G classification process appears to be a function
of his methodology. Judging from his description of his research methods, Robinson did not observe any
classification interviews firsthand. See id. at 113. Instead, his information about the process appears to have come
from interviews he conducted with, among others, “former inmates and others who have interacted with the K6G
unit.” Id. at 113. And from the perspective of those who have been interviewed for admission to K6G, the process
would seem like a “test” one could “flunk,” since Bell and Lanni do often ask interviewees questions with right
and wrong answers. As I explain in the text, however, this feature of the process is misleading, since the officers
are far less interested in whether interview subjects give “correct” answers than with what the resulting
conversation reveals about their sexual orientation.
180. Again, for classification purposes, this term generally means that a (male) person seeks out men and only
men for emotional companionship and sexual gratification. See supra notes 145–50 and accompanying text.

2011]

STRATEGIC SEGREGATION

35

who prefers men, who when free seeks out men and only men for sexual
gratification, or of someone who is just pretending that he does so.
To this, it may be objected that there is no one way to be gay. Some gay men
may be married to women. Others may be celibate or have children or may never
have had sex with another man. But it is clear from the many interviews I have
observed that Bell and Lanni are well aware of this diversity of experience among
homosexual men, which explains why one will find in K6G some men (especially
among the older residents) who are married, and many men who have children.181
In one sense, what these officers seek is the same thing predators in GP would look
for: whether this person is sexually attracted to other men, not as a proxy for
women, but because of a fundamental sexual orientation towards men. In most
cases, someone who fits this description will have had sexual relationships or
sexual encounters with other men and is likely to have had some other life
experiences that confirm that he is gay—coming out to family members,182
associating with other gay men whether in bars or clubs or at some other venue,
identifying with the gay community whether in concrete ways or merely in the
abstract, etc. For this reason, when facing someone who seems clearly to fit the
K6G standard, Bell and Lanni will often have many facts to which they can point
to confirm their conclusion.
One interview I observed during my time in K6G183 took precisely this form.184
Several physical cues—manner of speech, gestures, etc.—suggested at once that
the person in question was a gay man.185 When asked, he explained that he had
never in his life had sex with a woman186 and that he had known his whole life that
he was gay. When asked for the name of his last boyfriend and that boyfriend’s

181. I cannot say whether there are in fact people classified to K6G who were celibate or had never had sex
with men prior to their admission. But my sense of Bell and Lanni’s understanding of same-sex attraction is that
these factors would not preclude admission to K6G.
182. It is in these cases when the officers might resort to calling a family member. It might, for example,
happen that Bell or Lanni will ask an interviewee if his family knows that he is gay. An interviewee might say that
although he has never told his father, his mother knows. In that case, Bell or Lanni might call the mother for
confirmation. That confirmation provides powerful evidence that this person is telling the truth, and it is this
evidence that these officers seek in every interview.
183. In all, I observed approximately 50 interviews, and, of those, took notes on approximately 34. For a
description of this portion of the research design, see the Methodological Appendix.
184. Field Notes, July 6, 2007, at 6–7.
185. Although taken alone, these cues are never dispositive, the sixth sense colloquially referred to by some as
“gaydar” inevitably comes into play in the K6G classification process. Given the mission of the unit, taking such
cues into account is wholly appropriate, since what comes across to Bell and Lanni in the K6G classification office
may also be expected to come across to potential predators in GP.
186. The information revealed during a K6G classification interview is necessarily of a highly personal nature.
To ensure the confidentiality of the people whose classification interviews I observed and to make the interview
subjects (all of whom consented in advance to my presence), see infra Methodological Appendix, more
comfortable, I did not record any personal details that could allow for after-the-fact identification (i.e., name, date
of birth, booking number, etc.) and only recorded general information that seemed relevant to understanding what
transpired in the interviews. As a consequence, I do not know the exact age of the person whose interview I am
describing in the text. Since Men’s Central holds adult men, he was necessarily over eighteen.

36

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

birthday, he readily supplied this information—and then did the same with the
boyfriend before that. Also in response to questioning, he listed off several gay
clubs that he frequented, and volunteered unasked a description of what he and his
friends liked to do at their favorite clubs (have a drink on the patio before going in
to dance). By the close of this conversation, there was no doubt either in Lanni’s
mind or in my own that this person was gay and belonged in K6G.
Some might object that anyone could produce these same answers, and perhaps
even mimic gay mannerisms to confirm the picture. To be sure, such subterfuge is
possible. And were such a pretender extremely disciplined and a very good actor,
he might well succeed in gaining entry to K6G. But in such a case, the right
decision would be to classify that person to K6G, not because he in fact fit the
entrance criteria, but because the job of the classification officers is not to divine an
essential truth—even assuming there is an inherent truth of the matter as to one’s
sexual orientation. It is instead to make their best assessment, all things considered, as to the appropriate classification for each person they interview. When the
data that emerge from an interview strongly indicate a gay interviewee, the right
decision is to keep that person in K6G. Short of a truth serum, the system in place
for determining whether a person who appears to satisfy K6G’s admission
standard has only been pretending to do so is as effective as can reasonably be
wished.
Moreover, as is well known in legal circles, judgments of credibility rest on
more than the substance of answers given. Body language, tone of voice, hesitation
or lack of it: all these cues go into a determination of whether someone is telling
the truth, and they assist the K6G classification officers in making their assessments. An example: at the end of the interview just described, Lanni asked the
interviewee what he thought the biggest problem was for homosexuals in the
United States. In answer, he said that it was the way gay people are stereotyped, in
that most people “don’t realize that we are professionals just like everybody else
and not everybody takes drugs and not all of us party.” Later, Lanni asked me if I
had noticed the way the man had talked about gays in general, that he had not said
they, but we—a locution indicating his identification with the group. This is the
sort of subtle clue the officers look for in trying to formulate their judgments.187
Some readers may yet be troubled by the possibility that Bell and Lanni might
regularly refuse admittance to K6G to a substantial number of men on the grounds
that they disbelieve their claims of being gay. Is not sexual orientation more
complex than this true/false binary would allow? Who are these law enforcement
officers—both of whom have wives and children, neither of whom identifies as
gay—to tell a man who insists that he is gay that he is not? There is no question
that sexuality is subtle and complex and that sexual identity may be more elastic,

187. I recorded all the details of the interviews described in this section in July 2007. I drew on those field
notes to reconstitute what took place at that time.

2011]

STRATEGIC SEGREGATION

37

dynamic, and hard to classify than the reductionism of conventional labels like
“gay,” “bi,” and “straight” would allow. But having sat in on countless K6G
classification interviews, I think it safe to say that a good many of the people who
tell the IRC classification officer that they are gay are quite simply lying.188
Indeed, after an unsuccessful effort to lie their way into K6G, aspirants to the unit
will often admit to the lie and even explain their motivations.
Sometimes the lies are close to the surface and fairly readily detected. It is well
known among K6Gs that Lanni often asks those from L.A. County whether they
ever go to gay clubs or bars, and if so, to name them.189 Very often, even without
firsthand knowledge, an interviewee will have an answer to this first question at the
ready, since in the K6G holding tank, aspirants are frequently prepped or
“schooled” by the regulars to expect this question. But in many cases I observed,
the prep work went no further, so that the interview subject was unable to provide
the correct answers to the inevitable follow-up questions, which may be about the
location of the various clubs, their decor, or the number of floors.190 In this event,
the interview subject will be encouraged to talk even when his answers are plainly
wrong, so that later, he may be confronted with the many falsities he had offered, at
which point the person will often admit that he had been lying all along. Such a
revelation will not always end the matter, since as the officers well know, men may
be gay without being club-goers. But from what I have observed, having been
exposed in a lie, whether about a club or something else, people who are only
pretending to be gay will often lose heart as to the whole enterprise and admit to
the larger lie.
Sometimes the officers will open the interview by asking about a disclosure
made on the intake form, perhaps the fact that the subject is married or engaged to
a woman or, as in one case I observed, had four children by four different women.
Again, these facts do not necessarily mean the subject does not meet K6G’s
admission criteria. But they provide a point of departure for further questions. For
example, the officers might ask: how long have you been gay? Or: I notice you’ve
been in the L.A. County Jail five (or ten or fifteen) times before now. Why have you
never before told the intake officer that you are homosexual?191 In each case, the
goal is to get the subject talking and see if his story seems authentic, makes sense,

188. It should give some indication of how desperately scary, dangerous, and unpleasant life is in GP that men
who are not sexually attracted to men qua men, but who are instead sexually attracted to women—especially men
about to enter an institution in which gay men are known to be at serious risk of victimization—routinely claim to
be gay on admission to the Jail.
189. Both Bell and Lanni participate in the process. On the days I spent in the office, Lanni most often took the
lead on interviewing, with Bell playing the role of backup and provocateur.
190. Bell and Lanni take occasional field trips to the Santa Monica strip and elsewhere in Los Angeles to stay
up on the bar scene. They frequently take photographs of various hotspots, then present these photos for
identification to interviewees claiming to be familiar with the places they picture.
191. For discussion of the intake process, see supra Part II.C.1.

38

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

hangs together.192 In one case I observed, the starting point for conversation was
the charge. The interviewee was being held on charges of assaulting a boy in the
living room of the boy’s house. The subject was asked what he was doing in the
house, and he said that he was there to see the boy’s mother. It eventually came out
that this woman was his girlfriend, the suspicion of which had likely led Lanni to
ask the question in the first place.193
After sitting in on enough interviews, I myself learned to react with skepticism
to some responses. For example, a number of subjects, when asked about prior
marriages or why in previous visits to the Jail they had never said they were gay,
explained that they had only just become gay the year before. Others claimed that,
after years with women, they had recently decided they wanted to try men. These
answers contrasted sharply with those from obviously gay men who, when asked
how long they have been gay or have known of their homosexuality, typically
offered some variant of “all my life” or “as long as I can remember.”194 Certainly,

192. Sometimes the questions may be sexually explicit: do you pitch or catch? What does semen taste like? As
to the latter question, Lanni reports that pretenders will often wax lyrical about the delicious taste of semen, while
men with sexual experience with other men will offer more prosaic observations—that it can be salty, for
example, or that it depends on what the person ate that day.
193. Another case of this sort took some time to unravel. The subject had a number of tattoos featuring female
names. Lanni pointed to one and asked whose name it was. The answer was that it was the name of his daughter
who had died. Then Lanni looked at his intake form and saw that he had recently had sex with a woman. When
asked about this, the subject said she was his “ex-best-friend.” To the question of where the woman was now, the
subject answered that she had died. How had she died? After hesitating, the subject said, in an unconvincing tone,
“cancer.” A few moments later, when asked again about his daughter—whose name was supposedly tattooed on
his arm—the subject said “oh, it was only my stepdaughter.” Now consider Lanni’s position. He has to decide
whether the subject’s claim to be gay is credible. If there is nothing in this exchange to offer definitive proof of a
lie, there is nonetheless much that was suspicious. For one thing, given the recentness of the sexual encounter,
how plausible was it that his partner had already died of cancer? What does it mean to be an “ex-best friend”? And
how common is it for men to label a woman this way, especially a woman with whom he recently had sex, who
also just reportedly even more recently died of cancer? As for that tattoo, did it refer to his daughter or his
stepdaughter? Or, as was more likely the case, did it refer to a woman with whom he was emotionally involved, a
fact he sought to hide from Lanni by claiming the name to be that of his daughter? It is not surprising that a man
would tattoo his daughter’s name on his own arm, especially if she had died. But if one cared enough about that
person to have her name tattooed on his arm, would that same person, when asked about her apparent untimely
death, later wave away any questions about her death with a careless “oh, it’s only my stepdaughter”? In short,
there was much in this story that did not add up.
A more likely story was that this man had female lovers whose names he had tattooed on his body, that the
recent sexual encounter with a woman, whoever she was, was a standard part of a life lived as a heterosexual, and
that he was trying as best as he could to deflect questions that might reveal these facts. A further clue bears noting:
his reference to his recent lover as an “ex-best friend.” Even in my short time observing K6G classification
interviews, I noticed a particular pattern whereby interviewees would invoke their best friends either to supply the
name of a man who is supposedly their lover or to explain away the inconvenient fact of a recent female lover or
another woman in their lives. In another interview, for example, Lanni asked the subject if he had a boyfriend, and
if so, what his name was. The subject said yes and immediately supplied the name. Then Lanni asked where they
met, and the answer was that he was his best friend and that they had grown up together. It could have been that
this longtime close friend was in fact his lover, but it was more likely that his was the first male name the subject
could think of under pressure—and indeed, subsequent questioning confirmed this latter explanation.
194. One such case occurred on July 9, 2007. That day, there were five people who required classification
interviews. In my field notes I described the first case as “an easy one.” As I explained, this person “put on his

2011]

STRATEGIC SEGREGATION

39

many men may take years to admit to themselves that they are gay. But in a highly
homophobic culture, it is unusual to say the least that someone who had been
unproblematically attracted to women his whole life would suddenly decide to
“become gay” out of curiosity.
Again, none of these data points alone will determine the ultimate classification
decision. And, as I witnessed in several cases, when the evidence appears
genuinely conflicting, the person in question will typically be held in K6G. What I
am trying to convey with these examples is that classification to K6G is not
arbitrary, nor are the classification decisions based on an official refusal to
recognize, or an inability to understand, that all kinds of men—married, single,
celibate, black, white, Latino, professionals, manual laborers, gang members,
those with children and those without, those who have never set foot in a gay bar or
attended Gay Pride—can be gay. From what I have observed, the process currently
in place recognizes this diversity full well, and the officers in charge intelligently
and methodically use a variety of methods in their attempts to identify, consistent
with their official responsibility, which of the thousands of people they see every
year are telling the truth at intake about their sexual orientation as they themselves
understand it, and which are just pretending.
There are many reasons why one might object to the process I have just
described. It may seem to validate and even reify one particular and arguably
partial and even stereotypical notion of gay male identity. Even worse, it seems to
condition access to safety on the ability to demonstrate the markers of this
particular notion, as well as on one’s willingness to allow state officials access to
the most private and intimate details of one’s life. And, for all this, it yields a
protected class that is both overinclusive and underinclusive: overinclusive because it will invariably lead to classifying men to K6G who are physically able to
live safely on the mainline, and underinclusive because it refuses admission to
those who, although neither gay nor trans, nonetheless exhibit several other
characteristics known to make men vulnerable to sexual assault in custody. Indeed,
to the extent that the classification process seeks to identify those people who are
primarily attracted to other men, it may even be underinclusive as to its own goal
of protecting those prisoners at risk of victimization because of their sexuality,
since the qualities that would lead a person to be identified as gay within the prison
culture may on further examination prove to differ from the standard currently
employed in K6G.

form that he had never had sex with a woman, [and] that he knew his whole life that he was gay.” When asked
“about his former lovers and their birthdays,” he recited “names and birthdays without any hesitation.” At some
point during the interview, Lanni asked him where he met his new boyfriend and he said that “he’s a computer guy
and [this] interviewee had had a broken computer and called and this guy came to fix his computer.” The entire
exchange felt authentic and the decision to keep this individual in K6G was, as I noted at the time, “pretty
straightforward.” Field Notes, July 9, 2007, at 1.

40

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

These concerns are serious ones, and I examine them more fully below.195 For
now, my aim is simply to explain the K6G classification process. And to
understand this process, it is necessary to recognize that, every day, people
admitted to the Jail who are aware of K6G’s admission criteria and know
themselves not to satisfy it nonetheless claim to do so in a bid to get into the unit.
As best I could judge, those seeking access to K6G under false pretenses do so
for two main reasons, which are not mutually exclusive. First, those with a long
history of institutionalization, who have become accustomed to having sex with
men while incarcerated, know that in K6G, they will find many willing sexual
partners. In fact, it appears that many who fall into this category fully expect to be
sent back to GP.196 But as those who have been through this process well know,
Bell and Lanni do not start work until 5:30 a.m., and someone admitted to the Jail
the previous evening can thus expect to enjoy a long night in the K6G holding cell
with any number of people open to a sexual encounter.
More disturbing was the other motivation that often emerged after the lies were
exposed: the desire for safe housing in the Jail. In some cases, interviewees had
specific reasons to fear placement in a less well-protected unit. In one such
instance, after a lengthy exchange in which the interviewee made many inconsistent and self-contradictory claims regarding his supposed gay identity, it emerged
that this man was a witness in a murder trial that was in process at the time—and
that he had been arrested for an unrelated offense and now found himself detained
in the same facility that was also holding the two men against whom he was
testifying. When asked why he had not simply requested protective custody in the
K10 unit instead of pretending to be gay to get into K6G, he explained that the two
defendants in the trial had long arms and that, as he put it, “even in K10, people
always seem to run into those guys.”197 He thus figured K6G was his safest bet.
That this man felt compelled by what awaited him elsewhere in the Jail to seek
refuge in K6G gives some idea of the unimaginable dangers and deeply diminished
options that people behind bars in the United States routinely face. One interview
in particular provided a disturbing window on this world. In that case, the
interviewee appeared from the start to have serious mental health problems. Lanni
began the conversation with general questions about the man’s recent history,
which I eventually realized was a way to try to calm him down. In response, this
man gave a long and detailed account of going from shelter to shelter, trying to get
his psych meds, leaving a residential treatment program without permission, being

195. See infra Part III.A–B.
196. This is not, however, true of everyone who falls into this category. At least one person I observed, after his
efforts to sustain a pretense of being gay were exposed, admitted that he heard that the K6Gs “have a pretty good
time,” and that he wanted to have a good time too. I took this phrasing to be a reference to the consensual sex that
regularly occurs in the K6G dorm, and in which this person had hoped to participate.
197. Field Notes, July 4, 2007.

2011]

STRATEGIC SEGREGATION

41

refused entry to another program because he had left the first, and so on. Then
Lanni asked a series of questions about his sexuality, the answers to which strongly
pointed to this person having several gay friends but not to his being gay himself.
Eventually, Lanni asked him why, in all his many stints in the Jail, he had never
before said he was gay. What emerged were several reasons for his trying to get
into K6G that had nothing to do with his sexuality.198
Consistent with his apparent mental health profile, this man’s previous turn in
the Jail had been spent in Twin Towers, which houses those detainees with serious
mental illnesses, and he plainly did not want to go back there. As he described it,
Twin Towers was “dirty, nasty, crowded,” and people there “disrespected him and
threatened to kick his ass all the time.” On further questioning, he explained that
there are no gang politics in the psych units, that “it is pretty peaceful,” but that
“it’s just really overcrowded,” and the mats on the bunks are “too thin and not as
comfortable as the ones in Men’s Central.” But he also did not want to go to GP,
and with good reason. This man, it turned out, had done time in the Texas prison
system, where he had been a member of the Texas Aryan Nation and had the
tattoos to prove it. He explained that there is “gang warfare in prisons across the
country,” and that “California doesn’t like Texas.” He was afraid that if anyone
(meaning any prisoner) in California saw his Texas tattoos, he would be stabbed.199
And still worse, he feared that the strongly racialized gang politics of the
California system200 would force him to affiliate with one of the white power
gangs. The problem, he explained, was that with prison gangs, it is “blood in and
blood out,” meaning that anyone seeking membership must seriously hurt someone on the gang’s behalf before gaining admission (“blood in”), and that if he
changed his mind and tried to drop out, he would become vulnerable to attack by
the members of the gang he sought to leave (“blood out”). Even apart from the risk
of physical harm, being forced to undertake such violence on the inside could
expose him to serious criminal charges. To make matters worse, he already had
two strikes, which meant that any subsequent felony conviction in California could
earn him a 25-year mandatory minimum prison term.201 And as he explained, he
didn’t “really want to get a 25-year-to-life bid on a third strike just for dealing with
someone else’s problem and for someone else’s politics.”
In sum, here was a man who did not want to go back to Twin Towers where the
conditions were especially uncomfortable and possibly threatening, and who did
not want to go to GP out of fear of: (1) being stabbed for being a Texas gang
member; (2) getting caught up again in the race politics; and (3) as a consequence,
possibly being driven to violence that could earn him a third strike. Given all this,
who could blame him for pretending to be gay in order to access a unit that he
198.
199.
200.
201.

Details of this interview are drawn from my field notes dated July 9, 2007.
Or, as he put it, he’s “going to get stuck.”
See discussion infra, at Part II.D.2.
See CAL. PENAL CODE § 1170.12 (West 2011).

42

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

believed, not inaccurately, would be free from these problems?202
In most cases, those who sought refuge in K6G by lying about their sexuality did
so for a more prosaic reason: they wanted a respite from the stress and danger of
the gang politics that govern life in GP. I met one “OG” (Old Gangster, i.e., a
longtime gang member) “shot caller” (i.e., a gang member in a leadership role)
who was tired of the gang life and no longer wanted the pressure and responsibility
that came with his status. (Think corporate vice president in an extremely
high-stress business who is tired of managing the forty people who report to her.)
And I met several other men who, having spent considerable time in GP on prior
visits to the Jail, simply wanted a break from the pressure of an environment in
which gang politics govern. As one person put it, “it’s crazy down there.”203
More will be said in the next section about the Jail’s gang culture and K6G’s
insulation from that culture.204 For now, it is enough to consider what it suggests
about conditions in the Jail more generally that so many men would willingly
expose themselves to the stigma of homosexuality to protect themselves from the
stress and danger awaiting them in GP. The state has an obligation to ensure the
safety of those it incarcerates.205 But judging from the daily parade of people who
pretend to be gay in order to avoid the dangers of GP or other housing units, the
state has fallen well short of honoring its obligations206 to many of the 166,000
people on average who come through the Jail every year.207 This should trouble
anyone committed to ensuring that the state satisfies the carceral burden it assumes
when it decides to put people behind bars.208
This larger point should not be lost. Here is one way that making sense of the
K6G classification process turns out to provide a window on the system in general:
it reveals simultaneously the myriad dangers that people in custody routinely face,

202. Perhaps the most unusual explanation for trying to get into K6G under false pretenses that arose during
my field work was offered during an interview I did not observe, but which Lanni told me about later in the day. In
that case, three young men had driven down to L.A. together from Oakland for the ATB, a black gay pride party.
See AT THE BEACH: LOS ANGELES BLACK PRIDE TOY DRIVE & TOY GIVEAWAY, http://www.atbla.com (last visited
Mar. 6, 2011). They were arrested together and, on admission to the Jail, each said that he was gay and was sent to
K6G. During his interview, one of the young men gave Lanni odd and inconsistent answers to basic questions
such as, “do you have a lover?” and “how long have you known him?” At some point (as he often does), Lanni
told the interviewee to stop messing with him and tell him what was going on. It turned out that the man being
interviewed was not himself gay. However, his brother—who at that moment was sitting out in the hallway
awaiting his own interview—was gay, and this young man wanted to stay with his brother.
203. See Field Notes, July 16, 2007.
204. See infra, Part II.D.2.
205. See Farmer v. Brennan, 511 U.S. 825, 833 (1994) (“Having incarcerated ‘persons [with] demonstrated
proclivit[ies] for antisocial criminal, and often violent, conduct,’ . . . having stripped them of virtually every
means of self-protection and foreclosed their access to outside aid, the government and its officials are not free to
let the state of nature take its course.”) (quoting Hudson v. Palmer, 468 U.S. 517, 526 (1984)).
206. By my use of “the state” here, I mean to refer to the polity in a general sense. Strictly speaking, the L.A.
County Jail is administered by the municipality of L.A. County and not the state of California.
207. See supra note 101.
208. For more on the state’s carceral burden, see Dolovich, supra note 36, at 911–22.

2011]

STRATEGIC SEGREGATION

43

the many dimensions on which the state fails to protect those in its charge, and the
depth of the challenges facing those officials charged with fulfilling the state’s
obligation in this regard. Almost by accident, with K6G, L.A. County has created a
space relatively free from the gang culture that fuels much of the violence in GP. It
has also successfully managed to maintain a physical boundary between those
classified into K6G and the rest of the Jail’s population—a feat apparently not even
replicated among those people classified into K10 and designated as high-security
“keepaways.” The evident desperation many detainees feel to get classified into
K6G is both a measure of this larger failure and a way to understand its precise
shape, right down to the details. Someone looking to understand the ways L.A.
County is failing to ensure adequate protection for the people it detains could learn
a lot from spending a few weeks listening to the wide range of reasons why people
who know full well that they do not meet K6G’s admissions criteria try their
hardest to convince state officials otherwise.
As to the official aim of K6G’s classification process—identifying those
individuals who meet the admissions criteria—from what I have observed, the
system that has emerged in L.A. County does the job as well as one could
reasonably hope. No doubt, mistakes are made.209 But my sense is that, to maintain
the safety of the unit, it is not necessary that the system be 100% accurate on its
own terms. Instead, what is required is a demonstrated official commitment to: (1)
an accurate classification process and (2) persistent monitoring of the dorm to
identify (and remove) any predators. In this way, most of those who pose a threat to
K6G residents will be weeded out, and those who are misclassified into K6G,
recognizing that they are subject to removal if the error is discovered, will have a
strong incentive not to draw attention to themselves by bad (i.e., predatory)
behavior. With these pieces in place, there is a good chance of establishing living
conditions for gay men and trans women that do not include daily harassment and
fear of assault. And, given what life is typically like in GP for members of these
groups, this result would be—and, as will be seen, in Men’s Central, already
is—nothing short of astonishing.210

209. Just how many mistakes are made is difficult to gauge. I asked my interviewees to estimate how many
people in K6G at the time had been misclassified and did not belong there. Answers ranged from two percent to
twenty-five percent. This variation should not be surprising: given the desirability of placement in K6G, those
who are not in fact gay are unlikely to admit the fact, which means that my interview subjects could really only
guess.
210. This is not to say that life in K6G is a panacea. In many ways, conditions in K6G mimic the unpleasant
and even dangerous conditions of life in GP. And the absence of gang control in K6G actually means that there are
more frequent one-on-one physical altercations in the K6G dorms. For further discussion of these differences, see
Dolovich, supra note 45, at 1–3, 7–16.

44

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

D. Is K6G Safer?
1. Direct Evidence
K6G was established to keep its residents safe from rape and other forms of
sexual assault. Has it succeeded? It is impossible to know how much sexual abuse
K6G residents would have suffered in the absence of this unit. But in light of what
is known about the vulnerability of gay men and trans women in the general
population of men’s prisons,211 it stands to reason that being housed separately
from GP will make these groups safer.212 And indeed, K6G residents themselves
widely attest to the heightened security that they feel in the unit. In the summer of
2007, I conducted in-depth qualitative interviews with a random sample of
approximately ten percent of K6G’s residents.213 I asked subjects 176 questions,
several of which probed the issue of relative safety. Taken together, the answers
given to these questions overwhelmingly confirm the effectiveness of total
segregation as a protective measure.
My respondents almost unanimously reported feeling safe in K6G, and almost
all testified to feeling safer—from physical harm,214 sexual harassment,215 and
sexual assault216—than they would in GP.217 One question asked: how safe do you
feel in K6G218: very safe, safe, unsafe, or very unsafe?219 Of the thirty-one
interviewees to answer this question, only two reported feeling anything less than
safe—a remarkable finding given the well-recognized vulnerability of gay men
and trans women in carceral environments generally. Moreover, both of the people
in that minority of two reported feeling a mix of “safe and unsafe,” and offered
explanations for their mixed responses not inconsistent with an overall sense of
security as compared with GP. As to the first, he made clear that he “feel[s] safe
amongst the inmates,” but that he regarded the deputies who came into the dorms

211. See supra Part I.
212. Again, gay men and trans women are not the only groups in need of greater protection from prison rape.
See supra note 8. I return to this issue below. See infra Part III.B.
213. See supra note 23. For a detailed description of the research protocol, see the Methodological Appendix.
The questionnaire is reproduced below as Appendix B.
214. See Appendix B, Q121.
215. See Appendix B, Q123.
216. See Appendix B, Q91.
217. The exceptions here were two interviewees with extremely muscular physiques—one of whom a former
boxer—who both said they felt able to protect themselves and therefore felt safe in any environment. However,
even the former boxer, a trans woman, said elsewhere in the interview that if she were attacked by a group of men
bent on rape—a circumstance she implied was not unusual in the general population—she would be able to “take
out” some of them but ultimately expected she would be forced to succumb to the others.
218. Although officially, the unit is now referred to as K6G, at the time of my interviews, most of my
respondents—many of them with a long history of detention in the unit—still referred to it by its previous
referent, K11. My interview questions therefore used the term “K11” instead of “K6G.” To avoid confusion, in
quoting from my interviews, I have changed all references to K11 to reflect the current designation of K6G.
219. See Appendix B, Q125.

2011]

STRATEGIC SEGREGATION

45

as posing a real threat220—a feeling he explicitly connected to an incident that had
occurred in the K6G dorm the previous evening, in which several officers
reportedly used excessive force against a K6G resident.221 As for the second, this
person told a more particularized story, one that suggested an even more complicated set of gender dynamics in K6G than I was able to unearth in my time in the
Jail.222 But the source of the discomfort, however troubling, did not reflect the fear
of sexual or physical violence against which the unit was intended to protect.
A further set of questions asked interviewees to consider three different
locations—K6G, GP, and “out in the community”—and to rank the three in terms
of their feelings of relative safety from physical assault, sexual harassment and
“being yourself.”223 Here too, the results confirm the heightened security gay men
and trans women feel in the unit. In the majority of cases, interviewees felt safest
out in the community, followed by K6G, and then finally GP. Remarkably,
especially among the trans women, K6G was frequently named as the place they
felt safest, followed by out in the community. But with the exception of two or
three respondents whose physical size and long experience in prison made them
confident they could handle themselves equally well in any environment, interviewees uniformly named GP as the context in which they felt least safe from physical
assault and least able to be themselves without danger.224
Still further evidence of the extent to which the Jail has successfully created a
safer space for these vulnerable groups is found in a number of comments made by
interviewees during the course of the interviews. For example, one respondent, in
explaining why he preferred K6G to GP, said “I didn’t have to hide who I was. I
could be myself and not have to worry about . . . being in any kind of danger that
was only because of my sexuality.”225 Another, explaining his preference for K6G,
said simply “[b]ecause they’ll kill me in GP.”226 Another attested that in K6G, he
can “get in [his] bed and relax without having someone come and sexually

220. Int. 75, at B12. As this interview subject memorably put it, “[t]here’s a group of these deputies that they
literally scare the [expletive] of me, and I don’t scare very easily.” Id. For further discussion on interactions
between K6G residents and custody staff, see infra Part III.A.
221. This account of the incident was corroborated by several other dorm residents. Unfortunately, it does not
necessarily distinguish K6G from GP in terms of detainee safety vis-a`-vis deputies.
222. This subject described a personal preference for shifting back and forth between presenting as female and
presenting as male—in other words, to shift between the identity of “gay boy” and that of “queen.” Int. 68, at G14,
H1-2. In K6G, however, this person felt considerable pressure to suppress that desire and to make a choice. For
the most part, in the dorm, this person would present as a “girl.” But because s/he was not entirely comfortable in
that role, s/he felt at constant risk of being disciplined by those K6G residents who took it upon themselves to
police the gender line. Hence, s/he found K6G “both safe and unsafe,” a feeling that arose from the need to
“constantly monitor myself in my actions.” Int. 68, at G14.
223. See Appendix B, Q119–Q124.
224. See supra note 217.
225. Int. 48, at F2.
226. Int. 92, at B1. This respondent continued: “[In K6G], I won’t have to worry about, you know, when I’m
taking a shower, to watch my back. In the general population . . . [y]ou have to watch your back all the time.” Int.
92, at B8.

46

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

harassing [him]. [In GP, he] couldn’t relax because you have the guys that want to
run in the showers and want to have sexual things done to them.”227 Another said
that “a lot of these guys in [GP] would expect it from us because of the fact that oh
well, he likes guys, so might as well fuck him . . . .”228 One respondent, who felt
generally safe because of his “strong personality[,]” still felt safer in K6G because
in K6G, unlike in GP, you “don’t have people that have the need to punk
anyone.”229 Finally, one interviewee explained that in prison, most of the prisoners
in protective custody are homosexuals, and that they ask to be put in protective
custody because they “know they can’t handle theirself if they go to [GP]. You’re
getting raped constantly by six, seven people at one time. [And] that’s scary . . . even
for me.”230
Also noteworthy is the range of answers given to the question: if you had five
words to describe life in K6G, what would they be?231 As one would expect from a
description of life in jail, several of the listed words carried a negative connotation.
These included “noisy” or “loud” (3),232 “nasty” (2), “hateful” (2), “sour” (1) and
“depressing” or “sad” (7). But of the negative words offered, only two (“anxiety
producing” (1) and “stressful” (1)) suggested anything of the tension, danger, and
fear of violence one would expect to hear from people describing life behind
bars—especially people known to face a high risk of sexual victimization while
incarcerated.233 And taking the responses to this question as a whole, what is most
remarkable are the number of words that suggest a positive experience of
incarceration in K6G, including “fun” or “wow” (8), “exciting” (1), “easy,”
“easier,” or “easy-going” (4), “relax” or “relaxing” (2), “nice” or “good” (3),
“peaceful” and “calm” (3), “learning experience” (3), and “serene” (1).
That a sizable portion of K6G residents have positive associations with life in
the unit may seem shocking and hard to credit. After all, K6G is a unit of the L.A.
County Jail, its residents prisoners of the state. Denied their liberty, they are held in
often crowded, unhygienic, and unpleasant surroundings for as long as the state
chooses to keep them. No doubt, however much they might prefer K6G to GP, the
vast majority of residents would rather be back in the free world and out from

227. Int. 102, at D5.
228. Int. 71, at C5.
229. Int. 200, at D9.
230. Int. 53, at B13. This respondent, a trans woman, was an extremely muscular ex-boxer who said she would
fight rather than submit to forced sex. Even allowing for probable exaggeration, the message that GP is dangerous
for K6Gs was nonetheless clearly conveyed.
231. See Appendix B, Q35. I then followed up by asking for an explanation of each descriptor offered. These
questions, which proved very effective in eliciting a picture of life in K6G, were Joe Doherty’s idea.
232. The numbers in parentheses refer to the number of respondents who offered this particular term.
233. It bears noting that the responded who offered “anxiety producing,” along with several other words with
negative connotations (“miserable,” “hateful,” and “depressing”), stipulated that he would use the same words to
describe “being in jail, period.” Int. 131, at F11.

2011]

STRATEGIC SEGREGATION

47

under the Jail’s control.234 Yet the notion that many K6Gs nonetheless seem to find
life in the unit tolerable and even to some extent enjoyable is strongly suggested by
many of the descriptors my respondents offered. It is, moreover, entirely consistent
with my interviews as a whole, as well as with many casual conversations with
residents and my extended observations of life in the dorms. I have more to say
about this puzzling phenomenon elsewhere.235 For present purposes, it is enough
to observe that the picture of jail life painted by my respondents is a far cry from
what one would expect from those at high risk of sexual victimization behind bars.
Moreover, it offers powerful evidence that this segregation experiment has
succeeded in creating a relatively safe carceral space for otherwise vulnerable
people—one in which, in contrast to GP, vulnerable individuals need not sell
themselves to stronger prisoners to be protected from violent assault by (other)
predatory inmates.236
It might be argued that the data just described do not demonstrate that K6G in
fact keeps its residents safer than they would be in GP, but only that residents
subjectively feel safer—a feeling that may or may not correspond with the
objective fact of the matter. But further evidence suggests the direct relationship
between my subjects’ perceptions and their day-to-day experience. Subjects were
asked: have you ever had to do sexual things against your will with any other
inmates in K6G?237 The unanimously negative answers given in response to this
inquiry contrasted with the answers to a related question, which asked: have you
ever [over your whole incarceration history] had to do sexual things against your
will with other inmates? 238 In response to this latter question, one person reported
having been raped by a cellmate,239 another reported being twice forced into oral
sex,240 and a third offered an answer suggesting that he had faced the pressure to
engage in sexual conduct that arises in a protective pairing.241 Taken as given, the
answers reflect close to a 10% victimization rate in custody elsewhere than K6G.
But even this number may understate the actual experience of my subjects, since in
234. Although it would seem beyond peradventure that all K6G residents would prefer to be free, the fact that
a majority of the trans women I interviewed reported feeling safer from physical harm and more able to be
themselves in K6G makes me hesitate to offer what would ordinarily seem an uncontroversial generalization (i.e.,
that given the choice, all K6Gs would prefer not to be in jail at all).
235. See Dolovich, supra note 45, at 11–14.
236. Even in GP, people otherwise at risk of sexual assault may feel safe once they hook up with a stronger
person in a protective pairing, or once they secure membership in a gang. This feeling of security, however, is still
contingent. By contrast, the security K6G offers is independent of any protective arrangements individuals might
feel forced to create for themselves, and thus comes without the very high cost such coerced arrangements can
exact. I thank Jordan Woods for helping me to recognize this point.
237. See Appendix B, Q84.
238. See Appendix B, Q94.
239. Int. 102, at E10.
240. Int. 140, at C8.
241. See Int. 101, at A23 (“Not force, they don’t force you at all. They just give you the option that you can
choose for your own. So, you force yourself, actually.”). This answer recalls descriptions of the catch-22 faced by
vulnerable prisoners described in Part I above. See, e.g., supra note 51.

48

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

a number of cases, subjects who answered this question in the negative offered
responses at other points in the interview suggesting that an affirmative answer
may have been more accurate.242 By contrast, the uniformly negative answers to
the question regarding sexual victimization in K6G were corroborated by the
interviews generally, which overwhelmingly conveyed a sense of security in the
K6G dorms.
It is certainly true that a person’s subjective perceptions may at times be at odds
with reality. But when the question involves personal safety, it seems reasonable to
expect feelings of relative security to arise from actual personal experience, and
thus appropriate to credit the perceptions of otherwise vulnerable individuals that
they are safer in some conditions of confinement than in others. Ideally, the
conclusion as to K6G’s relative safety compelled by the data would be corroborated with more extensive quantitative findings. It is, however, unclear how one
could ever conclusively prove that particular individuals are in fact safer in
segregated housing (thereby confirming the subjective reports of my interview
subjects).243 Apart from the obvious lack of a control group, there are simply too
many variables to be confident in such conclusions. This being so, qualitative
assessments by otherwise vulnerable parties with direct personal experience in the
Jail may be the best available mechanism for appraising K6G’s relative success.244
2. Corroborating Circumstantial Evidence: Gang Politics in the Jail—But Not
in K6G
In addition, over the course of my research, corroborating evidence of K6G’s
relative safety arose from an unexpected quarter: those heterosexual men being
held in the L.A. County system who, although knowing themselves not to satisfy
K6G’s admissions criteria, pretended to be gay to get into the unit.245 Certainly,
these efforts may only reflect a widespread perception as to K6G’s relative safety
and not the reality of the matter. But if there were nothing to this belief, it is hard to
see how it would continue to be as widely and deeply held as it is. Yet it is not
unusual to find newly admitted detainees—sometimes four or five a day—lying
about their sexuality during their initial classification interview in order to be

242. This reticence in the face of a direct question about sexual victimization is perhaps not surprising. See
supra note 3. For this reason, the overall impressions conveyed in response to questions regarding perceptions of
personal safety may be the more trustworthy data.
243. See infra note 309.
244. It may of course be that, in spite of the randomness of my sample, see infra Methodological Appendix, the
experience and perceptions of my subjects were atypical, and that others with experience in K6G would offer a
different assessment of the relative success of the unit. And my sample necessarily did not include those
individuals who, although eligible for K6G, chose on the basis of prior experience in K6G to go instead to GP on
subsequent turns in the Jail, believing that they would be safer by doing so. Were such individuals to exist in any
numbers and could they be identified, they would prove a valuable source of insight into K6G—and possibly of
counterevidence as to K6G’s relative success at its institutional purpose.
245. See supra note 176, and Part II.C.2 generally.

2011]

STRATEGIC SEGREGATION

49

classified to K6G.246 Given the stigma that attaches to gay men in many
communities in the United States, not to mention the harassment and assault gay
men typically suffer while incarcerated, it may be hard to fathom why men who are
not gay—especially those facing imprisonment—would try to pass as such.
Certainly, it is not uncommon for people to pretend to be a different sexual
orientation, but these efforts are more typically made by gay people pretending to
be straight. How then to explain these daily efforts at reverse-passing? The reason
is simple: those familiar with L.A. County know K6G to be the safest place in the
Jail.247 That this is so, of course, stands as strong condemnation of the Jail as a
whole, for its failure to provide anything close to adequate security in its GP units.
But it nonetheless raises the possibility that L.A. County may have something to
teach other institutions about how to protect those who are most vulnerable to
harm behind bars.
Several possible reasons for this reverse-passing phenomenon have already
been explored.248 This section will focus on what appeared to be the most common
explanation: the respite K6G provides from GP’s endemic gang politics. To
understand this phenomenon, it is necessary to understand a little about the Jail’s
gang culture. The gang culture in L.A. County bears a close resemblance to that in
the California prisons.249 In L.A. County’s iteration of the system, the prisoner
population is divided into four “gangs” or “races” (in this context, these terms are
interchangeable) that structure to a considerable extent the behavior of and
interactions between individual prisoners: Blacks; Whites; “Surenos” or “Southsiders,” who are native-born Latinos from south of Fresno250; and “Paisas” (slang for
illegals), who are foreign-born Latinos.251 As these groupings suggests, the
organizing principle of the system is race-based, although, as with racial categories
in general, the meaning of race in this system is constituted within the cultural
context. Thus, Southsiders and Paisas are referred to as different “races,” although
each group is comprised of Latinos. And anyone who does not identify (or is not
identified by others) as white, black, or Latino—who is, say, Asian, Native
American, or Middle Eastern—is known as “Other” and is automatically assigned

246. For further discussion on the Jail’s intake process, including a description of the initial classification
interview, see supra Part II.C.1.
247. See supra Part II.C.2 (discussing this phenomenon in greater detail).
248. See id.
249. See Dolovich, supra note 45, at 3–7, for more on the interplay between the gang politics in L.A. County
and those in the California state prison system.
250. In the California prison system, there is a deep hostility between the Surenos and the “Nortenos,” who are
native-born Latinos from north of Fresno. The California Department of Corrections does its best to house
Surenos and Nortenos in separate facilities. L.A. County, unsurprisingly, is dominated by Surenos, and any
Norteno unlucky enough to find himself in L.A. County would likely request protective custody for his own
safety. For more on this division, see Dolovich, supra note 45, at 4–5, n.20.
251. Anyone who does not fit one of these four designated categories (i.e., who qualifies as an “Other”) is
expected to “run with” the blacks, although they may have to pay a tax to do so.

50

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

to the Blacks.252 This is the logic of “racial” distinctions in the L.A. County Jail. To
complicate matters, these racial groupings are distinguishable from the prison
gangs, organized on racial lines, which dominate the California prison system and
have the power to reach into the Jail. These gangs, which include (among others)
the Mexican Mafia, the Black Guerilla Family, and the Aryan Brotherhood, have
restricted membership, but the leadership of each makes the rules that must be
followed by all prisoners “assigned” to their respective racial groupings.253
In this culture, strict rules govern individual behavior and interaction. At their
most basic, these rules arise from two foundational principles: racial segregation
and mutual “respect.” The corollaries of these two principles are the two cardinal
sins: racial mixing and interracial disrespect. Behavior is strictly controlled and
rigidly policed by the gangs to guard against transgressions, and the commission of
any offense may bring swift and violent reprisal, often from the wrongdoer’s own
group.254
The gang-enforced rules that define life for prisoners in GP reflect this moral
universe. First, a set of detailed precepts designed to prevent racial mixing govern
everyday conduct. Between individuals of different “races,” there can be no
touching, no sharing of food or utensils, and no overt displays of mutual regard.
People of different races can talk to one another, and depending on the dorm, may
even be able to play cards or chess with one another. But even in that case, they
would not be able to sit on the same bunk to do so; one person would have to play
standing up next to the bunk on which the cards or game board are placed.
Bunks too are designated by race to ensure that the top and bottom of each bunk
bed are occupied by two people of the same race.255 Members of different races
cannot use the same toilets or phones; for this reason, in each dorm, each phone
and toilet is designated as “belonging” to one race or another.256 Even the water

252. See supra note 251.
253. Thus, for example, the leadership of the Mexican Mafia, along with leaders of the other Latino gangs
(e.g., La Nuestra Familia), makes the rules for the Southsiders; the Aryan Brotherhood, along with leaders of the
other white gangs (e.g., Nazi Low Riders), makes the rules for the whites, etc.
254. This description should in no way be taken as evidence of endorsement. I am well aware that the cultural
system I am describing here is deeply offensive and troubling. But to understand K6G and the difference it
represents, it is necessary that the larger gang culture be understood, which is why I am describing it in such detail
here.
255. This particular practice may have to change in the near future, at least in the California prison system.
After the Supreme Court handed down its decision in Johnson v. California, 543 U.S. 499 (2005), the state agreed
to a policy of racial integration in all its facilities. This policy commits the state to racially integrating as much as
possible both cells and dorms. The experience of such integration in Texas suggests that the racial integration of
cells should have no appreciable effect on the level of violence. See infra note 382. As to the dorms, however, the
public nature of the living quarters may mean that even those prisoners who do not object to sharing a bunk bed
with someone of another race will feel compelled to perform disgust, anger, and recalcitrance at the integration.
This difference may prove to complicate California’s integrationist efforts in its prison dorms.
256. There was some suggestion in my conversations that at least sometimes, the phones and possibly the
toilets were divided up between blacks on the one hand and everyone else on the other, as opposed to one per
group.

2011]

STRATEGIC SEGREGATION

51

fountains in the recreation areas have been assigned according to race. If you are in
the rec areas and there is no fountain designated for your group, you cannot get a
drink. Compliance with these rules is enforced by the gang leadership in each
dorm. In each dorm, each gang has a “rep,” for four in total, whose responsibility it
is to make sure that his “soldiers” comply with these rules. New arrivals are taken
aside and informed of their obligations, and violations are met with reprisals
ranging from strong rebukes to physical violence.
A second set of rules, designed to guard against any signs of mutual disrespect,
provides for an almost scrupulous equity in the distribution of benefits. What
benefits could possibly be had in a jail dormitory? Apart from the much-soughtafter commissary items, which are available to anyone who can afford them or who
cuts a deal with someone who can,257 the main benefits are two: being the first to
get your meal at chow time and getting a say in the channel to which the dorm’s
communal television is set. Consistent with the demands of racial consistency and
the imperative not to be disrespected, the gangs in the Jail have worked out a
system. As to both being first in line at mealtimes and choosing what TV show to
watch (within the parameters set for available programming by the Jail itself), the
prizes rotate. The gangs take turns being the first to line up for chow, and on any
given day, the group whose turn it is to go first at mealtimes gets to decide what to
watch on TV that day.258

257. As one interview subject explained, in the Jail there are:
these people that do two-for-one, and you go to them and you get, like, one cookie, you’ve got to
pay two back. And sometimes I get like $60, $75 in two-for-one, and by the time the store comes I
have to shovel out all my store, because you can only spend 130 bucks. So, if I go to the store and
spend 130 bucks, and $95 of it or 100 bucks of it is me getting two-for-one, then I’ve got like 30
bucks in there for myself. And I run out of that.
Int. 136, at C18.
What happens if people don’t pay their debts?
Oh, you can get in trouble. Some of them are very violent to people. Some of them want to whup
your butt and they threaten you. I mean, the people that are two-for-one are people who are like
drug dealers, actually, and they’re not very nice people. It’s not very good to deal with them,
actually. And one of them that I deal with is a very nice guy, I like to pay him, because he gives me
two-for-one. And I got myself in a pattern to where it’s like a cycle for me. I have to pay him back,
so then I have to go back to him again for more, because I ran out. So it’s like a cycle.
Id.
A number of my interview subjects recalled one serious violent incident that had occurred not long before in the
K6G dorms, which arose when a debtor could not pay. I was unable to confirm this account, nor am I able to say
how often violence stemming from unpaid debts occurs in K6G.
258. I witnessed these rules in action in the GP dorm that served as a (fortuitous) control during my time in the
Jail. The dorms in Men’s Central are arranged in groups of four, with a single officer’s booth affording visual
access to, and some interaction with, residents of all four dorms. (For more on this configuration, see infra
Methodological Appendix.) In the summer of 2007, one such grouping included the three K6G dorms and one
dorm that served for a time as a general GP medium-security dorm. I learned a lot about GP from observing that
medium-security GP dorm and from informal conversation with some of its residents. I later confirmed what I
learned from that vantage point in conversations with custody officials and residents of K6G.

52

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

The rigid observation of these rules means that life in GP can often appear
remarkably calm. Indeed, this relative calm is arguably what leads Jail officials to
tolerate and even facilitate gang control of the internal jail culture, since for much
of the time it ensures order and stability. But this seeming calm masks the intense
stress created for GP residents by the imperative to follow the rules or risk physical
punishment. It also masks the ever-present possibility of collective violence. In
this highly calibrated system, collective violence—i.e., riots—can break out at any
time, prompted by anything from a perceived slight of one person by someone of
another race, to long-brewing tensions among different racial groupings, to a
decision made by gang leadership—usually far off in Pelican Bay or some other
high-security California prison—to launch an all-out war with another gang.259
Often, the people who fight will not even know why they are doing so. But their
knowledge of the reasons is irrelevant. This brings us to the preeminent obligation
for all prisoners caught up in this gang culture: the imperative to “jump in” (i.e.,
join in the fight) whenever the signal is given. Those who fail to respond to this
signal know that they can expect to be violently punished by their own gang once
the dust has settled.260
The system just described, with its rigid code of conduct and penalties for
violations, is what is known in the Jail as “gang politics” or just “politics.” These
politics make life in GP scary, stressful, and dangerous.261 In K6G, however—and
here is the key to understanding the appeal of the unit for many of the people Bell
and Lanni meet—there are no politics.262 Anyone can use any phone, any toilet,
any shower, without fear of being violently “disciplined.” No one prevents people
of different races from sitting together, sharing food or utensils, touching, kissing
or otherwise being intimate with one another. As a result, this kind of interracial
mutual engagement is routine. At meal times, everyone crowds together and

259. The most senior gang leaders are typically housed in California maximum-security facilities. But they are
still able to reach into the Jail dorms through various means, most notably the frequent transfer between facilities
of individual gang members, who serve as messengers.
260. As one of my (black) respondents succinctly explained, “[i]f a Mexican and a black fight, and another
Mexican jumps on the black and beat on the black, I may be called to where I have to jump in and fight. And if I
don’t, then the blacks may all beat me up later.” Int. 119, at C4.
261. This effect came through clearly in my interviews, as subjects described their experiences of life in GP. As
one (white) respondent described,
I was scared to death. Because where I was [housed], I was with nothing but Mexicans. They were
all gang bangers [i.e., someone deeply involved in the gang culture], every one of them were gang
bangers. I forget what clique they were from. But in [the overhead light in my cell] we had 32
shanks, knives, handmade knives. And then one day somebody disrespected one of the Mexicans,
and the Mexicans they all went off on the whites. The only reason why they didn’t go off on me is
because our tank had all those shanks in them. And that’s the only thing that saved me from being
jumped on by six other gang bangers.
Int. 123, at E6.
262. In my interviews, the absence of such politics in K6G was raised repeatedly as a welcome feature of life in
the unit.

2011]

STRATEGIC SEGREGATION

53

people get their trays on a first-come, first-served basis. To be sure, there are gang
members in K6G, and, on occasion, some of them “tr[y] to turn it into a political
thing,” as once happened when “a couple of the inmates [in one of the K6G
dorms] . . . tried to segregate it with blacks, whites, [etc.].”263 Or, people might
“start throwing up those gang signs or where they’re from.”264 But these efforts are
readily put down by other dorm residents.265 The effect is a housing unit
remarkably free from the constant threat of gang-related violence.
The absence of gang politics in K6G is a puzzle I take up at greater length
elsewhere.266 For present purposes, it is enough to note that gang culture thrives
where people anxious not to be seen as weak and therefore vulnerable to
victimization are willing to enforce the rules.267 Those “gang bangers” who
violently compel others to submit to the demands of the gang culture thereby
secure their own power and status, and broadcast that they are not someone to be
messed with. Gang banging, in other words, protects against aspersions of
weakness, of cowardice, of being a sissy—in short, of being female. This dynamic
reveals the deep connection between the culture of prison gangs and the fear of
rape, as those seen to be unwilling to adhere to the code open themselves up to
being victimized. As one of my interviewees might have put it, in prison, the
baddest gang bangers are the least likely to be punked.268 In K6G, by contrast,
there is no pressure on residents to prove their masculinity, and thus no risk that
those who display “feminine” qualities will be targets of sexual assault. As a
consequence, there is no premium on seeming hard or tough, suppressing
emotion, or instilling fear in others—all of which behaviors are demanded by
GP’s gang culture. It appears, in short, that, in the absence of pressures to
perform a hypermasculine identity,269 the rigid, irrational, racist, stressful, and

263. Int. 71, at A7.
264. Int. 131, at F4.
265. See, e.g., Int. 119, at B11 (explaining that gang politics are “not taken seriously in [K6G],” and “if they
was causing too much of a problem, Bloods or Crips [or] whichever, I’m pretty sure that we would probably whup
them . . . to stop problems for everybody”); Int. 53, at C11 (“We don’t play [gang politics] up there. We don’t play
that at all. . . . If they going to come up there gang banging . . . [I tell them] you’re in the wrong place. I’m one of
the ones that let them know.”).
266. See Dolovich, supra note 45.
267. As Craig Haney explains in his contribution to this issue, “[g]angs only flourish in a jail or prison society
when there is a strong undercurrent of fear and reminders of one’s own vulnerability.” Haney, supra note 49, at
136. See also id. at 135–136 (explaining that through the “racial gang culture . . . [p]eople who live under
conditions of scarcity, threat and alienation band together to create a sense of security and safety”).
268. Haney explains this connection at greater length in his contribution to this volume. See id. at 137. As he
puts it, “there is a kind of symbiotic relationship between the threat of prison rape and the existence of prison
gangs.” See id.
269. To say that there is no hypermasculinity imperative in K6G is not the same as saying that no residents of
K6G have an investment in appearing masculine. As Jeannie Suk points out in her contribution to this volume,
heterosexual men have no monopoly on masculine performance. See Suk, supra note 67, at 116. To the contrary,
as she notes, “the phenomenon of gay masculinity is well known.” Id.

54

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

scary rule structure that governs (gang) life elsewhere in the Jail can get no
traction.
III. K6G OR NOT K6G: THREE CRITICAL PERSPECTIVES
The aim of Part II was to describe the way K6G works and its implications for
the security of its target populations. I now turn to three key objections that might
be raised against the K6G program in particular and the segregationist enterprise in
general. Part III.A addresses what it terms the “antisegregationist” objection,
which would condemn any state-sponsored, identity-based segregation as “demoralizing and dangerous.”270 It argues that although antisegregationist concerns are
valid ones that bear addressing, they are insufficient to justify abandoning the K6G
program itself. It also notes a key irony for those who would take this position: by
separating out detainees on grounds of sexual orientation and gender identity, K6G
has become one of the only places in the Jail where detainees of different races can
intermingle without consequences. Part III.B addresses the objection that K6G’s
classification standards are sorely underinclusive. As it shows, the underinclusivity
concern in fact represents a cluster of objections to the K6G model: that K6G
should be open to anyone who asks; that the Department’s stated admissions
criteria are too narrow because they self-consciously exclude bisexuals; that the
classification process is inadequate to identify all those who meet existing
classification criteria; and that those criteria fail to take account of all who are at
risk of sexual victimization in GP. Although the first three of these objections are
ultimately shown to have little weight, the last concern—that K6G’s admission
criteria exclude many other at-risk detainees—proves to pose a serious challenge
to the K6G model. As it happens, however, an alternative approach exists that
would simultaneously mitigate antisegregationist concerns and remedy K6G’s
underinclusivity. This alternative, consistent with the classification standards
recommended by the National Prison Rape Elimination Commission (the Commission),271 does not entirely avoid the risks of K6G’s identity-based strategy, nor is it
without its own dangers. It should not, moreover, be allowed wholly to crowd out
the K6G model, which may prove to best serve the needs of particular carceral
institutions. Still, Part III.B concludes that on balance, the Commission’s more
inclusive approach is the more advisable default model, although it argues that,
despite the shortcomings of the K6G approach, L.A. County’s K6G unit should
remain in place. Part III.B also responds to the objection that any strategy designed
to prevent prison rape by removing the most vulnerable prisoners from GP is
doomed to fail, because the hypermasculinity imperative will simply push the
remaining prisoners to target the weakest remaining individuals for victimization.
Finally, Part III.C responds to the argument that, even for jurisdictions like L.A.

270. COMMISSION REPORT, supra note 8, at 80.
271. See supra note 37.

2011]

STRATEGIC SEGREGATION

55

County, where the K6G approach has proved an effective way to protect their gay
and trans populations, the K6G model would be off-limits as unconstitutional after
the Supreme Court’s decision in Johnson v. California.272 As it shows, this
argument misconstrues the implications of Johnson, which, in cases where prison
officials can demonstrate the success of K6G at keeping target groups safer, would
not pose any constitutional obstacle to adopting or maintaining this model.
A. “Demoralizing and Dangerous”273: The Antisegregationist Objection
The history of race discrimination in the United States has left a legacy of
suspicion toward any official program of identity-based segregation. Even in the
prison context—not typically a site of meaningful constitutional scrutiny274—the
Supreme Court has condemned even temporary segregation on the basis of race as
a “highly suspect [policy] tool.”275
This suspicion is appropriate, and not only as to race; the experience in
Alabama’s Limestone prison of HIVϩ prisoners,276 who were divided from the
general population and crammed into a “drafty, rat-infested warehouse once
reserved for chain gangs,”277 stands as evidence that segregation to protect the
majority from contact with any despised and feared minority can be as cruel and
damaging to the segregated population as it was to African Americans in the Jim
Crow South. And, as Brown v. Board of Education famously taught, even where
“physical facilities and other ‘tangible’ factors may be equal,” state-sponsored
segregation implies an official judgment as to the “inferiority” of the targeted
population that can be demoralizing and psychologically damaging to members of
that group.278
Some version of this view appears to have motivated the Commission, which, in
its final report, “specifically prohibit[ed] housing assignments based solely on a
person’s sexual orientation, gender identity, or gender status,” on the grounds that
“[this] practice can lead to labeling that is both demoralizing and dangerous.”279

272. 543 U.S. 499 (2005).
273. COMMISSION REPORT, supra note 8, at 80.
274. See, e.g., Turner v. Safley, 482 U.S. 78, 89 (1987) (holding that policies burdening prisoners’
constitutional rights will nonetheless be upheld so long as they are “reasonably related to legitimate penological
interests”).
275. Johnson v. California, 543 U.S. 499, 506 (2005) (internal quotations omitted).
276. See generally BENJAMIN FLEURY-STEINER, DYING INSIDE: THE HIV/AIDS WARD AT LIMESTONE PRISON
(Univ. of Mich. Press 2008) (discussing the inhumane treatment of prisoners, such as being chained to their beds,
subjected to filth, and untreated for their medical needs in Limestone Prison’s segregated HIV ward).
277. Paul von Zielbauer, A Company’s Troubled Answer for Prisoners with H.I.V., N.Y. TIMES, Aug. 1, 2005, at
A1.
278. Brown v. Bd. of Educ., 347 U.S. 483, 493–94 (1954).
279. Id. The relevant draft standard, “SC-2: Use of screening information,” required that “[l]esbian, gay,
bisexual, transgender, or other gender-nonconforming inmates are not placed in particular facilities, units, or
wings solely on the basis of their sexual orientation, genital status, or gender identity.” COMMISSION REPORT, supra
note 8, at 77.

56

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

Viewed from this perspective, multiple aspects of the K6G program give cause for
concern.280 First, there is the classification process itself, which invites and even
requires state officials to make explicit judgments as to the sexual identity of the
people they detain, prying into the most private and intimate details of their lives to
do so. And second, there is the aftermath of this process, whereby individuals
determined by the state to be gay are explicitly and openly labeled as such and
placed in a unit widely known to house sexual minorities.
As to the first set of harms, it bears noting that a detainee who wants to protect
his privacy and perhaps his dignity from violation by the state could refuse to
answer any intrusive questions. In response to the initial question posed in
IRC—are you homosexual?—he could answer in the negative, and that would be
the end of the state’s intrusion into his personal business. Or, if he answers
affirmatively and winds up opposite Bell or Lanni, he could choose to reveal no
further personal information. The problem, of course, is that access to the
relatively safe space of K6G is conditioned on a willingness to expose the intimate
details of one’s life to state officials. This is a serious concern, but it has less to do
with official violations of privacy and dignity than with the limited availability of
safe space in L.A. County, a troubling but distinct issue to which I return below.281
Those who do reveal their private details to a classification officer may well feel a
sense of personal violation in doing so. But were this intrusiveness the only way
for County officials to identify likely victims in order to ensure their safety, and the
only harm to arise from classification to K6G, the privacy and dignity objections
would amount to a thin reed indeed on which to condemn the whole enterprise.
The carceral experience inevitably involves a loss of privacy282 and dignity.283
Much of this deprivation is unnecessary, and to this extent should be challenged
and remedied. But unfortunately, to some extent, such violations may be the cost of
keeping people safe in custody. In a deeply imperfect world, sometimes the best
one can hope for is to avoid the greater evil. And, in the decidedly imperfect world
of the contemporary American prison, the compromising of privacy and/or dignity
as a means to ensure physical safety is surely the lesser evil.
Of greater concern is the possibility of harm from openly labeling—“outing”—
certain people as gay in the carceral context. To this extent, being classified to K6G
carries a number of risks, the most obvious of which is the threat of stigmatization,
of being exposed to popular prejudice and hatred because of one’s identity. In my

280. I leave aside for now the concern that K6G’s admissions standards condition safety in the Jail on the
ability to convince classification officers that one is “really gay.” I return to this point and examine it in more detail
in Part III.B.
281. See infra Part III.B.
282. See Hudson v. Palmer, 468 U.S. 517, 526–28 (1984) (holding that prisoners cannot have a reasonable
expectation of privacy in their prison cells and that they are therefore not entitled to Fourth Amendment protection
against unreasonable searches while in custody).
283. See Dolovich, supra note 36, at 931–35 (discussing the many ways in which prisoners are often
unnecessarily dehumanized and deprived of their dignity).

2011]

STRATEGIC SEGREGATION

57

interviews, one subject explicitly noted the way harm of this sort can arise from the
official publicizing of difference. When asked for words to describe life in K6G,
this subject offered “discrimination” and “segregation.” As he explained,
[w]e’re separated as [K6Gs] and the blues, the light blues, they just stand out.
Everybody here know what a light blue is. When you walk through they know,
oh, that’s a homosexual or that’s a [K6G]. . . . We get segregated from people
because we’re gay. [And] that’s where the disrespect come in, when we go
outside of the dorm. We’re inside our own dorms, it’s perfect, it’s fine, it’s cool.
But when we step out, we have to get into the same frame of mind that some
people might not like us, some people might say something to us. And that’s
what discrimination is.284

On this account, what is objectionable about K6G is the way it singles out gay
men and publicly announces their sexual orientation, thereby exposing them to
disrespect and other discriminatory treatment when they come into contact with
others. Although this interviewee was the only one to link this treatment to the
color-coded uniforms, many others described the verbal harassment to which
K6Gs are subjected by GPs when outside the dorm.285 As was explained above,286

284. Int. 48, at F6.
285. Interview subjects were asked: Do you ever have interaction with GP? If so, would you describe the
interaction as negative, positive, both negative and positive, or neither negative or positive? How so? See
Appendix B, Q30–Q32. The responses made clear that with the exception of the trusties who come into the dorms
escorted by custodial staff to distribute meals and do clothing exchange, the interactions between K6Gs and GPs
occur almost exclusively when they pass each other in transit in the facility. (The one glaring exception was the
respondent who reported having been raped in the K6G court-line holding cell. See infra note 297.) The
characterizations of these hallway interactions were split almost equally between positive and negative. Of the
twenty-four inmates who reported some interaction, four characterized the interaction as positive, five as
negative, eleven as both negative and positive, and four as neither negative nor positive. Those who indicated only
negative interactions with GPs mainly described verbal harassment while waiting in line or during brief passing
interactions. See, e.g., Int. 131, at F10 (“[M]ost of [the GPs] got smartass mouths. . . . [They are] homophobic and
they call you faggot or things like that.”); Int. 75, at A14 (“They walk by our cells and spit on us and throw shit at
us and the deputies don’t do nothing but giggle.”); Int. 45, at A8 (“[T]hey cuss me out.”). Those who reported both
negative and positive interactions described a mix of unwelcome verbal harassment (and some spitting), and
welcome flirtation. Here is a brief sampling of such responses:
“[O]nly interaction . . . is pretty much when I go to court. . . . Sometimes there are some guys that
say hi, but there are some that wink their eye. . . . [Some] may make a comment and say
‘faggot’ . . . .” Int. 119, at A5.
“[T]hey do say names to us or they do call me a faggot or whatever, but I’m just, like,
‘Whatever.’ . . . [S]ometimes they flirt with you or sometimes they’re nice.” Int. 92, at B4.
“Some are positive . . . [but] some of the [GP], they don’t like us talking to them, for the simple
fact because we’re gay.” Int. 89, at C4.
“We [K6Gs] harass [GPs] . . . [K6Gs will] whistle . . . and try to embarrass [GPs] in front of their
people. And then, of course, they’ll look at us and give us very disgusting looks a lot of times.” Int.
79, at E9.
“More positive than negative. . . . Because you got guys that’s going to face time. A lot of them is
going to face a lot of time and stuff, and homosexuals is what they would probably mess with if

58

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

jail policies in place to physically separate K6Gs from GPs are reasonably
effective.287 But the siting of K6G within Men’s Central means that when K6Gs
are out of the dorms, they continually cross paths with GPs. When this happens,
the K6Gs will often be the targets of catcalls, whistling, or explicitly homophobic
epithets.
More troubling still, the verbal abuse K6Gs draw comes not only from fellow
inmates but also from staff—custodial officers in particular.288 Indeed, a number of
my subjects noted that at least some custody officers treat K6Gs more harshly and
less respectfully than they treat GPs. Interviewees offered several possible explanations for this differential treatment. A number of respondents suggested that
K6Gs themselves provoke the officers. As one person put it, “you can’t blame it all
on the officers, because sometimes [K6Gs] got smart mouths and they bring it on
[them]selves.”289 It was also implied that the officers are more intimidated by GPs.
As the same respondent explained, “the things they do in [K6G], they don’t do in
[GP]. . . . Like just fuck with us in general. Just talking crazy to us. They don’t do
that in the general population because general population tend to go off on them a
little bit.”290 Finally, others pointed to homophobia on the part of some officers.
One respondent found that “as a whole . . . they consider us less of a threat so
they’re a little more lenient. But then you have those [who] are homophobic, who
[are] just waiting for us to [not follow orders]” to lash out. And that, the subject
concluded, is “something I feel is so degrading.”291
Moreover, when the perpetrators are custodial staff, the verbal abuse to which
K6Gs are subjected is not limited to chance encounters in hallways. Many of my
respondents reported that some—though by no means all—staff assigned to the
booth overlooking the dorms felt free to use homophobic epithets over the
loudspeakers when making routine communications like announcing mealtimes or
count.292

they get the time. . . . [T]he next best thing to a whole female would be a homosexual. . . . And then
you got some that actually despises us, whether it have to do with their sexuality or a family
member that they have that’s the same gender as us, you know. And they can’t deal with that. We
get spit at a lot.” Int. 53, at B6.
“[When GPs walk past the K6G court cell], some of them just laugh and some of them just go with
it.” Int. 55, at 9.
286. See supra Part II.B–C.
287. For one notable failure, see infra note 297.
288. Jail staff includes not just custodial officers, but also nurses, teachers, chaplains, volunteers, etc. For the
most part, my subjects reported respectful treatment from members of these other groups.
289. Int. 140, at B1.
290. Id.; see also Int. 103, at D3 (“GPs get more respect. . . . I believe because they feel that, because we’re
wearing powder blue or because we’re gay, or transgender[], that we don’t pose too much of a threat, so they can
talk to us any way that they want.”).
291. Int. 50, at F7.
292. See, e.g., Int. 75, at A7 (reporting that one day before count, the officer in the booth made the
announcement by saying “[a]ll you faggots get on your bed[s]”).

2011]

STRATEGIC SEGREGATION

59

The verbal harassment of K6Gs in their dorms raises a further, still more serious
issue that might be thought of as the “sitting duck” concern: that concentrating all
sexual minorities in one unit would only draw attention to people who are already
extremely vulnerable, making them sitting ducks for any predators—whether
inmates or officers—bent on assault. This risk would of course exist as to any
vulnerable populations, but it may be greater still when the population is comprised of sexual minorities known to be targets of both popular prejudice and
physical violence because of their sexuality. Certainly, it seems fair to say that
K6Gs are sitting ducks when it comes to verbal harassment. Although unpleasant,
this may be inevitable. Were, however, the same to be true as to physical abuse,
this finding would be damning indeed, since maintaining the physical safety of
K6G residents is the whole point of the enterprise. Here, my research speaks
directly to this point and suggests reason for cautious optimism, at least as to
assaults by fellow inmates in the dorms.293 Even the subject quoted above, who
objected to the discriminatory treatment K6Gs often face outside the dorms, made
the point that “[when w]e’re in our own dorm, it’s perfect, it’s fine, it’s cool.”294
And still more to the point, the other words this same subject offered to describe
life in the unit were “comfortable” and “easy.”295
What, however, of staff? Unfortunately, restrictions placed on my data collection by Jail administrators precluded me from asking specifically about any sexual
contact, whether forced, coerced, or consensual, between K6Gs and custodial or
other Jail staff.296 Certainly, the readiness with which my respondents affirmed
their feelings of safety in K6G at least tended to suggest that unwanted sexual
contact with staff had not been a problem for them. Still, confirmation of this initial
impression must await further investigation, unhampered by the particular restric-

293. See supra Part II.D.1.
294. Int. 48, at F6.
295. Id. at F4.
296. In my original research protocol, I had indicated an intent to ask about sexual contact between K6G
residents and staff. But my protocol also explained that I would keep confidential all information given to me by
my interviewees (with the standard exception of anything suggesting that a particular interviewee was either
suicidal or homicidal). When I submitted my protocol to Jail officials for approval as required by the UCLA IRB
process, they protested that because any sexual contact between inmates and staff is illegal, they would need to be
informed if I discovered any information alleging such contact. Although I argued that if the stipulation were
insisted upon, Jail administrators could expect to learn nothing from my research on this crucial issue, these
efforts were to no avail, and ultimately, my commitment to disclose any such information to the Jail became a
condition of my research. As a consequence, at the start of each interview, I explained to each subject the
parameters of the confidentiality they could expect from me. Specifically, I explained that any information they
gave me regarding any sexual contact, whether forced, coerced, or consensual, between any staff member and any
inmate, whether themselves or someone else, would be disclosed to Jail staff. I further explained that for this
reason, if they did not want that information to be disclosed, they should not share it with me during the
interview—that, in other words, they should regard sharing such information as equivalent to making an official
report. (So as not to prejudice the interview in general, I was careful to make clear that it was only this thin slice of
information that would be treated this way.) Not surprisingly, I learned nothing in my interviews as to whether and
to what extent sexual contact occurs between staff and K6G residents.

60

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

tions I faced in my own research.
In any case, it is clear that, even if K6G succeeds in ensuring residents’ relative
safety, some residents are still victimized. At least one of my respondents reported
suffering the worst effects of a program that flags gay inmates with color-coded
uniforms: he was raped in the K6G court-line holding cell by a GP inmate who
threatened him with a razor.297 Even apart from this disturbing incident, the
persistent verbal harassment by custodial staff indicates that being publicly outed
as gay in the Jail carries risks. Given these dangers, what is the appropriate
response? At the most immediate practical level, it is incumbent on the Department
to take prompt action to address the most troubling abuses identified by my
research—the persistent verbal harassment of K6Gs by custodial staff, and,
considerably more serious still, the failure to protect K6Gs from physical assault
by GPs when outside the dorms.
As to the first, verbal harassment by staff, appropriate steps would include: (1) a
demonstration on the part of senior Jail officials of a commitment to zero tolerance
of such behavior; (2) the implementation of meaningful staff training, preferably
designed in consultation with K6G residents and leaders of the Los Angeles LGBT
community, as to appropriate ways to interact with K6Gs; and (3) the imposition of
administrative leave without pay for any staff member found to engage in the
maltreatment of K6Gs.
As to the second, keeping K6Gs physically safe, it is obvious that more must be
done to shore up the physical boundary between K6G and GP when K6Gs are out
of their dorms. At an absolute minimum, in response to the incident reported to me
of rape by another inmate, the Department should relocate the K6G court-line
holding cell so that it is directly in the sight line of officers, and should modify the
cell to have a key entry only. These simple modifications, moreover, are only a
starting point for considering what other changes are necessary to ensure the safety
of K6Gs when they are outside their dorms.298
297. The holding cells in the court line are designed for easy access. Each weekday morning in Men’s Central,
hundreds of men line up to be transferred to court. Officers determine which courthouse people need to get to and
send them to wait in the appropriate cell. The processing is facilitated by the cell design itself: the cells have
revolving entrances with horizontal metal bars like those one might see in a subway station, which allow someone
to enter simply by pushing the bars. As in a subway, the rotation is one-way. Once someone has entered the cell,
there is no exit except by an officer with the key. Unfortunately, the cell designated for K6Gs is not in the direct
sight-line of the officers, which creates the opportunity for GPs or other non-K6G prisoners to enter the cell freely
if they choose. One day, on his return from court, my interviewee reported that he was in the K6G holding cell
with fifteen or so others when a GP inmate with a razor in his mouth entered the cell and forced him to submit to
anal rape. The assailant then waited for an officer to release him from his cell. Certainly, had the victim wanted to
make a formal complaint about the incident, he could have readily done so. There were many witnesses, a few of
whom reportedly tried to intervene before being warned off by the razor, and the identity of the assailant was
obvious. But as is often the case with forcible rapes behind bars, the victim made a calculation that it was better to
say nothing.
298. The department should also take immediate steps to allow research into the incidence of sexual contact
between staff and inmates, which would require lifting the disclosure requirement that precluded the pursuit of
this issue in my own research. See supra note 296.

2011]

STRATEGIC SEGREGATION

61

Arguably, the risks created by explicitly flagging gay inmates may be necessary
to ensure their physical safety. But in creating that risk, the Department has a
heightened obligation to protect the people in K6G from the dangers to which its
own official strategy gives rise. Meeting this obligation would involve protecting
K6Gs not only from predatory inmates but also from predatory staff. Although
constraints placed on my research by the Department299 meant that I was unable to
determine the extent to which predatory staff members pose a threat to the physical
safety of K6Gs, evidence from other carceral contexts suggests that some such risk
certainly exists. The Department must therefore take all necessary steps to keep
unit residents safe from possible sexual predation by staff. Doing so would at a
minimum require ongoing monitoring of the unit, the sending of strong signals that
sexual contact of any kind between prisoners and staff will bring heavy consequences, and an official readiness to take seriously any and all allegations made by
unit residents of physical abuse by staff.300
It might be argued that, if L.A. County is going to retain its K6G unit, it ought at
a minimum to get rid of the color-coded uniforms that serve to publicize the sexual
identity of the wearers. Even taken alone, this system of color-coding may seem
distasteful and insidious, dehumanizing and stigmatizing the people in custody.
And, as a practical matter, given that gay men and trans women are favored targets
of sexual harassment and assault, one might well wonder whether distinctly
colored uniforms wouldn’t just make them more visible to potential predators and
thus more vulnerable. These concerns are legitimate ones. There is something
unpalatable about singling people out for special notice on the basis of a
stigmatized identity, an effect not lost on the residents; as has been seen, at least
one of my respondents noted the discomfort he felt when outside the K6G dorms
because of his light blue uniform. The choice of light blue for K6G is, moreover,
particularly unfortunate when juxtaposed against the GPs’ dark blue and thereby
appearing to suggest a feminized version of the male norm.
But for L.A. County to abandon the color-coding of uniforms on grounds of
these objections would be a grave mistake. Certainly, in an ideal world, any such
crude markers of difference would be avoided. The L.A. County Jail, however, is
not an ideal world. In this facility, as in any overcrowded, understaffed carceral
institution, violence is a constant possibility. With stakes like these, officials are
right to use every tool at their disposal to protect those in custody against assaults,
even tools about which people of good faith may be ambivalent—including
color-coded uniforms. Indeed, the physical safety of K6Gs arguably depends on
these uniforms. K6Gs use the same medical facilities, visiting rooms, court
transfer wing, and hallways as the general population. This situation creates

299. See supra note 296.
300. On the possibility that prisoner complaints might serve as a source of information regarding life in the
unit, see Dolovich, supra note 45, at 14–15.

62

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

endless opportunities for physical contact between K6Gs and GPs. To maintain an
impermeable boundary around the K6Gs is thus a constant challenge. To this end,
the color-coded uniforms are indispensable. They allow officers, for example, to
ensure that only those classified to K6G can access the unit’s dorms and to see
immediately if any non-K6Gs have gained admittance. The same holds for the
unit’s classroom, classification office, holding cell, and anywhere else exclusively
reserved for K6Gs. True, the uniforms signal to the general population that they are
in proximity to gay men. (Trans women are generally obvious to all regardless of
the color of their uniforms.) For this reason, K6Gs are often the targets of hooting,
catcalls, and homophobic epithets from GPs whose paths they cross. But remarkably, in all my discussions with residents of K6G, I learned of only one instance of
any breach of the physical boundary between K6Gs and GPs. And that tragic
instance—the rape referred to above301—strongly suggests that the better response
is not assimilation, but more effective segregation.302
All this, of course, begs the question: why should there even be a K6G unit in
L.A. County? Perhaps, given the risks involved in the state’s singling out of people
as members of sexual minorities, it would be better not to have any K6G at all.
However, assuming the choice were K6G or nothing, any move to disband K6G on
the grounds just enumerated would be a grave mistake. As the Supreme Court’s
Equal Protection doctrine explicitly recognizes, even the most suspect group
classifications may sometimes be necessary to realize urgent and legitimate state
interests.303 Under some circumstances, in other words, identity-based segregation
may be the lesser evil. When this is so, to prohibit such segregation may be to

301. See supra note 297. This incident occurred in the court line, to an individual waiting in the holding cell
reserved for K6Gs. As a policy matter, the answer to such an incident is not to put the K6Gs in dark blues, which
would only make it difficult to ensure their physical separation from other detainees, thereby exposing them to
physical assaults by anyone who discovers their sexual orientation. The better response is more effective
separation, which in this instance would involve relocating the K6G holding cell so it is directly in the sight-line
of officers, and requiring a key entry as well as a key exit to ensure the effective segregation of those inside it.
302. Without a doubt, there are risks involved in explicitly flagging certain prisoners as sexual minorities. But
notwithstanding the serious administrative challenges involved in running a facility as crowded, chaotic, and
complex as Men’s Central, the Jail does a reasonably good job of maintaining the boundary around the K6Gs.
There is certainly room for improvement. However, this achievement is still notable. And it is hard to see how the
same feat could have been accomplished without the color-coded uniforms. Again, however, the institution might
perhaps have done better to choose a more neutral color than light blue for K6G. Moreover, the fact that the
California prison system does not segregate gay men from GP raises the disturbing possibility that, by busing
K6Gs from L.A. County to state prison in light blue uniforms, the state is effectively announcing the sexual
orientation of these men to any GPs who are paying attention to the new arrivals. For this reason, it is imperative
that, if they do not do so already, Jail administrators must make a point of providing K6G transferees with the dark
blue uniforms of GP before they are removed from the K6G dorms in advance of being transferred.
303. See, e.g., Adam Winkler, Fatal in Theory and Strict in Fact: An Empirical Analysis of Strict Scrutiny in
the Federal Courts, 59 VAND. L. REV. 793, 794–98 (2006) (demonstrating that, contrary to the widely-held view
that strict scrutiny is “strict in theory and fatal in fact,” thirty percent of Equal Protection challenges analyzed
under strict scrutiny by federal courts between 1990 and 2003 resulted in the challenged law being upheld)
(quoting Gerald Gunther, The Supreme Court, 1971-Term Foreword: In Search of Evolving Doctrine on a
Changing Court: A Model for a Newer Equal Protection, 86 HARV. L. REV. 1, 8 (1972)).

2011]

STRATEGIC SEGREGATION

63

countenance and even facilitate potentially serious harm against the very population that the prohibition is intended to protect. This would be precisely the case
were the anti-segregationist concerns canvassed above found sufficient to justify
eliminating K6G. If the Department were to send the K6Gs back to GP, it would
certainly take the state out of the process of formally seeking to unearth and
publicly announce the sexual orientation of some of its detainees. But the effect
would only be to leave predatory inmates to do the scrutinizing, with very different
motives and a very different result.
There is, however, a second option: a more inclusive approach that would seek
to identify not which inmates are “homosexual,”304 but which inmates are, for
whatever reason, at a high risk of victimization in GP. This approach might well
eliminate the most troubling features of state-sponsored identity-based segregation
while correcting for the underinclusivity of K6G’s admissions standards. In what
follows, I explore the case for such an approach, and argue that it may be a wiser
choice for jurisdictions seeking to develop classification schemes aimed at
protecting inmates at heightened risk of sexual victimization. At the same time, on
a more local level, I argue that, notwithstanding the problems that attend a K6G
approach, L.A. County ought not to disband its existing K6G program, but should
instead supplement it with a second segregation unit for other vulnerable prisoners
who do not otherwise qualify for K6G.
This latter two-track approach may still not wholly satisfy those who object to
K6G on antisegregationist grounds. There is, however, an irony for those who
would on principle oppose any form of identity-based segregation: K6G, a unit
that explicitly segregates detainees on the basis of sexual orientation and gender
identity, is one of the few spaces in the Jail not defined by rigid segregation on the
basis of race. As noted, in K6G, there is no felt imperative to be hard and tough or
otherwise perform a hypermasculine identity.305 And, in the absence of pressure to
prove one’s manhood, the demands of the gangs—that members maintain strict
discipline, show no emotion or signs of weakness, and be ready to “jump in” and
fight at a signal from the “shot callers”—seem absurd. With the majority of K6G
residents not willing to play the game, and indeed affirmatively rejecting the game,
efforts to organize along gang lines, and thereby to impose the particular regime of
racial segregation the gangs demand, get nowhere.306 The complete racial integration of K6G is thus a byproduct—a happy accident—of the absence in the unit of
any hypermasculine imperative, which is itself the accidental byproduct of the
decision to segregate gay men and trans women from the rest of the Jail’s
population.
The point is not that the form of official segregation K6G represents should no

304. See Consent Decree, supra note 116.
305. See supra Part II.D.2.
306. See supra note 265.

64

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

longer trouble those committed to ending identity-based discrimination. But the
racial integration of K6G indicates how complex these issues are in the carceral
context. It also makes a clear case as to why, in this context, ideological purity is
generally both ill-advised and difficult to sustain.
B. Drawing the Line: The Underinclusivity Objection
The antisegregationist objection just explored rests on the harm it is feared will
be done to those admitted to K6G. By contrast, the underinclusivity objection
stems from the fear of harm to those not classified to K6G.307 Under this broad
heading, it is possible to identify a cluster of objections, each concerned with a
different subset of detainees excluded from K6G. First, it might be argued on
grounds of distributive justice that, given the considerable advantages of K6G,
access to the unit should be open to anyone who asks. Second, it might be argued
that the Department’s stated admissions criteria are too narrow to capture all those
who are attracted to men, and that, on grounds of fairness and consistency, the
criteria should be expanded to be more inclusive. Third, it might be argued that the
classification process fails to identify many detainees who actually meet the
Department’s stated criteria for entrance into K6G, and that, again on grounds of
consistency and fairness, the process should be modified to ensure their identification and classification to the unit.308 Fourth and finally, it might be argued that the
admission criteria themselves are the problem, because they only identify a subset

307. In this section, I focus on K6G in particular. But the objections explored here would arise in any facility
implementing a similar program.
308. In a forthcoming article, Russell Robinson at times appears to be making a version of this argument.
Robinson seems to object to the K6G classification process at least in part on the ground that it is likely to exclude
men of color and working-class men who, although they have sex with men (‘MSM’) are disproportionately likely
not to identify as homosexual or gay. See Robinson, supra note 115, at 122. Although Robinson challenges the
K6G classification process for privileging white middle-class men, he ultimately takes the position that the
segregation decisions should not turn on sexual identity at all but instead should be designed to identify all those
detainees who would be vulnerable to sexual victimization for whatever reason. Id. at 235–37. However, on this
approach (which I too endorse, along with the National Prison Rape Elimination Commission, see infra Part
III.B), many of those MSM on whose behalf Robinson condemns the current approach as underinclusive—in
particular, those MSM who because of reigning stereotypes of gay identity do not “seem” gay—would likely still
be excluded from a revamped K6G and sent to GP. This would not be, as is currently the case, because they chose
not to “come out” to the Jail’s classification officers, but because, not “seeming” gay, they are at lower risk of
being victimized on the mainline. At a minimum, this suggests that, to the extent that the process Robinson
condemns for relying on stereotypical notions of male homosexuality like effeminacy does so rely (which, it bears
noting, it does to a lesser degree than he imagines), this reliance may be appropriate. To be successful at its own
stated goal of protecting vulnerable prisoners, the K6G classification process need not succeed in identifying all
those individuals who are really, truly, fundamentally sexually attracted to men. Rather, it need only identify those
whose characteristics are likely to lead them to be identified as such in GP. To the extent that predators in GP are
likely to rely on the same features that Lanni and Bell take into consideration when making their decisions, the
process currently in place would seem to get it exactly right, notwithstanding any possible race or class bias of
relying on characteristics conventionally associated with gay identity.

2011]

STRATEGIC SEGREGATION

65

of those detainees likely to be at risk of sexual victimization in GP.309
Before turning to these concerns, it is necessary to underscore the limited scope
of the present inquiry and the inevitably partialist and unsatisfying nature of any
classification system given the current realities of mass incarceration. At base,
motivating each of the objections just articulated is the sense that the relative
safety enjoyed by K6G residents should be generalized to all people detained in the
Jail. Why should one have to be gay or trans to be protected from physical and
sexual violence while in state custody? And once the justness of this position is
granted—and who could oppose it?—the obvious conclusion is that the fault lies
not with the particulars of any given policy or practice, but with the carceral
system itself. And once the problem is viewed in this light, it begins to appear that
those seeking to protect people from being sexually abused behind bars should
focus not on crafting better classification standards for K6G, but on changing the
culture, structure, and operation of all carceral facilities to ensure the physical
safety of everyone society incarcerates.
It is hard to argue with the logic of this position. The problem is that there is no
reason to expect that these broad changes will be achieved anytime soon. The
challenges facing the L.A. County Jail—overcrowding, understaffing, aging and
decrepit facilities, entrenched gang control, etc.—are not exclusive to L.A.
County. They are shared by facilities all over the country and are the precise
conditions in which violence thrives.310 However fervently one might wish things

309. In conversations about this project, this last objection has often taken the form of a concern that, assuming
my account of the prison culture of hypermasculinity, see supra Part I, accurately explains why gay men and trans
women are at such a high risk of rape in prison, to remove these vulnerable populations to K6G would only shift
the attention of predators to other weak prisoners, thus exacerbating the vulnerability of those who are neither gay
nor trans, but are young, small, mentally or physically disabled, first-time or nonviolent offenders, etc. In the
absence of gays or trans individuals for predators to victimize, these other vulnerable people will become the
primary targets for being “turned out.” See, e.g., Suk, supra note 67; see also infra, Part III.B, text accompanying
notes 357–70 (addressing this concern at some length). This is not an unreasonable concern, at least absent
concerted efforts to provide healthier and more socially productive ways for all men in custody to assure others
(and themselves) of their own manhood. What it tells us, however, is not that gay and trans prisoners should not be
removed from GP, but that, until the broader causes of the culture of hypermasculinity can be addressed, all
prisoners known to be vulnerable to sexual victimization should be removed from GP. The answer, in other words,
is more targeted segregation, not less. As I discuss in the text, under present conditions, the key policy question
K6G raises for prison and jail administrators is whether all vulnerable people should be housed together, or
whether there might be reasons why gay and trans prisoners should be housed separately even from other
vulnerable populations. Of course, it would be far better if GP itself were not a site of such pressure and fear that
men feel compelled to be violent toward others to guard against being victimized themselves. The possibility that
K6G might only leave a different population vulnerable to the worst forms of abuse provides some indication of
the urgency of the need to reengineer the carceral experience not just for the especially vulnerable, but for
everyone.
310. As Terry Kupers explains,
[i]n crowded, noisy, unhygienic environments, human being[s] tend to treat each other terribly.
Imagine sleeping in a converted gymnasium with 150 to 200 prisoners. There are constant lines to
use the toilets and phones, and altercations erupt when one irritable prisoner thinks another has
been on the phone too long. There are rows of bunks blocking the view, so beatings and rapes can

66

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

were different, and however hard one might work to bring about the desired
change, there will remain the issue of what to do the meantime to keep safe the
people most vulnerable to abuse under current conditions. The question, in short, is
not whether, all things being equal, K6G is a perfect program, but instead whether,
given current realities, K6G represents the best possible approach. This brings us
back to the various iterations of the underinclusivity objection sketched above,
which I now address in turn.
First, it might be thought that, in light of the desirability of K6G, fairness
requires the unit to be open to anyone who asks. Given, however, that K6G is
widely known to be the unit that houses gay men and trans women, this strategy
would only represent an invitation for predators to gain direct access to those
whom they might well have sought to victimize in GP.311 This is not to say that gay
men or trans women may not also be predators. Indeed, given that access is
currently granted to anyone who can demonstrate their qualification to the
satisfaction of the classification officers, it is inevitable that some predators will
find their way into K6G. At present, when this happens, the response is to remove
such people to the K6G disciplinary wing as soon as they are found to pose a
threat. One might argue that the same approach could be taken with a more
inclusive policy, so that those who turn out to be predators could just be removed
as soon as their predatory nature comes to light, thereby not penalizing those who
would not otherwise qualify but who simply want to do their time in peace. There
is, however, a difference between accepting the inevitable risk that, within a broad
group of people in danger of victimization, some will inevitably turn out to be
predators, and creating the conditions whereby any detainees with predatory
tendencies may gain admission to the unit just by asking. In that case, to wait for
the predators to identify themselves by their predation would be to expose to
serious (and avoidable) risk the very people K6G is intended to protect.
Second, even granting the inadvisability of an open admissions policy, one
might still object that the current admission criteria are too narrow because they
self-consciously exclude bisexuals. This exclusion is of particular concern because

go on in one part of the dorm while officers sit at their desks in another area. The noise level is so
loud that muffled screams cannot be heard. Meanwhile the constant noise and unhygienic
conditions cause irritability on everyone’s part. Individuals who are vulnerable to attack and sexual
assault—for example, smaller men, men suffering from serious mental illness, and gay or
transgender persons—have no cell to retreat to when they feel endangered.
Terry A. Kupers, Prison and the Decimation of Pro-Social Life Skills, in THE TRAUMA OF PSYCHOLOGICAL
TORTURE 127, 130 (Almerindo Ojeda ed., 2008). Under these conditions, Kupers asks, “[i]s it any wonder that
research clearly links prison crowding with increased rates of violence, psychiatric breakdowns, rapes, and
suicides?” Id.
311. See supra Part II.A (explaining the dangers created for unit residents of L.A. County’s “homosexual”
module pre-K6G, when anyone claiming to be gay was admitted to the unit); infra note 392 (explaining how a
similar unit on Rikers Island wound up putting residents in jeopardy by housing them with “violence-prone”
inmates who falsely claimed to be gay).

2011]

STRATEGIC SEGREGATION

67

members of this group are also known to face an elevated risk of sexual abuse in
prison.312 Before the validity of this objection can be properly assessed, it is
necessary to say a little more about this aspect of the K6G admissions policy.
Recall that the unit’s classification officers define their mission as classifying to
the unit only those men who “live a homosexual lifestyle” on the outside, by which
they mean those who prefer men and thus seek out men—and only men—for
sexual gratification and intimacy.313 The seemingly flexible notion of a “homosexual lifestyle” notwithstanding, what motivates the K6G classification process is
a theory of (male) homosexuality that is both binary and essentialist: one is either
gay or not gay, and there is an essential truth of the matter. This approach is at odds
with contemporary understandings of sexuality, which regard sexual identity as
more complex and fluid than this reductionist account allows.314 But its lack of
theoretical sophistication aside, the dualistic theory of sexual identity that drives
the classification process has important practical significance: it allows for the
exclusion of what might be termed “situational homosexuals,”315 those men who
given the choice would prefer to have sex with women, but who, when in
circumstances not allowing access to women, will have sex with men. The original
federal court order that established K6G316 restricts access to K6G to “homosexual
inmates”—a term that as a technical matter excludes bisexuals. But my reading of
the determination on the part of Bell and Lanni to exclude bisexuals from
K6G—and they are determined to do so—is that it stems less from an unwillingness to deviate from the terms of the original court order,317 or from a view that
men who are sexually attracted to both men and women are not at risk in GP, than
from the sense that the vast majority of the men who seek admission to K6G on
grounds of bisexuality are really situational homosexuals.
Why should situational homosexuals not gain admission to K6G? The L.A.

312. According to the Bureau of Justice Statistics, “[jail i]nmates with a sexual orientation other than
heterosexual reported significantly higher rates of sexual victimization. An estimated 2.7% of heterosexual
inmates alleged an incident, compared to 18.5% of homosexual inmates, and 9.8% of bisexual inmates or inmates
indicating ‘other’ as an orientation.” Bureau of Justice Statistics Special Report: Sexual Victimization in Local
Jails Reported by Inmates, 2007, U.S. DEPARTMENT OF JUSTICE 6 (June 2008), http://bjs.ojp.usdoj.gov/content/pub/
pdf/svljri07.pdf.
313. There is, however, some flexibility in this definition, since Bell and Lanni often admit married men into
the unit, which suggests that a person can still be regarded as gay by K6G standards, even if he looks to women for
emotional intimacy and companionship.
314. See Jeffrey Weeks, Necessary Fictions: Sexual Identity and the Politics of Diversity, in SEXUALITIES AND
SOCIETY: A READER 122 (Jeffrey Weeks et al. eds., 2003). Ironically, it may turn out that the unsophisticated binary
approach does a better job of identifying those individuals most in need of a protection K6G provides. I return to
this point below.
315. See supra note 150.
316. See Consent Decree, supra note 116.
317. Indeed, that order said nothing about trans women, but K6G has housed the trans women together with
the gay men since the mid-1990s. See supra note 132.

68

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

County Jail houses close to 19,000 people at any given time, most of them men.318
Of these men, a very large proportion have previously spent time behind bars,
whether in juvenile hall, jail, or prison. It is thus likely that many in this
population, being human and thus having normal sexual desires and needs, will
have had sex with men while incarcerated—even if, given the choice, they would
prefer to have sex with women. Were K6G to admit everyone in the Jail who fell
into this category, not only would the size of K6G grow enormously, but
situational homosexuals would quickly become the overwhelming majority. Perhaps this shift would not put at greater risk those detainees who meet the current
admissions criteria. But there is reason to think that it would. As was explained in
Part I, the prison sexual culture puts pressure on men who have sex with other men
behind bars to simultaneously feminize and dominate their partners, both to prove
their masculinity to others and to ward off accusations that they themselves are
gay. This is a culture in which the situational homosexuals Bell and Lanni would
confront have experienced sex with men. The worry is that, were members of this
group admitted to K6G, they would reintroduce into K6G the notion that anyone
associated with femininity is someone to be dominated and forced into submission.
Indeed, given the way that the passive sexual partner in the sexual culture of GP is
defined not as male but as female, many of those detainees who could fairly claim
to have had sex with men in prison may not even regard themselves as having had
male sexual partners at all. Were they admitted to a new, expanded K6G, they
might well feel an even greater imperative to prove their manhood and thus even
greater pressure to perform a hypermasculine identity. This heightened pressure
could in turn place an even greater premium among residents on not seeming to be
gay, thereby reproducing the same cultural dynamics that put gay men and trans
women at risk in GP and which K6G has until now successfully guarded against.
Notice, moreover, that were the hypermasculinity imperative to reassert itself in
this way, all residents, whatever their sexual orientation, would feel pressure to
perform this identity,319 and could thus, as a preemptive move, become predatory.
This dynamic certainly exists in GP units, and there is no reason to think it would
not also exist in an expanded K6G. The point is thus not that men who self-identify
as gay are not capable of the abuses that can accompany the performance of
hypermasculinity in men’s prisons. It is instead that a widespread impulse to inflict
these abuses may be the likely effect of admitting to K6G all those who
self-identify as bisexuals solely because of sexual experiences they have had with
men while incarcerated.320 And with this effect could come the renewed victimiza-

318. In March 2010, there were 1,912 women in custody in L.A. County. See Personal Communication from
Deputy Bart Lanni, L.A. Sheriff’s Department, to author (Mar. 12, 2010) (on file with author).
319. Put differently, whether detainees are likely to feel pressure to perform the traits of hypermasculinity will
depend not on the sexual identity of the individuals involved but on the context.
320. This is not to challenge the sincerity of such self-identifiers. It is certainly possible and even likely that
men who before their detention would not have willingly turned to men for sex may, after having had sex with

2011]

STRATEGIC SEGREGATION

69

tion of trans women and men judged as gay, by whichever unit residents strong
enough to inflict it.
Still, even granting that the open admission of situational homosexuals to K6G
could have this dangerous effect, K6G’s exclusion of bisexuals is plainly overinclusive, since at a minimum it also denies access to people whose sexual attraction to
men predated their time in custody and is thus not a product of the prison sexual
culture just described. Given that this population is also known to be at heightened
risk of victimization in GP,321 this is without a doubt a troubling effect. Arguably, if
K6G’s admissions standards could be modified to capture only what might be
called, for lack of a better term, “pre-custody bisexuals,” this would represent an
improvement on the current model. However, failing this possibility, the need to
ensure the ongoing protection of K6G’s residents from likely predators may mean
that as a practical matter, pre-custody bisexuals may need to be housed with other
detainees who are at risk of sexual victimization but who are not otherwise eligible
for K6G. Some readers may find this way of dealing with the problem to be
non-ideal. But, as has already been observed, the complexities of prison administration in the present era of mass incarceration do not often admit of ideal
solutions. So long as vulnerable people—in this case, “pre-custody bisexuals”—
are kept safe, the issue of whether they are housed with gay men and trans women
or with other vulnerable detainees may not especially signify.
This brings us to the third objection to K6G’s terms of admission—that even
granting the legitimacy of the current standards, the Jail still routinely fails to
identify many people whose sexual practices or inclinations place them within the
category of people K6G is intended to house. The concern here is that some men
who would fit K6G’s admission criteria nonetheless answer in the negative when
asked in IRC if they “are homosexual.” Why would they do so? Such men may be
reluctant to “come out” to state officials in an environment in which being gay can
typically expose one to harassment and abuse.322 Or, although their primary sexual
men while in custody, come to find themselves having feelings of sexual attraction toward other men.
Nonetheless, it may be a wise strategic choice for Jail officials to refuse bisexuals access to K6G when this is the
sole asserted ground for their admission.
One possible troubling effect of excluding bisexuals from K6G is that it may disproportionately disadvantage
men of color. Russell Robinson reports that “black men and Latino men are more likely than white men to report
having had sex with both men and women.” Robinson, supra note 115, at 216. And, because bisexuals may also
face an elevated risk of sexual violation in prison, see supra note 312, it may be argued that a unit designed to
protect likely victims ought not to exclude bisexuals. For discussion on this issue, see text accompanying note
322.
321. See supra note 312.
322. Note that this concern may arise in the case of someone who knows nothing about K6G, but it need not,
since some men who know about K6G may still hesitate to identify as gay in any custodial setting. Because my
primary research focused on those who answered “yes” to the “homosexual” question in the IRC classification
process, I am unable to provide any sense of the motivating reasons of those who, although being otherwise
K6G-eligible, nonetheless answered this question in the negative. However, the fact that Latinos but not blacks
appear to be substantially underrepresented in K6G, see infra note 323, offers at least a starting point for
speculating as to possible motives. Here, it may be relevant that Southsiders, who make up the largest grouping of

70

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

orientation is toward other men, they may not self-identify as “gay” or “homosexual,” whether for cultural or other reasons.323
At base, the concern here is that K6G is underinclusive even as to its own
admissions standards. To adequately assess the weight of this objection, it is
worthwhile to distinguish between two groups: those individuals who because of
their own appearance and self-presentation would be able to pass as heterosexual
in GP, and those who would not.324 As to those who would be able successfully to
pass in GP, the fact of their exclusion from K6G is not of sufficient concern to
justify a change in policy. To imagine otherwise would require a view of K6G
admission as an entitlement owed to all those male detainees whose sexuality is
primarily oriented towards other men. Were this the case, not only would the
exclusion of successful GP-passers be grounds for condemnation of the program,
but it would also become L.A. County’s affirmative obligation to take all possible
Latinos in the Jail, are known to be especially intolerant of homosexuals. In the summer of 2007, for example, this
gang even put out a “green light” on all gay Southsiders, meaning that anyone in the gang could attack gay
members with impunity and even earn “stripes” for doing so. Under these conditions, a disproportionate reticence
among Latino detainees to admit to being gay in IRC should not be surprising.
323. Robinson notes that many men—especially black and Latino men—will not trust law enforcement with
this sensitive information. Robinson, supra note 115, at 209. Moreover, he reports that black and Latino men who
have sex with men (MSM) “are less likely to identify as gay than white MSM and more likely to reject
conventional sexual identity categories.” Id. at 215. Robinson’s discussion raises the possibility that whites will
be overrepresented in K6G at the expense of blacks and Latinos. Because the Jail population changes daily, any
statistics on this issue will at best be a snapshot of the facility. But data provided on two random days in June 2010
and March 2011 suggest a relatively stable racial distribution. And as it happens, K6G is disproportionately white.
In June 2010, for example, whites made up 29.3% of the K6G population as compared with 13.9% of the L.A.
County Jail as a whole. But, as it turns out, blacks too appear to be overrepresented in K6G, although not nearly to
the same degree as whites, making up 36.5% of K6G as compared with 33% of the Jail’s population. It is Latinos
who proved to be markedly underrepresented in K6G, making up 31.3% of K6G’s population and 49.7% of the
Jail population as a whole.
June 14, 2010

March 4, 2001

Total Jail Population

K6G

Total Jail Population

K6G

Black

33.0%

36.5%

33.0%

35.0%

Latino

49.7

31.3

49.6

29.7

White

13.9

29.3

13.7

30.1

Other

3.4

2.9

3.7

5.2

Personal Communication, Deputy Bart Lanni, L.A. Sheriff’s Department, June 14, 2010 & March 4, 2011 (on file
with the author). This discrepancy suggests that cultural factors relating to how Latino men in the Jail and perhaps
Latino culture as a whole conceive of sexual identity may well be influencing the way Latino detainees answer the
initial IRC inquiry. For a possible explanation of Latino underrepresentation in K6G arising from the particular
dynamics of the prison culture itself, see supra note 322.
324. As to the former—those able to pass—they may lack the characteristics that fit the stereotypical picture of
the gay man (effeminacy, flamboyance, etc.) or that would otherwise lead men in GP to perceive someone as gay.
Or, as was true of more than one individual I met in K6G, despite a preference for performing something of the
stereotypical gay identity, they may be wholly capable of suppressing any such behavior patterns and successfully
passing as heterosexual—and even as the toughest hard-core gangster.

2011]

STRATEGIC SEGREGATION

71

steps to identify all such individuals and house them in K6G. But this would be
absurd. For one thing, this approach would force people who have made the
decision to keep their sexuality private to reveal intimate details about their
personal lives—an enterprise even more unseemly than that of having state
officials pry into the private lives of those men who already readily volunteered
personal information about their own sexuality. And more to the point, it would do
so for no good reason. Although the idea behind K6G is to segregate sexual
minorities from GP, this is not its motivating purpose. There is no independent
value to identifying all detainees who meet K6G’s admission criteria and housing
them together. The point of the program is to create a space in which those
individuals most at risk of sexual assault behind bars can do their time free from
this danger. Men who, although satisfying the (inevitably overinclusive) admissions standards, “are masculine and physically strong,”325 or who are otherwise
able to self-present in ways consistent with the operative understanding of
heterosexuality in GP, quite simply do not face the same risk of victimization as
men who are less able to pass as straight. From a policy perspective, therefore, the
fact that the members of this first group may remain in GP is not of serious
concern.
There may be an irony here for those who would condemn the K6G program as
overly driven by stereotypes. As has been seen,326 K6G’s classification officers do
what they can to go beyond the obvious stereotypes of the gay man in order to
identify those men who, although not conforming to the stereotypes, are still
fundamentally sexually attracted to other men. But perhaps inevitably, in seeking
to identify those men who satisfy their working definition of homosexuality, Bell
and Lanni will have to rely to some extent on socially constructed and culturally
legible markers of sexual orientation. Some might regard this fact as proof of the
biased, insufficiently sophisticated, and even dangerous nature of the K6G
classification process when it comes to understanding sexual identity. However,
given the mission of K6G, it is far less important that Bell and Lanni effectively
identify those individuals who are “really gay”—even assuming this category has
any real meaning—then that they are able to identify those people who would
come to be identified as gay within the prison culture.327 To some extent, there will
be an overlap between those who demonstrate some of the cultural markers of gay
identity and those who would be judged as gay within the Jail’s GP. But this will

325. Robinson, supra note 115, at 239.
326. See supra Part II.C.2.
327. I am grateful to Noah Zatz for encouraging me to draw this distinction. It does, however, bear
emphasizing that this is not how Bell and Lanni understand what they are doing. Nor would they be fulfilling
K6G’s official mandate if instead of seeking to identify which inmates are “really” gay, they sought instead to
identify who would be regarded as such in GP. As acknowledged throughout this Article, a system designed to do
the latter may well be more normatively appealing than the present K6G admissions process. But that approach is
necessarily distinct from what L.A. County is legally obliged to do under the consent decree establishing the unit
now known as K6G. See Consent Decree, supra note 116.

72

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

not be a perfect overlap. Indeed, it would be a grave mistake for the K6G
classification process to rely only on the most overt cultural markers of gay
identity, since those predatory inmates determined to unearth the sexual orientation of fellow detainees will have more time and more opportunity to observe an
individual’s behavior at his most intimate moments, when his guard is down.328 An
effective classification process would thus have to go beyond stereotypes. As a
practical matter, however, the possibility that, at some level, the two-step K6G
classification process is tilted toward those who are on average more likely to be
perceived as gay by others and thus less able to pass as straight in GP, and against
those who are able to pass—and who may not even self-identify as gay—is simply
not a plausible ground for objection. To the contrary, it seems exactly right.
The more worrisome possibility concerns those who bypass the K6G classification process by answering “no” in IRC to the question are you homosexual? and
yet who are not able to pass as heterosexual on the mainline. It is impossible to
know how many people are in this position. But it seems reasonable to expect that
many are.329 This means that, notwithstanding the protection K6G offers some
detainees whose personal characteristics put them at risk of victimization in GP,
there continue to be people who are predictably vulnerable to such abuse who are
routinely housed in GP.
This fact is of serious concern. But—and here we come to the fourth and most
weighty objection to K6G’s classification process—even if every person who fits
K6G’s admissions criteria were somehow identified and sent to K6G, there would
still be people routinely placed in GP who are known to be vulnerable to sexual
abuse. Gay men and trans women are perhaps the most obvious and ready victims
of the prison culture of hypermasculinity, but they are by no means its only
victims. To the contrary, it is well recognized that a host of factors besides being
gay or trans increase the vulnerability of male prisoners. Most notably, these
factors also include “mental or physical disability, young age, slight build, first
incarceration in prison or jail, nonviolent history, prior convictions for sex offenses
against an adult or child . . . [and] prior sexual victimization . . . .”330 K6G is thus
underinclusive, not only as to those gay men who, although at risk in GP, fail to

328. See, e.g., PARSELL, supra note 60, at 78–79, 86–89 (recounting an experience shortly after his arrival in
prison when, having been manipulated by predatory inmates into feeling comfortable and letting down his guard,
he let slip a comment revealing his own status as gay—a comment particularly noted by his companions at the
time, who included his future rapist).
329. L.A. County has a total population of over 9.8 million, see http://www.google.com/search?sourceidϭ
navclient&ieϭUTF8&rlzϭ1T4ADBF_enUS331US332&qϭtotalϩpopulationϩlosϩangeles, and an estimated
LGB population of over 440,000, or approximately 4.5% of the total. See GATES, supra note 107. Yet K6G’s
population of approximately 350 is only about 2% of the total male detainee population in the Jail. These numbers
suggest—albeit roughly—that K6G is not capturing the whole eligible detainee population in LA County.
330. COMMISSION REPORT, supra note 8, at 75.

2011]

STRATEGIC SEGREGATION

73

self-identify in IRC, but also as to all detainees with these other risk factors.331
True, L.A. County does have a policy designed to protect otherwise vulnerable
inmates who do not qualify for K6G. This policy involves identifying and
separating from GP those prisoners, unfortunately known in official Jail parlance
as “softs,” who because of physical attributes or other factors are also likely to be
victimized in GP. But this policy falls well short of meaningful implementation. In
the summer of 2007, the “softs” unit consisted of just five individuals332 out of a
total male prisoner population of approximately 17,000,333 a number which plainly
indicates that little institutional attention was being devoted to identifying those
at-risk prisoners who, although not appropriate for K6G, nonetheless needed to be
housed separately from GP for their own protection.
Part of the problem may be the design of the policy itself. At some point, it was
explained to me that “softs” were identified during the initial classification process,
by intake officers who “eyeball” new admits and assess their likely ability to
handle themselves on the mainline.334 But to anyone observing the intake process,
the flaw in this strategy becomes immediately obvious: classification officers sit in
raised booths and speak to incoming inmates through a glass wall with the aid of a
speakerphone. This set-up makes it difficult for officers to get a good look at the
people to whom they are speaking, even if they were bent on doing so. In any case,
during the intake process (which can take as little as one minute), the officer is
focused on the computer screen into which data is being entered. There is thus little
opportunity for the officer even to form an impression of the physical appearance
of each incoming individual, much less to make a judgment about likely victimization. If the Jail’s claim to protect at-risk inmates not otherwise eligible for K6G is
to be credible, at a minimum, a new mechanism for identifying otherwise at-risk
detainees needs to be developed, one that does not place the burden on intake
officers who are not in a position to fulfill it.
One strategy for addressing K6G’s underinclusivity would thus be to shore up
the “softs” program to assure protection for those at-risk individuals who do not

331. Some members of this latter group do find their way into K6G, although admittedly on the pretense of
actually being gay. It is unclear whether this occurs with the tacit assent of the classification officers or in spite of
their best efforts to discover the truth. I myself witnessed several instances in which, faced with such a case, the
K6G classification officers denied admittance to K6G but then did their best to find another relatively safe place in
the Jail for the person to land, whether in protective custody, or in the “soft tank,” or, if the person had any serious
mental illness (as do many of the detainees in L.A. County), in Twin Towers. But, although laudable, these
measures will seem beside the point to those who think a segregation unit for potential victims of prison rape
should accommodate all likely victims, and not merely those who happen to be gay or trans or able to pass as such,
or who are fortunate enough to encounter a sympathetic officer.
332. These men were housed in a pod in Twin Towers with a handful of detainees who were developmentally
disabled.
333. L.A. County houses approximately 1900 women out of a total inmate population of 19,000. See supra
note 318.
334. See also IRC Classification JICS Security Level Assignment (undated form on file with the author)
(directing intake officers to “observe inmate” to decide whether he is “soft”).

74

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

qualify for K6G. This approach would amount to a two-track strategy for
protecting vulnerable detainees: one track (K6G) for trans women and those men
who satisfy the Jail’s conception of homosexuality, and another track (“softs”) for
those who exhibit other risk factors for sexual victimization.
This brings us to the heart of the matter from a policy perspective. There is no
question that all people known to be at risk of sexual abuse in custody must be
protected.335 And K6G teaches that systematically separating vulnerable people
from GP is a relatively effective way336 to ensure their safety.337 The question is
whether carceral institutions would do better to create a single segregation unit to
house all those individuals known to be highly vulnerable to sexual victimization—
call this the unified approach—or to maintain two separate units (two “tracks”),
one on the model of K6G and another for all other vulnerable people.
In its final proposed standards, submitted to the U.S. Attorney General in
June 2009, the National Prison Rape Elimination Commission338 adopted in no
uncertain terms the first, unified approach. The Commission’s standards provided
that all inmates be screened on arrival (and at “all subsequent classification
reviews”) to assess their risk of being victimized or a victimizer,339 and included
the full range of those features known to increase the vulnerability of male
prisoners340 among the criteria to be considered when “screen[ing] male inmates
for risk of victimization.”341 Although it recognized the particular vulnerability of
gay men and trans women,342 the Commission nonetheless “discourage[d] the
creation of specialized units for vulnerable groups,” and “specifically prohibit[ed]
335. This is both a moral obligation on the part of the state and, after the adoption by the United States
Attorney General of National Standards to Prevent, Detect and Respond to Prison Rape, will also be a legal one.
See National Standards, supra note 7.
336. Given the realities of incarceration in the United States today—including, most notably, chronic
overcrowding and understaffing as well as an inmate culture in which predation and the readiness to use violence
brings the greatest status and respect—relative safety may at present be the best that can be hoped for.
337. There remains, of course, the question of the implications of such segregation for the people who remain
in GP. I take up this question at the end of this section.
338. See supra note 37.
339. COMMISSION REPORT, supra note 8, at 217.
340. See supra note 8 (listing the Commission’s screening criteria for male inmates).
341. SC-1 listed separate screening criteria for female inmates. In contrast to the long list provided in the case
of male inmates, see supra note 8, the Commission included just two criteria to consider in “screen[ing] female
inmates for risk of sexual victimization: prior sexual victimization and the inmate’s own perception of
vulnerability.” Id. The proposed standards issued in January 2011 by the U.S. Department of Justice do not
distinguish between men and women as to screening criteria for risk of victimization and abusiveness. See
National Standards, supra note 7, at 6280 (§ 115.41).
342. In its final report issued in June 2009, the Commission acknowledged the particular vulnerability to
prison rape of both gay men and individuals “whose sex at birth and current gender identity do not correspond
([i.e.,] transgender or intersex).” COMMISSION REPORT, supra note 8, at 73 (“Research on sexual abuse in
correctional facilities consistently documents the vulnerability of men and women with non-heterosexual
orientations (gay, lesbian, or bisexual).”). The Commission also condemned the practice of holding “vulnerable
prisoners” in conditions of protective custody, which as it noted, “may be as restrictive as those imposed to punish
prisoners.” Id. at 79. The Commission’s draft standards therefore allowed the use of protective custody of
“victims or potential victims . . . only as a last resort,” and then “only on a short-term basis.” Id. at 80.

2011]

STRATEGIC SEGREGATION

75

housing assignments based solely on a person’s sexual orientation, gender identity,
or genital status . . . .”343
Although the Commission said little to explain this latter prohibition—which,
read broadly, could have put K6G itself in jeopardy344—it did condemn as
“demoralizing and dangerous” the practice of officially labeling some detainees as
members of sexual minorities. As has been seen, this concern is not unwarranted,
and thus a unified approach may well be the better one for jurisdictions seeking
ways to protect their most vulnerable prisoners. A unified approach would
arguably reduce—although not eliminate345—the need for state officials to delve
into the private sexual lives of people newly admitted to custody, for whom the fact
of their incarceration alone would no doubt be traumatic enough. It would guard
against the unpalatable and risky practice of officially labeling some people as
members of sexual minorities. And it may even yield an administrative benefit by
avoiding unnecessary duplication of services, as well as simplifying the challenge
for prison administrators, who with a larger population needing segregation from
GP might enjoy greater flexibility in terms of housing and staffing. For these
reasons alone, a unified strategy would seem in most cases to be the better
approach.
Still, a unified strategy would not be wholly free of the dangers posed by K6G.
First, although there would be less need for officials to pry into the details of
detainees’ private lives in order to determine an individual’s sexual orientation,
there will still be occasion when such intrusion is necessary. A segregationist
343. Id. (emphasis added). See also id. at 217 (stipulating at SC-2 that “[l]esbian, gay, bisexual, transgender, or
other gender-nonconforming inmates are not placed in particular facilities, units, or wings solely on the basis of
their sexual orientation, genital status, or gender identity”). But see infra note 344.
344. On January 24, 2011, the United States Department of Justice (DOJ) issued a Notice of Proposed
Rulemaking (NPRM), in which it presented for comment proposed National Standards to Prevent, Detect, and
Respond to Prison Rape. See National Standards, supra note 7. Prior to its review of the Commission’s final
standards, DOJ had issued an Advance Notice of Proposed Rulemaking (ANPRM) “to solicit public input on the
Commission’s proposed national standands.” See id. at 6248. Partly in response to comments on SC-2, see supra
note 343, submitted pursuant to that ANPRM (including my own, see National Standards, supra note 7, at 6257
(noting that “[o]ne commenter discussed the success of the Los Angeles County Jail in housing gay male and
transgender prisoners in a separate housing unit”)), DOJ eliminated any prohibition at all “on assigning inmates to
particular units solely on the basis of sexual orientation or gender identity.” See id. at 6257. In responding to the
January 24, 2011 NPRM, many commenters, myself included, argued that this response went too far, and that a
better approach would be a qualified prohibition allowing that programs like K6G would still be possible so long
as established in connection with a consent decree, legal settlement or legal judgment for the purpose of
protecting lesbian, gay, bisexual, transgender or other intersex inmates. See Protecting Lesbian, Gay, Bisexual,
Transgender, Intersex and Gender Nonconforming People from Sexual Abuse and Harassment in Correctional
Settings, Comments Submitted in Response to Docket No. OAG-131; AG Order No. 3244-2011, National
Standards to Prevent, Detect, and Respond to Prison Rape, April 4, 2011 (comments submitted collectively by the
National Center for Transgender Equality, the National Center for Lesbian Rights, the ACLU, the National
Juvenile Defender Center, the Sylvia Rivera Law Project, The Equity Project, Lambda Legal Education and
Defense Fund, and the Transgender Law Center); Sharon Dolovich, Comments on National Standards to Prevent,
Detect, and Respond to Prison Rape, submitted pursuant to Notice of Proposed Rulemaking, January 24, 2011 (on
file with the author).
345. See infra, text accompanying notes 347–48.

76

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

strategy will only work to protect vulnerable prisoners if access to the protected
unit is restricted to those judged to be at risk. Otherwise, as with L.A. County’s
pre-1985 efforts to segregate gay men for their own protection, anyone wanting
access to likely victims need only assert his own feelings of vulnerability.346 And,
although many likely victims may be classified on appearance or mannerisms
alone,347 there will still be instances where further inquiry is required to assess
claims of vulnerability—including claims made on the basis of sexual orientation.348 As with K6G, individuals wishing to bypass this process may simply opt to
keep silent. And to the extent that they are able to pass as straight in GP, their
decision to do so may pose no issues for the project of protecting at-risk prisoners.
But there will inevitably be some people who fear themselves unable 24/7 to keep
hidden their sexual orientation toward other men, and who would thus be at risk
notwithstanding their lack of any immediately obvious characteristics suggesting
this orientation. The need to protect this population while at the same time
maintaining some gate-keeping process means that there may continue to be some
measure of official intrusion into the private lives of men in custody with the aim
of assessing claims of sexual-minority status.
Second, although with a unified approach, the state will avoid overtly labeling
some detainees as sexual minorities, it will still be flagging some subset of
prisoners as likely victims, thus potentially reproducing the sitting duck objection
raised against K6G. Possible predators—whether inmates or staff—may yet be
inclined to seek out and victimize individuals in this population, whether because
some among them are known to be gay or trans or simply because, by seeking
protection, they are by definition exposing themselves as weak. Equally, as with
K6G, the fact of their segregation may make residents of a protected unit
vulnerable to the kind of systematic verbal harassment from both staff and fellow
inmates that K6Gs report today.
Other dangers also attend a unified approach. For example, by explicitly
combining gay men and trans women with other vulnerable populations, there is a
risk of reintroducing into a combined unit some of the anxiety about possibly being
thought to be gay that already exists in today’s GP. Call this the identity insecurity
effect. This risk may well be mitigated by a collective recognition among residents
in the unit of the way the pressures to perform a hypermasculine identity create
stressful and even dangerous conditions for everyone, and the desire of all
residents of a unified dorm to avoid these effects. But whether this collective
recognition would emerge, and result in dorms free from such pressure, would

346. See also infra, note 392 (explaining the problems with the unit for gay male detainees at Rikers Island,
which arose from the minimal screening procedures for admission and led to the mixing of vulnerable and
predatory inmates).
347. See, e.g., Man and Cronan, supra note 8, at 164–75 (identifying a range of factors known to correlate with
vulnerability in men’s prisons, including age, projecting feminine characteristics, and physical size).
348. I thank Carole Goldberg for this observation.

2011]

STRATEGIC SEGREGATION

77

depend on the institution and on the makeup of the unit,349 as well as the extent to
which it remains free of gang control. This means that officials would still need to
be mindful of the danger of a possible reprise of the very cultural dynamics that
made the segregation of vulnerable people necessary in the first place.
Equally, officials adopting a unified strategy would still need to develop
mechanisms for dealing with one group that K6G overtly excludes: men who have
sex with women on the outside and sex with men while in custody. Some of these
men may themselves have been victimized and thus urgently require protection.350
But others will have adopted the dominant role in a protective pairing. Men in this
latter group will thus not only not need protection, but the argument for their
admittance—that they have sex with men—assumes a view of their sexual
experiences that contradicts the construction put on them in the prison culture in
which they occur.351 These men, in other words, may not even have regarded
themselves as having had sex with men, and, indeed, may be inclined to spend
considerable energy ensuring that no one doubts their status as “real men.” To put
members of this group in a protected unit for possible victims is particularly likely
to generate the identity insecurity effect noted above and thus to put at serious risk
the very people this segregationist strategy is intended to protect.352
Officials adopting a unified strategy would also need to be mindful of the danger
of diluting the protective enterprise. The worry here is that creating a catchall
classification for at-risk prisoners could ultimately distract official attention from
the need to protect against prison rape, leading instead to the creation of a large
multipurpose unit that houses anyone who, for whatever reason, is unable to “make
it” on the mainline. This concern is not wholly abstract. Some years ago, the
California prison system created what are known as “sensitive needs yards.” These
units, which in size and functioning are no different than GP units, house a host of
likely victims of prison violence: gang dropouts, sex offenders, prison informants

349. For example, the population of a unified unit might include a large number of elderly men, first offenders,
and nonviolent offenders, as well as individuals who are physically or mentally disabled in some way. Members
of these groups may be less likely than others to be drawn into the culture of hypermasculinity. I thank Jody
Marksamer for this point.
350. See Schwenk v. Hartford, 204 F.3d 1187, 1203 n.14 (9th Cir. 2000) (“Once raped, an inmate is marked as
a victim and is subsequently vulnerable to repeated violation.”).
351. In that context, those who take the dominant sexual position (whom Wooden and Parker memorably label
“inserter[s]”), are defined as heterosexual men whose passive sexual partners (Wooden and Parker’s “insertee[s]”)
are viewed as women. See WOODEN & PARKER, supra note 91, at 15.
352. It bears noting that many of the men I have described here as possible predators—indeed, possibly all of
them—are themselves afraid of being victimized. See Haney, supra note 49, at 128 (explaining the dynamics that
lead some men in custody to become predators out of fear for their own safety). Moreover, as Suk rightly notes,
removing the most vulnerable prisoners from GP may simply result in the victimization of the most vulnerable
individuals who remain, see Suk, supra note 67, at 114—at least absent meaningful steps to counter the
hypermasculinity imperative. These two aspects of the reality of prison life powerfully indicate the necessary
limits of the segregationist strategy, and underscore the broader imperative that, if the state is to continue to
incarcerate, ways must be found to ensure safe and humane custodial conditions for all the people consigned to
live in prison.

78

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

(i.e., “snitches”), and anyone else who requests protective custody, but for whom a
single cell is found by officials not to be appropriate or necessary (or available). It
is unclear whether these yards are safer than GP. But experience with this policy
has taught that sensitive needs yards are easily accessed by anyone with a story to
tell. If a “shot caller” in GP wants a snitch dispatched, he can order one of his
“soldiers” to pretend to be a gang dropout to get access to the yard with the target.
The soldier can bide his time, and when the opportunity arises, assault the target.
The worry, in other words, is that housing multiple populations with different risk
factors together in one unit can expose some vulnerable prisoners to assault.
Officials committed to creating a truly safe space for likely victims of sexual
violence would need to take steps to ensure that access really is limited to those for
whom it is intended. In particular, officials facing multiple, somewhat compatible
policy goals (i.e., separating out certain at-risk populations for different reasons)
would need to guard against the temptation to roll them all together the way the
California prison system has done. Having smaller units with more narrowly
defined missions may provide some defense against the pressure to consolidate,
and thus against the dangers such consolidation may generate.
Still, even if a unified segregationist approach is not wholly free of the dangers
to which K6G gives rise, it nonetheless promises to reduce those dangers
considerably. In most cases, therefore, officials looking to develop a protectionist
strategy may be well-advised to adopt such an approach. It is worth noting,
however, that a unified approach would not come without a cost. In particular, the
resulting unit would be unlikely to feature a carceral climate in which, as in K6G,
detainees feel free to express their emotions and to openly forge mutually
supportive friendships and even loving relationships.353 The fact that such a
climate already exists in K6G—and, most significantly for policy purposes, may
well be an antidote to the hypermasculinity imperative and the fear and violence it
creates—is something of which those committed to increasing the humanity of the
carceral environment should not lose sight.
What of L.A. County itself? Notwithstanding the dangers associated with the
K6G approach, it would be a serious mistake to dismantle a program widely
acknowledged as a success. K6G has broad community support and has ensured
relatively safe and humane conditions for populations that are otherwise extremely
vulnerable. As a purely theoretical matter, this position may seem indefensible; if
state-sponsored, identity-based segregation is wrong, surely L.A. County must
adjust, especially when a unified approach could mitigate the harms of such
segregation while continuing to ensure the safety of K6G’s current residents. The
argument against such a change, however, is not theoretical but pragmatic, of the
“if it ain’t (that) broke, don’t fix it” variety. Prisons are complex institutions that
can be very difficult to manage. Programs succeed or fail for all kinds of reasons

353. I describe this aspect of life in K6G in more detail elsewhere. See Dolovich, supra note 45, at 12–14.

2011]

STRATEGIC SEGREGATION

79

particular to a given institution, reasons that may be entirely independent of the
theoretical wisdom of the program design.354 To dismantle a relatively successful
program in the hope of improving on what already exists could wind up leaving
everyone worse off. And if for policymakers, the consequence would be professional disappointment or perhaps wounded pride, the effect on the prisoners could
be unspeakable harm. In my view, this is not a risk worth taking.355
The challenge for L.A. County going forward is to create an equally safe and
successful program for those detainees who do not qualify for K6G but who are
still at risk of sexual victimization. In facing this challenge, L.A. County may not
be alone. Although for the reasons just canvassed, officials seeking strategies for
protecting vulnerable prisoners will in most cases be well advised to adopt what I
have called a unified approach, there may yet be other facilities in which, for
institutional reasons, a two-track strategy would be the better approach. One such
context may be male juvenile detention centers, in which the imperative among
detainees to prove themselves “real men” (i.e., not gay) is likely to be extreme,
raising the possibility that gay and trans youth might be at risk of victimization
even in a unit filled with otherwise vulnerable individuals.356
K6G’s success in L.A. County is arguably reason enough for this approach to
remain an available tool in the toolkit of prison administrators. That other
institutions may find this approach the best way to ensure protection of their most
vulnerable prisoners only further strengthens the argument against its foreclosure.
As has been seen, segregating gay and trans prisoners even for their own protection
is not without its hazards. But in the existing penal context, few policies are
risk-free. So long as American society continues on its current carceral path, the
best that can be hoped for is some appreciable reduction in violence. And given
what such a reduction would mean for those who would escape victimization as a
result, pursuing half-measures may be the most immediate moral imperative.
Plainly, the foregoing assumes that a segregationist strategy—whether unified
or two-track—will lead to an overall reduction in sexual victimization. There is,
however, a troubling alternative possibility, one raised by Jeannie Suk in her
thoughtful commentary on this Article.357 Suk’s comments suggest that, assuming
the accuracy of the account of the prison sexual culture offered in Part I,
354. For a compelling argument as to the significance of institutional context to the success of rehabilitative
programming in custodial settings, see generally ANN CHIH LIN, REFORM IN THE MAKING: THE IMPLEMENTATION OF
SOCIAL POLICY IN PRISON (2000).
355. If K6G should remain in place, it is still true that L.A. County must do better at protecting other
vulnerable prisoners. And it is also true that the pragmatic path-dependency concern just offered against
dismantling K6G does not apply to jurisdictions facing the challenge of newly establishing programs to protect
their most vulnerable prisoners. In the larger context, for the reasons canvassed above, the more inclusive
approach is arguably the preferable default option.
356. But see National Standards, supra note 7, at 6298 (§ 115.342 (d)) (proposing a prohibition in juvenile
facilities of placing “[l]esbian, gay, bisexual, transgender or intersex residents . . . in particular housing, bed, or
other assignments solely on the basis of such identification or status.”).
357. See Suk, supra note 67.

80

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

segregating likely victims will make no difference to the total amount of forced or
coerced sex in a given facility.
How can this be? As Part I argued, the all-male environment of the prison
creates anxiety among residents that they are insufficiently “masculine.” This
anxiety is only exacerbated by prisoners’ lack of access to healthy, socially
productive means to express (and perform) their gender identity. In the absence of
alternatives, the predatory domination of weaker inmates becomes the main
mechanism prisoners use to gain status and power. This predation, moreover,
becomes a means of self-protection, since by dominating weaker inmates, predatory prisoners demonstrate that they are “real men” and thus not someone to be
messed with. This dynamic, referred to as the hypermasculinity imperative, has a
gendered aspect, with the weakest inmates being redefined as female and thereby
rendered eligible for victimization.
If this account is accurate, it suggests that the removal of vulnerable inmates,
rather than reducing the amount of victimization overall, will only “redistribute” it
to others.358 As Suk puts it, “[i]f gay men and trans women are not present to be
forced to become ‘women,’ other men will be forcibly classified as female.”359
Something of this understandable concern implicitly motivates the recognition,
emphasized throughout this Article, that it is not enough to segregate gay men and
trans women; instead, all vulnerable prisoners—the young, the mentally or
physically disabled, the small in stature, the elderly, etc—should be housed
separately from GP. But as Suk observes, even this move could be inadequate,
since on the logic of the hypermasculinity imperative, those prisoners who remain
behind in GP will continue to need to dominate others in order to prove their own
strength and ensure their own protection from victimization. If, as I have argued,
gender in this culture is ascribed, then once the most obvious targets for “feminization” are removed, other candidates for the female role will have to be identified.
The strongest of the bunch will simply turn on (and “turn out”) the weakest
prisoners who are left.
This insight offers a chilling reminder that the problem of sexual victimization
behind bars will not be fully resolved unless the carceral experience can be
reengineered in a more humane direction, not just for the especially vulnerable, but
for everyone. To understand the source of the danger, moreover, is to recognize the
most obvious avenue for reform: the provision of more personally healthy and
socially constructive ways for all men in custody to assure themselves—and
others—of their own manhood.360
Meanwhile, and perhaps on a less utopian note, Suk’s concern points to the need
for an even more finely grained classification process than that proposed by the

358. See id. at 114.
359. Id. at 113.
360. I leave for another day investigation of what these alternatives might be.

2011]

STRATEGIC SEGREGATION

81

Commission.361 That is, it may not be enough to divide the population into two
groups (i.e., those with characteristics known to make them vulnerable and those
without), but instead to house people in smaller groupings with others of like size
and strength, and to be prepared to reshuffle housing assignments should any
residents emerge as victims or victimizers. This approach is something like what is
already being done in San Francisco County.362 To facilitate this method, classification officers in San Francisco can spend up to 45 minutes interviewing each new
admit, to determine not only whether he might be vulnerable or predatory, but also
his relative strength and where he might fall in the pecking order of the unit to
which he is assigned (i.e., might he emerge as a victim in a standard GP unit? a
predator in a unit of vulnerable inmates?).363 The dedication of this time and
attention to each incoming detainee in San Francisco is in contrast with L.A.
County, in which the initial classification interview can be as short as one minute.
San Francisco’s more highly calibrated approach is enabled both by its relatively
small population—in recent years its annual admissions rate has been between
30,000 and 37,000 annually,364 as compared with L.A. County’s 166,000365—and
by the modern podular design of its new facility,366 which allows for smaller and
more readily monitored housing units.367 Unfortunately, however desirable as a
policy matter, this more fine-grained approach is highly labor-intensive. And in an
era of prison overcrowding and understaffing, the conditions simply do not exist
for its immediate implementation more widely—although jurisdictions able to
follow San Francisco’s lead ought certainly to do so.368
Given the present practical limits on the classification process, there remains the
question of how to respond to Suk’s redistribution point in the shorter term. If
removing from GP people known to be vulnerable will simply shift the burden to
those who remain and leave the overall victimization rate the same, we could be
left, as Suk suggests, making unpalatable normative arguments as to which

361. See supra note 40.
362. Personal Communication, Undersherriff Jan Dempsey, San Francisco Sheriff’s Department, May 29,
2008 (on file with the author).
363. See id.
364. Personal Communication, Sgt. Joan Scannell, San Francisco Sheriff’s Department, May 9, 2011 (on file
with the author).
365. See supra note 101.
366. See supra note 362.
367. L.A. County’s Twin Towers facility is of this same design, but understaffing has meant that much of Twin
Towers has stood empty almost since it opened.
368. In order to adopt San Francisco’s approach, moreover, it is of considerable benefit to have a physical plant
of the podular structure of San Francisco’s new jail. Unfortunately, although such modern buildings are far better
both in terms of safety and living standards for detainees, they are simply not the norm around the country. The
reason is simple: they are expensive, and both continued prison overcrowding and budget woes nationwide mean
that even those jurisdictions sorely in need of new facilities are being forced to rely on outdated housing stock
with old-style cellblocks that most corrections professionals have long since condemned for, among other things,
excessive noise, obstructed sightlines, and demoralizing effects on the people who live in them—not to mention
their incompatibility with a strategy of smaller, more tightly controlled housing units.

82

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

prisoners most deserve protection.369 Absent such arguments, moreover, there
would be no justification for any segregationist strategy at all. But to justify
abandoning a segregationist approach, it would have to be the case that segregation
would make no difference at all to the overall victimization rate. If instead, the
segregation of likely victims would put at least some downward pressure on the
incidence of abuse, even if not squelching the danger entirely, segregation would
still be justified on the ground that, all things considered, it is better to have fewer
victims. And although any such assessments in this regard may be only speculative, my strong sense—and, presumably, that of the Commission, which studied
the issue and recommended that inmate classification systems take account of
relative vulnerability—is that removing the most obvious and ready victims will
result in an overall reduction in sexual victimization, even if not its wholesale
elimination. Victimizing another person is costly; it requires a readiness to fight
and the capacity to overcome the strong resistance of the target (not to mention a
willingness to treat another person cruelly). Removing the easiest prey and leaving
the remaining prisoners more evenly matched physically increases the cost of
predation, making it harder to accomplish. It stands to reason that as the cost of
predation goes up, fewer prisoners will be able and willing to make the attempt,
thereby keeping the threat of assault more in check. Furthermore, assuming that
those units housing otherwise vulnerable prisoners can be kept free of a resurgence
of the hypermasculinity imperative, there will simply be fewer people left to
engage in the cycle of predation.
None of this is to minimize the danger faced by men in any housing units who
become the default victims once the more obviously weaker inmates are removed.
To the contrary, the possibility of their victimization demands that all available
protective measures be implemented. But as I have emphasized throughout this
Article, in the universe of American incarceration, we are very far from the
optimal. Although ideally, no one would be raped in prison, in reality it happens
every day. Failing a magic bullet, the best that can be hoped for in the short term is
an improvement over what currently exists. And unless one concludes that the
strategic segregation of all identifiable victims will make no appreciable difference—a position that on its face seems difficult to defend—support for the
segregationist approach proposed in this Article seems the only tenable position.
C. Unconstitutional: The Equal Protection Objection
I have argued that, notwithstanding the dangers associated with the K6G
approach, it would be a serious mistake to dismantle a program widely acknowledged to be a success. In addition, I have argued that the K6G model should be
available in cases where prison administrators determine this approach to be the

369. See Suk, supra note 67, at 115.

2011]

STRATEGIC SEGREGATION

83

best way to meet the particular challenges their institutions present.370 To adequately defend this position, it remains to address one final objection—that
identity-based segregation of the sort K6G represents violates Equal Protection,
and thus that no institution may constitutionally follow the K6G model, however
convinced officials might be of its desirability.
At first, the Supreme Court case of Johnson v. California371 may seem to
support this assertion. In Johnson, the Supreme Court heard a challenge to the
official policy of racial segregation then in place in the California prison system’s
reception centers. As is common practice, California designates certain prisons as
reception centers to which new admits are sent to be classified.372 In these
facilities, prisoners may spend up to sixty days having evaluations that will
“determine their ultimate placement” in the system.373 In 1995, Garrison Johnson,
a longtime prisoner in the California system, filed an equal protection claim
alleging that the California Department of Corrections (CDC) was segregating
prisoners in its reception centers on the basis of race.374 This allegation was
confirmed by a CDC officer, who admitted under oath “that the chances of an
inmate being assigned a cellmate of another race [in the reception centers] are
‘pretty close’ to zero percent.”375
Johnson ultimately prevailed in the Supreme Court. But the Court’s holding in
no way requires the conclusion that K6G is unconstitutional. The question before
the Court in Johnson involved the appropriate standard of review. Johnson argued
that, because the classification at issue was race-based, the Court should apply
strict scrutiny. The CDC argued that the appropriate standard of review was the
lower standard for reviewing prisoners’ constitutional claims established in Turner
v. Safley.376
Turner is a highly deferential standard that gives prison officials considerable
leeway in designing institutional policy.377 In this respect, it fits squarely within a
370. Ideally, this decision would be made in conjunction with community advocates, to ensure the community
involvement and support that has proved so crucial to K6G’s success. See infra, Part IV.
371. 543 U.S. 499 (2005).
372. All prisoners being transferred between facilities also go through the reception centers. Id. at 502.
373. Id.
374. Id. at 504.
375. Id. at 502.
376. 482 U.S. 78 (1987). In Turner, the Court held that even when a regulation or practice “impinges on
inmates’ constitutional rights,” the practice is nonetheless valid “if it is reasonably related to legitimate
penological interests.” Id. at 89.
377. This deference is by design. As the Turner Court explained,
[i]n our view, such a standard is necessary if prison administrators, and not the courts, are to make
the difficult judgments concerning institutional operations. Subjecting the day-to-day judgments
of prison officials to an inflexible strict scrutiny analysis would seriously hamper their ability to
anticipate security problems and to adopt innovative solutions to the intractable problems of prison
administration. The rule would also distort the decisionmaking process, for every administrative
judgment would be subject to the possibility that some court somewhere would conclude that it
had a less restrictive way of solving the problem at hand. Courts inevitably would become the

84

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

long constitutional tradition of according deference to prison officials.378 But in
Johnson, the Court sided against the state, holding that strict scrutiny was the
correct standard for reviewing Johnson’s claim of race-based classification. In that
case, the Court in effect decided that race trumps prison—that the constitutional
protection against undue racial classification is sufficiently weighty to overcome
the strong presumption of constitutionality generally accorded official actions that
“impinge on prisoners’ constitutional rights.”379 However, the Johnson Court did
not find the procedure of racial segregation in California’s reception centers
unconstitutional. It simply held that if the state wished to continue engaging in the
challenged practice, it had to show this practice to be consistent with strict
scrutiny, i.e., that the racial segregation was a “narrowly tailored” means to
“further compelling governmental interests.”380 All along, California had defended the policy challenged in Johnson as “necessary to prevent violence caused
by racial gangs.”381 Had the state been able to prove this assertion and also show
that the measure was narrowly tailored to the harm, it could have ultimately
prevailed.382

primary arbiters of what constitutes the best solution to every administrative problem, thereby
unnecessarily perpetuating the involvement of the federal courts in affairs of prison administration.
Id. (internal citations omitted).
378. Indeed, judicial deference to prison officials is perhaps the strongest theme to emerge from a historical
survey of prisoners’ rights litigation in the federal courts. For elaboration on this point, see Dolovich, supra note
36, at 962 n.306.
379. Turner, 482 U.S. at 89. In fairness to the Johnson Court, there was prior case law that “applied a
heightened standard of review in evaluating racial segregation in prisons,” to which it could (and did) point to
support its claim that it was only following precedent. Johnson v. California, 543 U.S. 499, 506–07 (2005) (citing
Lee v. Washington, 390 U.S. 333 (1968)). In Lee v. Washington, the Court had rejected an appeal challenging a
district court order striking down on Fourteenth Amendment grounds an Alabama statute that allowed for racial
segregation in the state’s prisons and jails. See Lee, 390 U.S. at 333–34. But the Court’s decision in Lee was issued
in a one paragraph per curiam opinion, and although this paragraph is open to the reading offered by the Johnson
majority (that it granted the state limited leeway to racially segregate “for the necessities of prison security and
discipline”), that decision by no means demands such a reading. Johnson, 543 U.S. at 539 (quoting Lee, 390 U.S.
at 334). In Lee, the Court simply rejected the State’s “contention . . . that the specific order directing desegregation of prisons and jails makes no allowance for the necessities of prison security and discipline,” noting that it did
“not so read the [order] of the District Court.” 390 U.S. at 333–34.
380. Johnson, 543 U.S. at 505.
381. Id. at 502.
382. As it happened, California chose to settle the case rather than seeking to defend the policy under strict
scrutiny at trial. This was a wise move, since at trial, Johnson could have produced evidence that California was
the only state in the union that racially segregated its reception centers, a finding calling into question the claim
that the move was necessary to avoid serious violence. As Justice Stevens noted in his Johnson dissent,
California’s policy is an outlier when compared to nationwide practice. The Federal Bureau of
Prisons administers 104 institutions; no similar policy is applied in any of them. Countless state
penal institutions are operated without such a policy. An amici brief filed by six former state
corrections officials with an aggregate of over 120 years of experience managing prison systems in
Wisconsin, Georgia, Oklahoma, Kansas, Alaska, and Washington makes clear that a blanket policy
of even temporary segregation runs counter to the great weight of professional opinion on sound
prison management.

2011]

STRATEGIC SEGREGATION

85

There are strong reasons to think that in Johnson, California could not have met
this burden.383 But it is far less clear that the same may be said of L.A. County as
regards K6G. As has been seen, in the culture of hypermasculinity that prevails in
the Jail’s GP, gay men and trans women are at constant risk of harassment and
assault. Reducing this risk and keeping members of these at-risk groups safe are
plainly compelling state interests under the first prong of the strict scrutiny test. As
for the second prong, narrow tailoring, it is certainly arguable that a different
strategy—the single unified approach explored in Part III.B above—would ensure
the protection K6G offers while minimizing the kind of overt identity-based
housing decisions that troubled the Johnson Court.384 But there are also reasons to
fear that opening up K6G to other populations could lead to the resurgence of a
hypermasculinity imperative among residents—whatever their sexual orientation—
afraid of being labeled as gay, thereby negating the protective benefits of K6G.385

Id. at 519 (Stevens, J., dissenting).
Yes, California might have tried to respond, as Justice Thomas argued in dissent, that California’s racially
aligned prison gangs are unlike those in any other prison system, making the need for racial segregation unique to
that state. But the credibility of this claim would have been sorely tested by the evidence from Texas, where
despite a prison culture as racialized as California’s, prison officials have long since abandoned officially enforced
racial segregation. Even more damning for California’s position, in 1977, in Lamar v. Coffield, the Texas
Department of Corrections entered into a consent decree under which “administrators agreed to [racially]
integrate inmates in double cells to the maximum feasible extent . . . .” See Chad Trulson & James W. Marquart,
The Caged Melting Pot: Toward an Understanding of the Consequences of Desegregation in Prison, 36 LAW &
SOC’Y REV. 743, 753 (2002) (citing Lamar v. Coffield, 951 F. Supp. 629 (S.D. Tex. 1996)) (emphasis added).
Compliance took fourteen years, but in 1991, after a decade of “footdragging,” Texas prison officials finally
implemented the order. See id. at 754, 755–56. And, despite predictions that “any efforts to racially mix the cells
would lead to catastrophic results—a violent race war,” see TRULSON & MARQUART, supra note 33, at 13, after the
integration of all prisoners found able to be safely housed in a cell with someone of another race, the race of the
parties was found to make virtually no difference to levels of violence between cellmates. Id. at 183. In fact,
analyzing data of violent incidents in the Texas prisons between 1991 and 1999, Trulson and Marquart found that
“the rate of cell incidents after 1992 was always less among desegregated cellmates.” See id. at 182. Trulson and
Marquart locate a possible explanation in what they call the “equal status contact” hypothesis. On this theory,
when individuals who are members of antagonistic groups interact together on terms of equal status, they are
more likely to see one another as independent human beings and less through a lens of group membership—and
are thus more likely to respect and judge each other as individuals. See Trulson & Marquart, The Caged Melting
Pot, supra at 745, 770. How might this thesis apply in the prison context? To radically oversimplify, when two
men are confined in a small cell, they can get to know one another as people and learn to see each other not
primarily as members of a race to which they are potentially hostile, but instead as individuals to be judged as
such. Id. These results bode well for California, which, in the course of settlement talks with Garrison Johnson
after the Court’s decision in Johnson v. California, committed to a program of racial integration modeled on that
of Texas.
383. See supra note 382.
384. Indeed, this is precisely the argument offered in Part III.B above.
385. Admittedly, this particular worry is speculative. As a practical matter, therefore, the extent to which a
given court might credit it as sufficient in itself to demonstrate the narrow tailoring prong would depend on the
standard of proof the court demands of defendants. As Justice Stevens noted in his opinion in Turner v.
Safley—and repeated in his dissent in Johnson—“[h]ow a court describes its standard of review . . . often has far
less consequence for the inmates than the actual showing that the court demands of the State in order to uphold the
regulation.” Turner v. Safley, 482 U.S. 76, 100 (Stevens, J., concurring in part and dissenting in part); Johnson,
542 U.S. at 523 n.3 (Stevens, J., dissenting).

86

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

This concern is particularly salient in L.A. County itself, given the long history of
K6G and the widespread understanding among those with experience in the Jail as
to the sexual identity of K6G’s residents.
In any case, the fact that in many carceral contexts, a unified approach might
achieve the same level of protection for gay men and trans women while reducing
the harm that arises from official identity-based classification should not be
enough to justify putting an end to K6G. Given the complexity of the L.A. County
Jail and the many contingent factors that determine whether a given housing
program is successful, it seems wiser to favor the certain present success of K6G
against the speculative protection that might come from dismantling K6G and
establishing a new unified unit. A court should seriously hesitate before concluding
that K6G fails strict scrutiny only because a theoretical alternative approach exists.
In sum, even were K6G or other like units designed on this model to be assessed
under strict scrutiny, the state would have a strong argument in defense of the
program—and especially in defense of K6G itself. Moreover, although there is
reason to lament this fact, neither sexual orientation nor gender identity is
presently considered a protected class under prevailing Equal Protection doctrine.
This means that claims of discrimination on these grounds receive no heightened
scrutiny. At best, an argument could be made for the searching rational basis
review—what some scholars have termed “rational basis with bite”386—that, it
might be argued, is available in claims of official discrimination based on sexual
orientation after Romer v. Evans.387 And in the case of K6G, this is a standard that
could readily be met.388 This being so, it seems hard to credit the assertion that
being housed in K6G violates Equal Protection. In any case, for such a challenge to
succeed, a claim must be brought by someone with standing. Because those people
currently being housed in K6G strongly prefer it to GP, residents are unlikely to
bring a case challenging the constitutionality of the unit. A court would thus have
to find standing on the part of someone who sought admission to K6G but was
refused. And even assuming someone in that situation would have the resources to
litigate the case, there are few incentives to bring suit, since even a victory would

386. See e.g., Kenji Yoshino, The New Equal Protection, 124 HARV. L. REV. 747, 759–61 (2011) (referring to
the Supreme Court’s review of Equal Protection claims based on sexual orientation as “rational basis with bite”)
(quoting Gayle Lynn Pettinga, Note: Rational Basis with Bite: Intermediate Scrutiny By Any Other Name, 62 IND.
L.J. 779, 780 (1987)); Jeremy B. Smith, Note: The Flaws of Rational Basis with Bite: Why the Supreme Court
Should Acknowledge its Application of Heightened Scrutiny to Classifications Based on Sexual Orientation, 73
FORDHAM L. REV. 2769, 2770 (2005) (same).
387. See Romer v. Evans, 517 U.S. 620, 632 (1996) (striking down an antigay Colorado constitutional
amendment on rational basis review in part because it was “inexplicable by anything but animus toward the class
it affects”). But see Yoshino, supra note 386 at 761 (cautioning that “rational basis with bite” has not proved in
practice to operate like “formal heightened scrutiny”).
388. The Romer Court emphasized the animus motivating the state constitutional provision struck down in that
case. See Romer, 517 U.S. at 632. Yet in the case of K6G, it is the desire to protect the affected class and not to
injure it that motivates the enterprise.

2011]

STRATEGIC SEGREGATION

87

not bring the desired relief.389 Were such a plaintiff to prevail, the likely remedy
would be an injunction ordering the unit disbanded—a result antithetical to the
interests of someone who wished to gain access to it.
IV. CONCLUSION: THE PROSPECTS FOR REPLICATION
Part III addressed several objections to the K6G model, on which gay men and
trans women in custody are segregated from the rest of the detainee population for
their own protection. It concluded that in most cases, the better approach may well
be the unified strategy endorsed by the National Prison Rape Elimination Commission in its 2009 draft standards. At the same time, it argued that the K6G model
should remain available to those prison officials who find it the best way to keep
safe the sexual minorities in their custody.390
To show how such a policy might be effectively implemented, this Article has
drawn on the example of the L.A. County Jail, which for over twenty-five years
has been housing gay men and trans women separately from the Jail’s general
population. Although by no means impervious to criticism, the K6G program has
created a surprisingly safe environment for people who in other carceral contexts
would be at great risk of abuse. For this reason alone, it merits the attention of
anyone committed to reducing the incidence of rape and other forms of sexual
victimization behind bars.
But a final question remains: is L.A. County’s K6G program even replicable? Or
is the success of K6G a product of features unique to L.A. County? My study of the
K6G unit reveals a number of factors that would increase the chances of this model
succeeding elsewhere. First, it helps to have a sizable population of gay men and
trans women in the jurisdiction. L.A. County is home to a high number of gay men
and trans women.391 At any given time, there is thus likely to be a sizable number
of detainees in the Jail who belong in K6G. This circumstance has a number of
benefits for the program as a whole. With a large enough population, devoting an
entire housing unit would not yield a net loss of bed space, which is a serious
concern in a perennially overcrowded facility. A large enough population allows
vulnerable prisoners access to both security and community, ensuring that no one

389. Conceivably, a party could seek to bring suit against K6G on principle, but in doing so may face standing
and other jurisdictional hurdles. And for the reasons I have suggested here, any such case should lose on the
merits.
390. Arguably, the best way to ensure that this model is employed only when strong grounds exist to think it
the most effective way of protecting LGBT prisoners is to restrict its availability to cases where LGBT advocates
agree with this assessment. In the case of L.A. County, for example, more effective segregation was the legal
remedy sought by the ACLU of Southern California on behalf of its clients, gay men detained in the Jail. This
experience informed the proposal advanced by a number of LGBT advocacy organizations, and in my own
separately submitted comments, see Dolovich, supra note 344, in response to the January 2011 Notice of
Proposed Rulemaking issued by DOJ regarding its proposed PREA standards. For the substance of this proposal,
see supra note 344.
391. See supra note 107.

88

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

has to choose between safety from sexual assault and satisfying his or her basic
human need for the company of others. Finally, such a large population allows for
multiple housing units, which in turn means that jail officials have options besides
protective custody for housing people who have developed mutual enmities.
It is difficult to say how small is too small. But even recognizing the need for
some critical mass of gay men and/or trans women to make a segregated unit the
more desirable alternative to protective custody, it is safe to say that in many prison
systems across the country, there are likely to be sufficient numbers to form the
basis for a viable independent unit. This is especially so in any system encompassing a sizable urban area. Certainly, the L.A. County Jail is enormous, but there are
many state prison systems with a much larger overall prisoner population, which
depending on the demographic make-up may make it possible to successfully
implement the L.A. County approach.392
Second, for the K6G model to work, it is crucial that the facility’s command
staff has a genuine commitment to the protection of vulnerable groups—and that
this commitment be clearly communicated to the line officers and other staff who
run the facility day-to-day. Such a commitment has been crucial to the success of
K6G. The evidence can be hard to notice, but only because its most potent aspects
lie in the seamlessness with which K6G has been absorbed into the daily
functioning of the Jail. In an operation the size of L.A. County, it is no small thing
to introduce specialized classifications into the central intake mechanisms. To
make this happen, senior administrators have to make clear that it must happen,
that successful implementation of the program is non-negotiable. K6G works
because the processes that emerged in response to that directive became routine,
and because when issues arise, institutional resources are devoted to their resolution—although perhaps not always as expeditiously as friends of the unit would like.
This brings us to the third point: it helps to have the support and participation of
the LGBT community on the outside. In the case of K6G, the Los Angeles LGBT
community has been deeply involved in the project from its inception, and has
worked with jail officials in an ongoing way to develop and adapt the classification
process and to help make sure that residents’ needs are met. Because of this
community involvement, K6Gs enjoy a range of services tailored to the particular
needs of the population,393 and Jail officials know that they will hear from outside

392. However, a sizeable gay and transgender population is not sufficient to ensure the success of a K6G-style
unit. The New York City Corrections Department for many years operated a segregation unit at Rikers Island for
gay men and trans women, to which detainees could gain admittance merely by declaring themselves eligible. See
Paul von Zielbauer, City Prepares to Close Rikers Housing for Gays, N.Y. TIMES, Dec. 30, 2005. As a
consequence, the unit mixed genuinely vulnerable individuals with “violence-prone inmates” who claimed to be
gay in order to prey on other residents of the unit. Id. The experience of Rikers recalls that of L.A. County’s
“homosexual” housing prior to the establishment of K6G, and suggests that the present tight control over
admissions, see supra Part II.C.1-2, is a key component of K6G’s success.
393. The condom distribution program, for example, was initially conceived by CorrectHELP (now the Center
for Health Justice), an independent advocacy organization that has administered the program since its inception.

2011]

STRATEGIC SEGREGATION

89

advocates if conditions in the unit fall below minimally acceptable standards.394
Finally, the example of K6G teaches that the attitudes of the officers in charge of
the unit, both toward the residents in particular and the enterprise in general, matter
enormously. L.A. County has been fortunate in the two men who have run the unit
for the past several decades. Senior Deputy Randy Bell and Deputy Bart Lanni are
committed to maintaining the boundaries K6G places between its residents and the
other inmates in the Jail. But they are equally committed to doing what they can to
meet the needs and improve the prospects of the people in their custody.395 And
perhaps most importantly, they regard themselves as advocates for K6G in the Jail
itself, readily taking concerns and complaints to the institution’s command staff
and just as readily addressing problems with peer deputies.396
That Bell and Lanni are often called upon to address issues arising from the
behavior of fellow deputies points to one component of the K6G program that
needs ongoing attention, and to one aspect of any similar program to which
attention would assuredly need to be paid: the conduct of the custodial officers
who interact with unit residents. In L.A. County, Bell and Lanni have responsibility for who gets classified into the unit. But once housing assignments have been
made, the dorms are administered just like any others. Non-specialized line
K6G also has full-time program officers from Tarzana Treatment Center, a reentry project aimed at those former
prisoners who are HIV-positive, and on-site sexually transmitted infections (STI) testing administered by the L.A.
County Department of Health.
394. The ACLU of Southern California has a full-time staff member assigned to the Jail, who monitors
conditions and works with Jail officials to resolve issues. See, e.g., Jails Project, AMERICAN CIVIL LIBERTIES UNION
OF SOUTHERN CALIFORNIA, http://www.aclu-sc.org/jails/ (last visited Jan. 18, 2010). But the Jail is enormous, and
one person can only do so much. That K6G is on the radar screen of so many outside organizations ready to
advocate on behalf of its residents helps to explain the relative appeal of K6G compared with other units in the
Jail.
395. The primary responsibility of these two officers is classification, i.e., determining which of the Jail admits
seeking access to K6G actually belong there. However, over the years, they have taken it upon themselves to
expand their own job descriptions to include a range of initiatives directed toward improving the lives and
prospects of the people classified to K6G. The two officers have developed a full schedule of educational
programs open to anyone in the unit. See supra note 138. They have worked with outside providers to make a
range of services—counseling, drug treatment, STI testing, etc.—available to residents of the dorm. See supra
note 165. And they work with outside advocates to identify and fulfill the particular needs of the K6G population,
efforts that often require them to navigate multiple channels of the extraordinarily cumbersome jail bureaucracy.
For example, in the last half of 2009, Lanni worked with Commander Robert Olmstead and Captain Buddy
Goldman to get permission for trans women in K6G to have cosmetics in the dorms on the same terms as
detainees in the women’s facility. Lanni also worked with Dr. Keith Markley, Supervising Psychiatrist at Men’s
Central Mental Health Service, to ensure access to hormone therapy for a number of the trans women in the unit.
Bell and Lanni, when they judge it appropriate, also serve as counselors to those K6Gs who seek them out, doing
what they can to help improve an individual’s prospects and reduce the likelihood of a return visit to the Jail.
Bell and Lanni are no pushovers. They know when they are being played. But from what I have observed, they
approach their work with a strong sense of respect for the people in the unit. In fact, they have shown me by
example that this is perhaps the key requirement on the part of custodial officers if carceral conditions are to be
humane.
396. This readiness to challenge the behavior of their peers vis-a`-vis K6G has not endeared Bell and Lanni to
their colleagues. That they do it anyway is a testament to their commitment to the well-being of the people in the
unit.

90

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

officers staff the booths overlooking the K6G dorms and go in and out of the dorms
to supervise the delivery of meals and clothing exchange, to conduct count, and to
deal with any security issues. Officers assigned to the floor on which K6G is
located understand the distinct status of the unit and follow policies designed to
maintain the necessary boundaries between K6Gs and GPs.397 But they may have
no particular sympathy for the residents of K6G, and in many cases may be
uncomfortable working with the populations K6G serves. As a consequence,
officers may verbally harass or otherwise maltreat the K6Gs. In engaging in such
misbehavior, line officers may be prompted by the same anxieties some detainees
feel that others might suspect their sexuality. They thus may be tempted to prove
their own masculinity by abusing those not considered “real men.” But whatever
the cause, mistreatment of K6Gs by line officers in the Jail indicates that Jail
leadership must do much more to ensure that K6Gs are treated with respect.
Such mistreatment highlights the risks involved in efforts to replicate this
program elsewhere. One danger of segregating sexual minorities is that they will
become lightning rods for abusive manifestations of the sexual anxieties of the
other men who come into contact with them, whether fellow prisoners or officers.
As regards fellow inmates, the best an institution may be able to do is to keep them
as far apart as possible. As to custodial officers, however, it is the institution’s
responsibility to make clear that abusive or otherwise disrespectful behavior will
not be tolerated and to back up this imperative with disciplinary action for those
found to be in violation. It may be impossible to wholly eliminate such abuses, but
that is no reason not to reduce their incidence as much as possible.
I have tried in this brief conclusion to identify some considerations that may be
relevant to any officials contemplating following the lead of L.A. County. Again,
my aim throughout has not been to promote the K6G approach as against all other
strategies for reducing prison rape. To the contrary, I agree with the Commission
that prisons and jails must carefully screen every individual on arrival to determine
the likelihood of that individual being victimized by or sexually abusive to others,
and that housing assignments should be made accordingly, with ongoing monitoring to reassign any emergent victims or predators to more appropriate housing. My
argument has simply been that K6G has been a relatively successful program in
L.A. County, and that, in some cases, the best practice for ensuring the safety of
vulnerable inmates may instead be a two-track program, modeled on K6G, in
which gay men, trans women and other sexual minorities are segregated out and
housed apart even from other vulnerable prisoners.
For some readers, the suggestion that this sort of state-sponsored identity-based
segregation would ever be appropriate may seem anathema. And there is no doubt
that this strategy is extremely far from what would exist in an ideal world. But

397. These policies include escorting K6Gs moving through the facility and making sure that no GPs mix with
the K6Gs when the K6Gs are out of their dorms. See supra Part II.B.

2011]

STRATEGIC SEGREGATION

91

especially in an era of mass incarceration, prisons are an ugly business. The
problem of prison rape admits of no ideal solutions. If we await a policy fix with no
risks, with no downsides, we will wait forever. And while we wait, people in prison
will continue to be raped. It may seem odd to champion an official policy of
identity-based segregation in the cause of civil rights, but as Equal Protection
doctrine plainly recognizes, under some circumstances, such policies may be the
only way to overcome the greater evil.398 If the vulnerability to rape of gay men
and trans women behind bars represents such a circumstance—and I believe it
does—a segregationist approach may at times be justified. And, in that case, it is
worth understanding how such an approach might effectively be put into practice.
It is to this end that I have described L.A. County’s K6G program in such detail.
Although not without its problems, K6G offers proof that, done right, a unit of this
sort can do far more good than harm. For this reason, it deserves our attention,
consideration, and esteem, regardless of whether any other jurisdictions follow its
lead.

398. See supra note 303.

92

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

METHODOLOGICAL APPENDIX
The research on which this Article is based was conducted in the Los Angeles
County Jail over approximately seven weeks in the summer of 2007.399 Data was
collected through five channels:
(1) observations of jail operations;
(2) informal conversations with prisoners, staff, volunteers, and other service
providers in the facility;
(3) observation of K6G classification interviews;
(4) formal interviews with a random sample of K6G residents; and
(5) in-depth interviews with Senior Deputy Randy Bell and Deputy Bart
Lanni of the L.A. County Sheriff’s Department.

1. Observation of Jail Operations
The L.A. County Jail is composed of eight different facilities. This research
focused on the two main facilities located in downtown Los Angeles: Men’s
Central Jail, and Twin Towers Correctional Facility. Over the course of the data
collection, I observed the operation of various areas of these two facilities in order
to understand how things worked at each location and the way the institution
functioned as a whole.400 To this end, I visited and observed many different parts
of the Jail, including the Inmate Reception Center (IRC), visiting room, court line,
hospital wing, library, law library, investigative line-up room, and a range of
general population and specialized housing units. The Jail has a number of
specialized units besides K6G. Over the course of my time in the Jail, I observed
several such units, including those housing prisoners who were seriously mentally
ill,401 deaf, developmentally disabled, former law enforcement officers or relatives
of law enforcement officers, and sexually violent predators.402
During these observations, I took notes when it seemed like doing so would not
draw undue attention to myself or otherwise influence the course of what I
observed. When taking notes in the moment did not seem advisable, I noted down
my observations at the first available opportunity. Each evening, I dictated my
notes into an audio recorder. These recordings were then transcribed and became
part of my field notes. My note-taking was atheoretical. The aim was just to record
what I was seeing for later analysis.
In most cases, these observations took place between 7:45 a.m. and 4:30 p.m.,
Monday to Friday. One exception was my visit to IRC. The busiest time of day for

399. UCLA IRB #G07-01-106-03.
400. At the start of my research, Chief Alex Yim granted me open access to these facilities, which made it
possible for me to move unescorted through the Jail.
401. There are many pods in Twin Towers that house this population. I observed one or two of these.
402. Those in this last group had been civilly committed to extended custody following the completion of their
criminal sentences, and at the time of my observations were housed in a single pod in Twin Towers.

2011]

STRATEGIC SEGREGATION

93

IRC is approximately 6 p.m.–11 p.m. This is because the majority of new admits to
the Jail are referred from court. Each courthouse that feeds the Jail has a holding
tank for the people taken into custody, and these groups are typically not
transferred to the Jail until court is over for the day. There are no weekend
transfers, so Mondays and Tuesdays—when weekend arrestees are finally brought
to the Jail—are the busiest days in the IRC. On those days, as many as 1000 people
may arrive for processing, and it can take up to five hours for everyone to be
processed. For this reason, I visited the IRC on a Monday evening, arriving at
6 p.m. and staying until things slowed down, which that evening was approximately 10:30 p.m. During this period, I toured the various stages of the intake
process, including the unit in Twin Towers where the Jail houses the new admits
needing medical care on arrival. I then sat for several hours behind a custody
assistant, observing her as she conducted classification interviews,403 an experience that provided an important contrast to the detailed second stage of the K6G
classification process.
Observations of the Jail’s general operations occurred unsystematically during
the period of data collection. By contrast, I daily observed the workings of K6G, as
a way to understand how the unit operates, the experience of residents, the nature
of the interaction between staff and residents, and the particular role played by the
officers charged with running the unit. K6G occupies two areas of Men’s Central.
Upstairs is the administrative center of K6G, made up of the classification office,
the classroom, and a third room used by the various service providers who work
with K6G residents.404 The office and classroom are directly opposite one another,
divided by a narrow hallway. This hallway contains benches on which new admits
sit awaiting their classification interviews or blood tests.405 The third room is
somewhat apart from the office and classroom, about 50 feet down the hall. One
floor down are the three dorms where K6G residents are housed.
Each day, I divided my time between K6G’s administrative center and its
housing units, starting upstairs in the morning and moving downstairs later in the
day. This allocation was dictated by the schedule of Officers Bell and Lanni, who

403. I thank Los Angeles Sheriff’s Department Custody Assistant Regina Fowlkes for her gracious willingness
to have me observe her work and for her patient explanations as to what she was doing and why.
404. These providers include the Tarzana Treatment Center, which provides reentry services and drug
treatment for HIV-positive detainees, and the Center for Health Justice, which, in addition to distributing condoms
in the dorms once a week, provides HIV counseling and prevention education to unit residents.
405. Prior to being sent to the dorms, all incoming K6Gs ordinarily have a syphilis test administered by staff
from the LA Department of Health, who are permanently assigned to K6G. These technicians also provide
optional testing for other sexually transmitted infections (or “STIs”), including HIV, gonorrhea, and Chlamydia.
This program was initiated in March 2000, after the Centers for Disease Control in Atlanta traced a syphilis
outbreak to the Jail’s K6G unit. The syphilis test is not mandatory, but Jail practice is to deny access to the K6G
dorms to those who decline to have the test and to house them instead in K6G-designated single cells in the
administrative segregation wing. This practice is justified by Jail administrators as a precaution against the
transmission of syphilis. Interview with Deputy Bart Lanni, L.A. Sheriff’s Department, in L.A., Cal., February 12,
2011.

94

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

run the K6G classification office. These officers start their shift at 5:30 a.m. and
leave between 1:30 p.m. and 2:00 p.m. each day. Each morning I would arrive
around 7:45 p.m., and stay upstairs until Bell and Lanni left for the day. I would
then move downstairs to the housing units, where I would remain until 4:30 p.m. or so.
While upstairs, I moved back and forth between the classification office, the
classroom, and the third room where I conducted my formal interviews. During
this time, when I was not conducting my own interviews, I sat in on classification
interviews (on which more below), observed classes in the classroom, and was a
fly on the wall in the office as Officers Bell and Lanni addressed the issues that
arose daily in the unit.
In the afternoons, I moved downstairs to the officer’s booth overlooking the
three K6G dorms. In the area of Men’s Central where K6G is located, dorms are
configured in groups of four, with one officers’ booth per four dorms. The officers’
booths are H-shaped, with one dorm extending straight up from each of the four
tips of the H (ie, two at the top and two at the bottom). The booths have large glass
windows at each end that allow officers to watch the goings-on in the dorms,406
and narrow openings in the glass, measuring 18 inches by 3.5 inches, at the bottom
of each window through which officers and residents can communicate. I spent
hours each day in the booth overlooking the K6G dorms, observing the distribution
of meals, clothing exchange (during which residents exchange dirty clothes and
linen for clean), count, and the general goings-on. Because dorms are in groups of
four and there are only three K6G dorms, I also wound up observing life in a
general population dorm, which by happenstance became my control.
As noted, most days I left the Jail at approximately 4:30 p.m. On two separate
occasions, however, I stayed through the evening, observing events in the dorms
until close to midnight in one case, and past midnight in the other.
2. Informal Conversations with Staff, Residents, and Others
Over the course of my observations, I encountered many different parties,
including Sheriff’s deputies, custody assistants, nurses, chaplains, teachers, volunteers, and Jail detainees. As often as possible, I engaged in informal conversations
with the people I met, with the aim of learning about the workings of the Jail and
the experience of all those who occupy it, staff and detainees alike. When visiting
various areas of the Jail, I made a particular point of talking to the officers in
charge of those areas, as a way to understand how things worked in the particular

406. The configuration of the dorms off the officers’ booth has a serious flaw: the dorms are rectangular in
shape, and the officers are situated at one end, making it hard to see what is happening at the other end. Officers go
daily into the dorms when meals are delivered (when residents are lined up preparatory to receiving their meals)
and to perform count (when residents are face-down on their bunks, with their wrists crossed above their heads
and dangling over the top of the bunk so their inmate numbers, worn on plastic wristbands, may be confirmed).
But from what I observed, officers almost never do random walk-throughs at other times. This leaves plenty of
opportunity for residents to engage in unobserved behavior, especially in the back corners of the dorms.

2011]

STRATEGIC SEGREGATION

95

location and the challenges faced in each. While upstairs in K6G’s administrative
center, I had innumerable informal conversations with Bell and Lanni, with any
K6G residents who happened to be around, and with the volunteers, service
providers, and other staff who worked in or came through the unit. During the
periods of observation in the officers’ booth overlooking the K6G dorms,407 I had
frequent informal conversations with the duty officers assigned to the dorms, with
many of the deputies assigned to the floor, and with many K6G residents, who after
they became accustomed to my presence would initiate conversations with me
through the opening in the booth’s window.
These conversations took place during the same time frame as my observations,
as described above. In each case, I obtained verbal consent. The content of these
conversations was recorded in the same way as my observations: I took detailed
notes at the time or as soon as possible after the conversation took place, and each
evening I dictated my written notes. This dictation was later transcribed and
became part of my field notes.
3. Observation of Classification Interviews
Each morning at 5:30 a.m., Officers Bell and Lanni arrive to find as many as 20
people in the K6G holding tank awaiting classification. Of these, as many as 10
may require an in-depth interview to determine whether they satisfy the K6G
admissions standard.408 For approximately two to three weeks, I met with every
individual to be classified and sat in on approximately 80% of the interviews. Once
I started conducting formal interviews myself (see below), my observation of these
interviews became more random. Observation of the classification interviews
conducted by Officers Bell and Lanni was a crucial component of this research.
Although new arrivals are asked during the brief IRC classification interview
whether they are “homosexual,” with those answering in the affirmative initially
classified to K6G, the final determination as to who will be admitted to the K6G
unit is made entirely based on this second-stage interview. To fully understand how
the program works and who winds up in K6G, it is necessary to have a clear and
accurate picture of this second-stage interview process.
Prior to each interview, I met in a separate room with each person awaiting their
classification interview, to explain my project and obtain verbal consent to my
sitting in on that exchange. The room used for this purpose was the third room of
what I have called K6G’s administrative center, that used by the various service
providers who work with K6G’s population. At that time of the day, this room was
typically empty. To ensure valid consent, it was important to assure prospective
407. During any given shift, there were two custody assistants in the booth, each responsible for two of the four
dorms in the configuration. In addition, there were five to seven deputies on the floor assigned as “prowlers”,
whose job it is to supervise distribution of meals and clothing exchange, and conduct count in each of the floor’s
ten dorms, to monitor comings and goings, and generally to keep order on the floor.
408. For extended discussion of this interview process, see supra Part II.C.2.

96

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

subjects that Officers Bell and Lanni would not know if they declined to have me
sit in on their interviews. But unless steps were taken, it would be obvious to these
officers which individuals declined to consent to my presence, since during their
interviews I would be conspicuously absent from the classification office. To get
around this problem, I told the two officers that I would only be sitting in on
approximately eighty percent of the interviews of those subjects who consented.
That way, in any given case where I did not observe the interview, Officers Bell
and Lanni would not know if I remained outside because the subject did not
consent or because the subject was one of the twenty percent whose interviews I
would not be observing notwithstanding their grant of consent.409
To make interview subjects feel comfortable, I did not audio-record the
classification interviews that I observed. Instead, I took detailed notes over the
course of each interview, which I dictated each evening. The transcripts of these
recordings became part of my field notes. In all, I observed approximately 50
interviews, and, of those, I recorded notes on approximately 34.
4. Formal Interviews with a Random Sample of K6G Residents
Over the course of approximately four weeks, I conducted formal interviews
with 32 K6G residents,410 based on a 176-question questionnaire.411 These
interviews took place in the early part of the day shift, from approximately 8:00
a.m. to 1:30 p.m. Interviews took anywhere from one to two hours. The content of
these interviews provided a detailed understanding of the experience of residents
of K6G that could not be acquired any other way.
The sample was selected as follows. On the first day I commenced my
interviews, Deputy Lanni gave me copies of three lists with the names of K6G’s
residents that day,412 one for each dorm. I went down the lists, highlighting every
third name. Those highlighted lists became the pool from which I drew my sample.
Each day, I worked off a different list, rotating among the dorms. On arrival, I
would pick three or four highlighted names, and, starting with the first name,
would ask Senior Deputy Bell to send a pass for that individual to be sent up from
the dorms. In almost every case, the pass was sent for the person I requested.413
When the proposed subject arrived from downstairs, I initiated a brief conversation, explaining who I was and that I sought to enlist their help with my research.

409. As it happened, in the end, only one subject declined to give consent.
410. I consented and commenced the interview process with thirty-three subjects, but one subject proved to be
developmentally disabled, an incapacity that had not been evident during the consent process. In that case, I
terminated the interview without completing the questionnaire.
411. This questionnaire is reproduced below as Appendix B.
412. Every day, hundreds of people are admitted to the Jail and hundreds more are released. For this reason, the
list of individuals housed in any given unit will change from day to day.
413. In a very few instances, on learning the identity of the person I was asking to see, Senior Deputy Bell
declined to send the pass, either because he felt the person was not someone I could safely be alone with in the
interview room or because the person had psychological problems that made them incapable of participating.

2011]

STRATEGIC SEGREGATION

97

In perhaps ten percent of the time, this brief conversation was enough to make
clear that the proposed subject was not appropriate to be interviewed, and the
individual was therefore excluded from the sample. In most such cases, the
individual only spoke Spanish. Because I do not speak Spanish and because the
substance of the interviews was sensitive enough that introducing a translator
would have risked complicating and even compromising the interview process, I
excluded non-English speakers from participation. In other cases, it was immediately clear that the proposed subject was unable to engage in the interview process,
whether for reasons of mental illness, developmental disability, or other psychological
incapacities.
In the remaining cases—the vast majority—I escorted the proposed subjects to
the interview room,414 where I explained that they had been chosen at random
from among K6G residents, described the nature of their proposed participation,
and sought written consent from those who chose to participate.415 Among other
assurances, I made clear that the decision whether to participate would have no
bearing on their cases, circumstances in the Jail, or prospects for release. A
difficulty existed, however, because those who chose not to participate would be
out of the interview room in just a few minutes, whereas those who did choose to
participate could remain for as long as two hours. In this way, the subject’s
decision whether or not to participate could become known to any officers or
fellow detainees on the floor at the time, a possibility that could create pressure to
participate on the part of subjects worried that the institution would frown on those
who declined. To address this problem, I employed the following procedure, which
I explained to Officers Bell and Lanni as well as to each person from whom I
sought consent. I asked each person who consented to participate to roll a single
die. If they rolled a 1, 2, 3, or 4, we proceeded to the interview. If they rolled a 5 or
6, their participation was at an end. This way, when a proposed subject left the
interview room after only a few minutes, Bell and Lanni would not know whether
it was because the individual had declined to participate, or because the die came
up 5 or 6. As it happened, not one person I invited to participate declined to do so,
and with only one exception,416 those who rolled a 5 or 6 expressed disappoint-

414. This room was reserved for my exclusive use during this period. During the interviews, the subject and I
were alone. But the room is equipped with a closed-circuit camera, by means of which Officers Bell and Lanni
could observe (but not hear) the interview, thus enabling them to ensure my safety should any problems arise
without compromising the privacy of the exchange.
415. Copies of the consent forms were not given to interviewees. Jail residents have no possibility of keeping
their papers private. Any record of participation would have therefore risked compromising the assurance of
confidentiality. Though Jail staff and/or fellow K6G residents might know which residents were considered for
participation in the study, they would only know which residents agreed to be subjects if they saw a signed
consent form.
416. The lone exception was a man who seemed to have no affect, and who was the only prisoner I met during
my time in the Jail of whom I was affirmatively afraid. I was relieved when he rolled a 5, although it was pointed
out to me afterwards that even had his roll qualified him to participate, I ought myself to have terminated the

98

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

ment with having to leave.417
Interviews were audio-recorded and transcribed, in part by staff of the UCLA
School of Law, and in part by commercial transcription services.418 Research
assistants then went through the transcripts systematically, summarizing the main
findings for each grouping of questions. Personal data from Qs 1–6 & Qs 140–50
were collated into a spread sheet and where possible analyzed for trends (e.g.,
breakdown of subject group according to age, race, etc.). For each question that
yielded data amenable to categorization or quantification (e.g., Qs 119–24, Q125,
Qs 151–56), answers were coded and the range calculated. The answers to Q35,
which asked subjects for “5 words to describe life in K6G,” were catalogued and
analyzed for the range of positive and negative words, and explanations for each
answer were collected and assessed. Because the aim of the interviews was to learn
generally about life in K6G, much of the data collected were qualitative. Analysis
therefore involved carefully reading through the transcripts to identify themes and
trends and to develop a picture of the K6G experience. In many cases, the wide
variation in the answers pointed to a lack of veracity on the part of some subjects.
Reading through the transcripts with an eye to such instances contributed to an
overall impression of which answers were more reliable—and also revealed
potential weaknesses in the questionnaire itself, which informed data analysis.419

interview and had him returned to the dorms. I resolved to do this in future should I feel that way again, but
thankfully the issue never came up.
417. The reason for this disappointment became apparent over the course of my time in the Jail. The unit is
small enough that by the end of my first week of interviews, many K6G residents knew about my research. In Jail,
there are very few distractions, and even fewer opportunities for residents to make any kind of positive
contribution or be the focus of positive attention. Interview subjects seemed to enjoy the interview process, and
reported their experience to their friends and colleagues back in the dorm. As a consequence, when I appeared in
the officers’ booth overlooking the dorms, I frequently had people coming to ask me if I would interview them.
These requests had no bearing on whether a given individual ended up in my sample, and because I ultimately
interviewed only around ten percent of K6G’s residents, in most cases the individuals offering themselves as
research subjects did not wind up being interviewed. But this did occasionally happen, and one day in particular I
was surprised to find that I had sent a pass for one of my most dedicated petitioners, whose moniker in the Jail, as
is often the case, turned out not to be his legal name. This individual was very pleased to have been invited to
participate, and no doubt believed his persistence had paid off.
418. Each UCLA staff member who transcribed these interviews, along with each research assistant who
subsequently worked on the project and each law review editor who consulted the transcripts during the editorial
process completed an on-line course on the protection of research subjects administered through the UCLA IRB.
Each commercial transcription services employee who worked on this project gave written assurance that they
would keep the contents of the interviews confidential.
419. For example, Q61 asked what percentage of K6Gs have no consensual sex at all while in K6G? The
answers ranged widely, with the majority suggesting that the number of people who remained celibate while in
K6G was extremely small. Yet to Q79, which asked have you ever had consensual sex while incarcerated?, a
sizeable number responded in the negative. After a while, I realized that unless many of those in my sample were
just coincidentally among the very few celibate residents of K6G, I was getting unreliable answers to either Q61
or Q78. Given that a constant theme throughout the interviews was that K6G was the site of a considerable
amount of consensual sex, the high numbers of negative answers I received to Q78 may simply have reflected a
reticence on the part of those I interviewed to admit to me in a face-to-face interview that they themselves had sex
in the unit. It is, however, instead (or also) possible that the estimates of sexually active K6Gs made in response to

2011]

STRATEGIC SEGREGATION

99

5. In-Depth Interviews with Officers Bell and Lanni
Lastly, I conducted in-depth interviews with Senior Deputy Randy Bell and
Deputy Bart Lanni. These two officers are responsible for deciding which of the
detainees who report being gay during their initial IRC interview meet K6G’s
admissions criteria. They also run K6G’s educational programming, manage the
various outside providers who work with K6G’s population, and are generally
responsible for the unit’s operation. Together, they have a combined forty years
working with the K6G population. Their long history with the unit made them the
best existing sources of accurate information about the evolution and functioning
of the unit. I therefore spent a total of several hours in focused conversation with
each. My interview with Deputy Lanni was conducted in one sitting. My interview
with Senior Deputy Bell took place over several days. As to each officer, I asked
questions relating to:
Y
Y
Y
Y
Y
Y
Y
Y

their background prior to joining the Los Angeles Sheriff’s Department;
how they came to be assigned to K6G;
the early history of K6G (i.e., prior to their assignment to the unit);
their understanding of the reason for K6G within the LA County Jail
system;
their sense of the differences between life in K6G and life in the Jail’s
general population;
their precise responsibilities in the unit;
their understanding of their own professional mission vis-a`-vis K6G; and
anything else they’d like to tell me about K6G or their impression of the
people housed in the unit.

The interview with Deputy Lanni, being the more formal of the two, was
audio-recorded and transcribed. Because my interview with Senior Deputy Bell
was more ad hoc, I took written notes that were then transcribed with my field
notes. These interviews provided valuable background on K6G rather than data to
be reported. Thus, the resulting transcripts were not analyzed, but have simply
served as a source of information and thematic insight.

Q61 were overinflated, and that my subjects overestimated the number of K6Gs who engaged in consensual sex in
the dorms.

100

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

APPENDIX B
DOLOVICH K6G PROJECT QUESTIONNAIRE420
FINAL VERSION
Respondent ID# _______
Interviewer Initials: _______
Date: _______________
Interviewer Start Time:___________
First, I’m going to ask you some questions about your history in K6G, your
impressions about how K6Gs are treated in the Jail, and about life in K6G. Please
remember that your answers to these questions are not going to be shared with any
staff (including Bell and Lanni) or other inmates. Please also remember that if you
tell me anything related to any sexual activity, consensual or otherwise, between
any inmate (whether yourself or anyone else) and any employee of the L.A. County
Jail or any other staff member or service provider, I will be obliged to report it to
Jail officials. In other words, if you tell me of any such activity, it will be as if you
are making an official report of the incident.421
1. Is this your first time in the LA County Jail? [if yes, go Yes
to question 7]
No
2. [If no] How many times have you been in the Jail?
3. Is this your first time in K6G?

Yes
No

4. [If no] How many times have you been in K6G?
5. Have you ever been in the Jail and not been in K6G?

Yes
No

6. If so, where were you housed?
7. When did you first learn about K6G?
8. [If other than once admitted to LA County]: How
did you learn about it?
9. [If heard about it on the outside] What were you told
about K6G? How was the unit described to you?
10. Do you prefer being in K6G to being in the general
population?

Yes
No

11. Why?/Why not?

420. This questionnaire was developed in close collaboration with Joe Doherty, Director of the UCLA School
of Law’s Empirical Research Group. Although the unit is now referred to as K6G, at the time of my interviews,
most of my respondents—many of them with a long history of detention in the L.A. County Jail—still referred to
it by its prior name of K11. My interview questions therefore used the term “K11” instead of “K6G.” To avoid
confusion, all references to K11 have been changed to K6G.
421. See supra note 296.

2011]

STRATEGIC SEGREGATION

101

12. What are the best things about being in K6G rather
than GP?
13. Anything else?
14. Anything else?
15. Anything else? [Ask until answer is no or until 5
things]
16. What are the worst things about being in K6G instead
of GP?
17. Anything else?
18. Anything else?
19. Anything else? [Ask until answer is no, or until 5
things]
20. Is life in K6G more restrictive than life in GP?

Yes
No

21. In what way?
22. Any other ways?
23. Any other ways? [Keep following up until 5
answers, or they run out of answers]
24. Do the guards treat K6Gs any differently than GP?

Yes
No

25. In what way?
26. Any other ways? [Keep following up until 5
answers, or they run out of answers]
27. Does the staff treat K6Gs any differently than GP?

Yes
No

28. In what way?
29. Any other ways? [Keep following up until 5
answers, or they run out of answers]
30. Do you ever have interaction with GP?

Yes
No

31. If so, would you describe the interaction as negative,
positive, both negative and positive, or neither
negative
or positive?

Negative
Positive
Both negative and
positive
Neither negative or positive

32. How so? [Probe for location and circumstances]
33. Any other ways?
34. Any other ways? [Keep following up until 5
answers, or they run out of answers]

102

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

35. If you had to give me 5 words to describe life in K6G,
what would they be? Why?
36. In general, do you think K6Gs enjoy being in K6G?

Yes
No

[Four modules follow: Community, Sex, Gangs, and Safety. Order of modules to
rotate from interview to interview]
37. COMMUNITY MODULE Do you think anyone
commits a crime just so they can go to K6G?

Yes
No

38. Is K6G better than the outside world for some people? Yes
No
39. In what ways?
40. Anything else?
41. Anything else? [Keep following up until 5 answers,
or they run out of answers. ]
42. Do you think that people who have been in K6G
would be less likely to commit a crime if they knew
they had to go into GP and not back to K6G?

Yes
No

43. Why/Why not?
44. Any other reason? [Keep following up until 5
answers, or they run out of answers. ]
45. Some people who have been in K6G more than once
say that coming back to K6G is like coming back to a
summer camp or a clubhouse, and others say it is just
like any other jail. What do you think?
46. Can you tell me more about how it is like a
[clubhouse or whatever they say it is]?
47. Anything else?
48. Anything else?
49. Think back to your first time in K6G. When you got
there, was anyone there you already knew?
50. [If yes] how many people did you already know?
51. Where did you know them from?
52. [If no to #3 and they’ve been in K6G before] Do
you hang out with your fellow K6Gs on the outside?
OR [if yes to #3 and this is their first time in K6G]
do you think when you get out that you will hang out
with anyone you got to know in K6G?
53. [If yes] Approximately how many?
54. [If yes] Without giving me any names, what do you
do when you hang out with them?

Summer camp
Clubhouse
Like any other jail
Other____________

2011]

STRATEGIC SEGREGATION

103

55. Often, in a large group of people, people tend to
divide up into smaller groups so they can hang out
with people like themselves. Does that happen in
K6G?
56. If so, what are the different groups?
57. How do you feel about the people in the other groups?
58. Do the groups divide up in K6G on racial lines?
59. Are the people you hang out with in any particular
group?

Yes
No

60. [If yes] Which ones?
[SEX MODULE] Now I’m going to ask you about consensual sexual practices in
K6G. Remember that if you tell me anything related to any sexual activity,
consensual or otherwise, between any inmate (whether yourself or anyone else)
and any employee of the L.A. County Jail or any other staff member or service
provider, I will be obliged to report it to Jail officials. In other words, if you tell me
of any such activity, it will be as if you are making an official report of the
incident.422
61. What percentage of K6Gs have no consensual sex at
all while in K6G?
62. What percentage of K6Gs would you say has
consensual sex with other K6Gs every day?
63. What percentage of K6Gs would you say has
consensual sex with other K6Gs a few times a week?
64. What percentage of K6Gs would you say has
consensual sex with other K6Gs once or twice a week,
but not more?
65. How often does any K6G inmate have consensual sex [don’t read options]
with non-K6G prisoners?
every day
3–5 times per week
1–2 times per week
1-2 times per month
Never
66. What percentage of K6G inmates would you say is in
a monogamous sexual relationship?
67. What percentage of K6G inmates would you say has
consensual sex with more than one other person
during their time in K6G?

422. See supra note 296.

[don’t read options]
Everyone; almost everyone;
most people; about half;
less than half; just a few; no
one

104

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

68. How many different sexual partners would you say
the average person in K6G has during their time in
K6G?
69. What time of day does most of the consensual sex in
K6G take place?

[don’t read options]
morning, afternoon,
evening, nighttime (ie after
lights out)

70. Where does most of the consensual sex in K6G take
place?

[don’t read options]
bunks; back of the dorms;
showers
other____________

71. How often do K6Gs trade sex for money?
72. How often do K6Gs trade sex for other things?
73. What other things besides money do people exchange
for sex?
74. When K6Gs trade sex for money or other things, how
often do they keep the money or other things for
themselves?
75. When K6Gs trade sex for money or other things, how
often do they give the money or other things to
someone else?
76. [If sometimes], who usually gets the money or other
things?
77. Is it always the same person or people, or does it
vary?

Same;
Varies

78. Have you ever had consensual sex while incarcerated? Yes
No
79. [If yes to #79] Without telling me any names, was the [ ] A cellmate?
consensual sex with. . ? [read list and check all that [ ] Another prisoner?
apply]
[ ] Someone else?
80. [If yes to #79] Where in the correctional facility have [circle all that apply]:
you had consensual sex?
Cell, Shower, Dorm; Living
Area (other than cell such
as dorm or gym), Day
Room/TV Room, Yard,
Kitchen/ Dining Hall,
Chapel, Hospital/
clinic/infirmary;
Other____________
81. Have you ever had consensual sex while incarcerated Yes
in K6G?
No
82. [If yes to #82] Without telling me any names, was the [check all that apply]
consensual sex with . . . ?
[ ] Another prisoner?
[ ] Someone else?

2011]

STRATEGIC SEGREGATION

105

83. [If yes to #82] Where in the Jail have you had
consensual sex?

[circle all that apply]:
Shower, Dorm, Day
Room/
TV Room, Yard, Kitchen/
Dining Hall, Chapel,
Hospital/clinic/infirmary;
Other____________

84. Have you ever had to do sexual things against your
will with any other inmates in K6G?

Yes
No

85. Please explain.
86. Anything else?
87. Anything else? [Keep following up until 5 answers,
or they run out of answers]
88. [If yes] Did you ever file an incident report regarding Yes
these incidents?
No
89. Have you ever been raped by another inmate in K6G? Yes
No
90. [If yes] Did you ever file an incident report regarding Yes
these incidents?
No
91. Do you think you are safer from being forced to do
sexual things against your will with other inmates in
K6G than you would be in GP?

Yes
No

92. Why/why not?
93. The next few questions deal with your experiences
during prior terms of incarceration. Remind me again
just so I get it right: have you been previously
incarcerated in other prisons, jails, juvenile hall or any
other youth correctional facility?
94. Let me direct your attention to your whole
Yes
incarceration history, including experiences in prison, No
jail, juvenile hall, or any other youth correctional
facility. Have you ever during that time had to do
sexual things against your will with other inmates
while incarcerated?
95. [If yes] how many times?
96. Just to be sure, have any of the following things ever
happened to you with other inmates while
incarcerated: groping or fondling, kissing, genital
contact, oral sex, or penetration against your will?
97. [If yes] How many times?
98. Can you tell me a little about what happened?
99. When these incidents happened, did you file an
incident report?

Yes
No

106

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

100. Have you ever done sexual things while incarcerated Yes
that you were not forced to do, but which you would No
rather not have done?
101. [If yes] About how many times?
102. [If yes] Was it always with the same person?
103. Can you tell me about one or two of these incidents?
104. When these incidents happened, did you file an
incident report?

Yes
No

105. [GANG MODULE] Are there any gang members in Yes
K6G?
No
106. Are there any gang politics in K6G?

Yes
No

107. [If yes] Tell me more. What sort of gang politics?
108. Would you say the gang politics are different in K6G Yes
than in GP?
No
109. If so, how?
110. Any other ways? [Keep following up until 5
answers, or they run out of answers]
111. Is K6G easier time for gang members than GP?

Yes
No

112. If so, how?
113. Any other ways? [Keep following up until 5
answers, or they run out of answers]
114. [If say nothing re putting up a front in answer to
previous question:] Would you say that gang
members have to put up a front in K6G, or are they
able to let down their guard?

Yes
No

115. Why/Why not?
116. Anything else?
117. [SAFETY MODULE] Now I’m going to ask you
some questions about safety. Some people feel that
in certain environments they have to pretend to be
something they are not in order to feel comfortable
or safe. Do you ever feel like that?
118. [If yes] Where?
119. [Hand them 3 3x5” cards reading K6G, General
Population, Out in the Community]: As between
these three places—K6G, GP, and out in the
community—where do you feel most comfortable
and safe being yourself?

Yes
No

2011]

STRATEGIC SEGREGATION

107

120. Of the remaining two places, where do you feel more K6G
comfortable and safe being yourself?
GP
Out in the community
121. Of these three places—K6G, GP, out in the
community—in which one do you feel the safest
from physical harm?

K6G
GP
Out in the community

122. Which one is the next safest?

K6G
GP
Out in the community

123. Of these three places—K6G, GP, out in the
community—in which one do you feel the safest
from sexual harassment?

K6G
GP
Out in the community

124. Which one is the next safest?

K6G
GP
Out in the community

125. How safe do you feel in K6G?

Very safe
Safe
Unsafe
Very unsafe

126. Can you explain? Why do you feel [very safe, safe,
unsafe, very unsafe]?
127. Any other reason? [Keep following up until 5
answers or until they run out of things to say]
128. Have you ever been hit, kicked, punched, or
otherwise assaulted without a weapon in K6G?

Yes
No

129. Have you ever been assaulted with a weapon in
K6G?

Yes
No

130. Do you think you are safer from physical violence in Yes
K6G than you would be in GP?
No
131. Why/why not?
132. Anything else? [Keep following up until 5 answers,
or they run out of answers]
[FINAL MODULE: not to be rotated]
Now I’m going to ask you a few questions about the K6G classification process.
133. Do you think the K6G classification team is
successful at distinguishing the gay prisoners from
those who are not?
134. Is there anything you would do differently if you had
the job of figuring out which prisoners were gay and
which were not?

Yes
No

108

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

Now I’m going to ask you about the condoms given out by The Center for Health
Justice (aka CorrectHelp).
135. How often do K6Gs use the condoms given out once
a week by the Center for Health Justice?
136. If K6G inmates could get more than one condom per Yes
week, would they use them?
No
137. Do K6Gs who have consensual sex without condoms Yes
worry about getting HIV or other sexually
No
transmitted diseases?
138. If there are no more condoms available, how do
K6Gs who are having consensual sex protect
themselves from getting HIV or other sexually
transmitted diseases?
Okay, we’re almost done. Just a few more questions.
139. Where were you born?
140. What year were you born?
141. Race?
142. What is the highest grade you completed in school?
143. Are you currently married, separated, single or
divorced?

Married
Separated
Single
Divorced
Living with someone

144. What gender is your [spouse; partner]?
145. Do you have any children?

Yes
No

146. [If yes] How many?
147. [If yes] What are their ages?
148. Before being incarcerated how often did you have
contact with them?

All of the time
Most of the Time
Occasionally
Rarely
Never

149. Has anyone in your family ever been incarcerated?
150. [If yes] who?

Mother, father
Brothers, sisters
Children
Aunts, uncles

151. Before being incarcerated, did you have any mental
health problems?

Yes
No

152. Since being incarcerated, have you had any mental
health problems?

Yes
No

2011]

STRATEGIC SEGREGATION

153. Before being incarcerated, did you have any
problems with alcohol?

109

Yes
No

154. Since being incarcerated, have you had any problems Yes
with alcohol?
No
155. Before being incarcerated, did you have any
problems with drugs?
156. Since being incarcerated, have you had any problems
with drugs?

Yes
No
Yes
No

157. Are you currently a member of a gang?

Yes
No

158. [If yes to #157] What gang?
159. [If no to #157] Were you ever a member of a gang?
160. [If yes to #157] What gang?
161. When did you leave?
162. How long have you been/were you a member of this
gang?
163. How old were you the first time you were arrested?
164. Where was that?
165. [If outside CA] when was the first time you were
arrested in CA?
166. How old were you the first time you were put in jail?
167. Where was that?
168. [If outside CA] when was the first time you were put
in jail in CA?
169. How many times have you been incarcerated in a
juvenile hall? [if 0, skip next question]:
170. Altogether how much time have you spent in a
juvenile hall?
171. How many times have you been incarcerated in a
CYA (California Youth Authority) facility [if 0, skip
next question]
172. Altogether, about how much time have you spent in
CYA facilities?
173. How many times (including this time) have you been
incarcerated in jail? [if 0, skip next question]
174. Altogether, about how much time have you spent in
jail?

Yes
No

110

AMERICAN CRIMINAL LAW REVIEW

[Vol. 48:1

175. How many times have you been in prison? [if 0, skip
next question]
176. Altogether how much time have you spent in prison?
This is the end of the interview. Thank you very much for your time and willingness to
participate. I won’t tell anyone here what you’ve said and you don’t have to tell anyone
either. At this point, you have the right to ask that the information you just gave me during
this interview not be used for purposes of my research. If you would prefer that your
information not be used, this form will be destroyed and the audiofile of the interview will
be erased. Are you still willing to have your information used for purposes of my research?