Madrid v. Gomez

889 F. Supp. 1146, 1995 U.S. Dist. LEXIS 841, 1995 WL 17092

• Court: District Court, N.D. California
• Decided: January 10, 1995
• Docket: C90-3094-TEH
• Status: Published

Opinion

FINDINGS OF FACT, CONCLUSIONS OF LAW, AND ORDER

THELTON E. HENDERSON, Chief Judge.

TABLE OF CONTENTS

Page

I. Introduction.1155

II. Findings of Fact

A. Excessive Force.1159

B. Medical Health Care .1200

C. Mental Health Care.1214

D. Conditions in the Security Housing Unit.1227

E. Cell Housing Practices. I> CO C<3

F. Segregation of Prison Gang Affiliates. o ^ oq

III.Conclusions of Law

A. Eighth Amendment Overview.1244

B. Excessive Force.1247

C. Medical and Mental Health Care.1255

D. Conditions in the Security Housing Unit.1260

E. Cell Housing Practices.1267

F. Segregation of Prison Gang Affiliates.1270

IV. Summary.1279

V. Appropriate Relief and Further Proceedings.1280

Appendix A (Glossary of Terms).1283

I.

INTRODUCTION

Plaintiffs represent a class of all prisoners who are, or will be, incarcerated by the State of California Department of Corrections at Pelican Bay State Prison, which is located in the remote northwest corner of California, seven miles northeast of Crescent City and 363 miles north of San Francisco. Pursuant to the civil rights statute 42 U.S.C. § 1983, 1 plaintiffs challenge the constitutionality of a broad range of conditions and practices that intimately affect almost every facet of their prison life. They seek redress from the Court in the form of injunctive and declaratory relief.

Although referred to in the singular, Pelican Bay State Prison (“Pelican Bay”) actually consists of three completely separate facilities. The first is a maximum security prison which houses approximately 2,000 “general population” maximum security inmates. The daily routine for these inmates is comparable to that in other maximum security prisons in California. The second is the Security Housing Unit, commonly referred to as the “SHU.” Located in a completely separate complex inside the security perimeter, the SHU has gained a well-deserved reputation as a place which, by design, imposes conditions far harsher than those anywhere else in the California prison system. The roughly 1,000-1,500 inmates confined in the SHU remain isolated in windowless cells for 22 and % hours each day, and are denied access to prison work programs and group exercise yards. Assignment to the SHU is not based on the inmate’s underlying offense; rather, SHU cells are reserved for those inmates in the California prison system who become affiliated with a prison gang or commit serious disciplinary infractions once in prison. They represent, according to a phrase coined by defendants, “the worst of the worst.” Finally, there is a small minimum security facility that houses approximately 200 prisoners. All in all, there are between 3,500 and 3,900 prisoners confined at Pelican Bay on any given day.

Just over five years old, Pelican Bay was activated on December 1, 1989. Considered a “prison of the future,” the buildings are . modern in design, and employ cutting-edge technology and security devices. This, then, is not a case about inadequate or deteriorating physical conditions. There are no rat-infested cells, antiquated buildings, or unsanitary supplies. Rather, plaintiffs contend that behind the newly-minted walls and shiny equipment lies a prison that is coldly indifferent to the limited, but basic and elemental, rights that incarcerated persons — including “the worst of the worst” — retain under the First, Eighth, and Fourteenth amendments of our United States Constitution. In particular, plaintiffs allege that defendants (1) condone a pattern and practice of using excessive force against inmates, (2) fail to provide inmates with adequate medical care, (3) fail to provide inmates with adequate mental health care, (4) impose inhumane conditions in the Security Housing Unit, (5) utilize cell-assignment procedures that expose inmates to an unreasonable risk of assault from other inmates, (6) fail to provide adequate procedural safeguards when segregating prison gang affiliates in the Security Housing Unit, and (7) fail to provide inmates with adequate access to the courts.

Named in their official capacity as defendants are Pelican Bay Warden Charles Marshall, Chief Deputy Warden Terry Peetz, Chief Medical Officer A.M. Astorga, and James Gomez, Director of the California Department of Corrections (“CDC”). 2 They deny that any of plaintiffs’ allegations have merit, and assert that Pelican Bay operates well within constitutional limits in each of the areas outlined above. Moreover, they argue, Pelican Bay, and the SHU in particular, does exactly what it was designed to do: it isolates the most brutal and disruptive elements of the inmate population while reducing violence in California state prisons overall.

The case was tried before the Court between September 14 and December 1, 1993. Immediately prior to the trial, the Court spent two days touring Pelican Bay, accompanied by counsel for both parties and prison officials. During the course of the trial, the Court heard testimony from 57 lay witnesses, including class members, defendants, and correctional employees at all levels. It also received into evidence over 6,000 exhibits, including documents, tape recordings, and photographs, as well as thousands of pages of deposition excerpts. 3

The Court recognizes that neither the inmates at Pelican Bay nor the Department of Corrections personnel can be considered neutral witnesses. For reasons that are self-evident, class members, as well as defendants and other prison staff, are interested in the outcome of the case. We also take into account the undeniable presence of a “code of silence” at Pelican Bay. As the evidence clearly shows, this unwritten but widely understood code is designed to encourage prison employees to remain silent regarding the improper behavior of their fellow employees, particularly where excessive force has been alleged. Those who defy the code risk retaliation and harassment. 4 We have considered all of the above, as well as the manner and demeanor of the witnesses, in assessing witness credibility and making our factual findings.

The Court was also aided by the testimony of ten experts in the areas of medicine, psychiatry, psychology, and prison management and operation. 5 With respect to the claims regarding excessive force and cell assignment practices, plaintiffs presented three experts: Charles Fenton, a former warden of two maximum security prisons, 6 Steve Martin, who spent more than 20 years working in varying capacities for the Texas Department of Corrections, 7 and Vince Nathan, who has worked for nearly 20 years as a court-appointed monitor and expert in prison cases. 8 Defendants presented two experts: Daniel McCarthy, who worked for almost 40 years in varying capacities for the California Department of Corrections, most recently as its Director, 9 and Larry DuBois, the Commissioner of the Massachusetts Department of Corrections. 10

With respect to the claims concerning medical care, mental health care, and conditions in the SHU, plaintiffs presented Dr. Armond Start, an associate professor at the University of Wisconsin Medical School and former director of health care services for the Oklahoma and Texas prison systems, 11 Dr. Stuart Grassian, a psychiatrist and faculty member at Harvard Medical School and expert on the effects of solitary confinement, 12 and Dr. Craig Haney, a professor of psychology at the University of California at Santa Cruz, who has specialized in the psychological effects of incarceration. 13 Defen dants presented Dr. Jay Harness, a professor of surgery at the University of California at Davis and former director of health care services for the Michigan prison system, 14 and Dr. Joel Dvoskin, a clinical psychologist and director of the Bureau of Forensic Services for the New York State Office of Mental Health. 15

We are mindful that the opinions of experts are entitled to little weight in determining whether a condition is “cruel and unusual punishment” under the Eighth Amendment. Toussaint v. McCarthy (Toussaint IV), 801 F.2d 1080, 1107 n. 28 (9th Cir.1986). As such, we have not relied upon expert opinion to make this ultimate legal determination. It is appropriate, however, for this Court to consider expert opinion in assessing subsidiary issues which inform the court’s final determination. For example, expert opinion may be properly considered in assessing the effects of challenged conditions or practices. See Helling v. McKinney, — U.S. —, —, 113 S.Ct. 2475, 2482, 125 L.Ed.2d 22 (1993) (making reference to the “scientific and statistical inquiries]” that will be used to determine the seriousness of the harm caused by challenged conditions); Jordan v. Gardner, 986 F.2d 1521, 1526 (9th Cir.1993) (en banc) (relying on expert testimony to establish psychological impact of challenged measure on inmates). See also Slakan v. Porter, 737 F.2d 368, 378 (4th Cir.1984) (correctional expert’s opinions concerning punitive nature of prison’s water hosing practices properly admitted).

After the trial was completed, in December 1993, the parties filed proposed findings of fact and conclusions of law on January 28 and February 1, 1994. The case was taken under submission at that time.

II.

FINDINGS OF FACT

A. EXCESSIVE FORCE

Perhaps the paramount responsibility of prison administrators is to maintain the safety and security of both staff and inmates. Bell v. Wolfish, 441 U.S. 520, 546-47, 99 S.Ct. 1861, 1878, 60 L.Ed.2d 447 (1979). In the setting of a maximum security prison, where inmates are more likely to pose security risks, this is a remarkably difficult undertaking. As the Ninth Circuit has previously said, prison officials have the “unenviable task of keeping dangerous men in safe custody under humane conditions.” Spain v. Procunier, 600 F.2d 189, 193 (9th Cir.1979). In a place like the SHU, which houses some of the most anti-social and violence-prone prisoners in the system — including those who suffer from mental illness — the task is that much more difficult.

There is no question that this demanding and often thankless undertaking will require prison staff to use force against inmates. Indeed, the responsible deployment of force is not only justifiable on many occasions, but absolutely necessary to maintain the security of the institution. As one expert at trial succinctly stated, when it comes to force, it is “as dangerous to use too little as it is to use too much.” Fenton Tr. 6-766.

At the same time, the prison setting offers a tremendous potential for abuse. Custody personnel are in constant contact, day after day, with a difficult, frustrating, and sometimes openly and actively hostile inmate population. They also have powerful weapons and enormous manpower at their disposal, and exercise nearly total control over the inmates under their supervision. Adding to this volatile mix is the fact that the prison setting, and particularly the SHU, is far removed from the usual sights and sounds of everyday life. From the outside, the SHU resembles a massive concrete bunker; from the inside it is a windowless labyrinth of cells and halls, sealed off from the outside world by walls, gates, and guards. The physical environment thus reinforces a sense of isolation and detachment from the outside world, and helps create a palpable distance from ordinary compunctions, inhibitions and community norms.

If, in addition to all of the above, prison administrators fail to adequately supervise and monitor the use of force, the potential that force will be misused increases significantly. See Haney Decl. at 23-24. At trial, plaintiffs sought to prove that this potential for abuse was in fact realized at Pelican Bay, leaving in its wake a pattern of excessive force against inmates.

A substantial portion of the trial was devoted to this claim. The parties presented testimony from dozens of witnesses, including several inmates and CDC personnel at all levels. The documentary evidence presented included incident reports, Internal Affairs reports and investigative files and tapes. Several hundred pages of deposition testimony were also admitted into evidence.

As described above, the parties also presented five experts, Charles Fenton, Vince Nathan and Steve Martin (for plaintiffs) and Daniel McCarthy and Larry DuBois (for defendants), all of whom are distinguished experts in the correctional field. There were, however, significant differences in the amount of preparation they undertook prior to testifying. On the whole, plaintiffs’ experts did substantially more to familiarize themselves with the particulars of Pelican Bay than did defendants’ experts. For example, plaintiffs’ expert Martin reviewed over one thousand documents, including all incident reports for each cell extraction and shooting that occurred at Pelican Bay through the end of discovery, all fetal restraint memoranda and Internal Affairs investigation files produced (including a number of related audio tapes), training materials, rules violations reports, grievance appeals submitted by prisoners, and 30 deposition transcripts. He also spent several days at Pelican Bay, from September 14-17, 1992, and July 7-9, 1993, during which time he toured the prison, conducted interviews with inmates, and met with correctional staff. 16 In contrast, defendants’ expert Larry DuBois testified that he spent approximately 20 hours reading documentary evidence; in his own words, he “reviewed or cursorily re viewed” about a box and a half of documents relevant to the case (incident reports, training materials, Internal Affairs reports and policies). Tr. 29-4689. He also spent two days at Pelican Bay in October 1993, during which he met with the Warden and correctional staff, and observed demonstrations of various weapons at the prison’s firing range.

After a thorough review and consideration of the testimonial and documentary evidence, the Court is compelled to conclude that the Eighth Amendment’s restraint on using excessive force has been repeatedly violated at Pelican Bay, leading to a conspicuous pattern of excessive force. In many instances, there was either no justification for the use of force, or alternately, the use of force was appropriate, but the amount of force applied was so strikingly disproportionate to the circumstances that it was imposed, more likely than not for the very purpose of causing harm, rather than in a good faith effort to restore or maintain order. Although this pattern was probably more pronounced during the initial years of the prison’s operation (and prior to the prosecution of the instant class action), the Court is satisfied that it continues to exist. 17

Plaintiffs’ experts forcefully opined that the level of force used at Pelican Bay is well beyond the norm of any facility with which they are familiar. Nathan, for example, testified that “Pelican Bay State Prison exists in a very different universe.... [I]n 18 years of involvement with a number of the most repressive and unlawful prisons in the United States, I have never observed ... the level of officially sanctioned unnecessary and excessive force that exists at [Pelican Bay].” Nathan Deck at 12. 18

As previously noted, the risk that force will be misused is considerably enhanced when prison administrators fail to implement adequate systems to regulate and monitor its use. Plaintiffs have demonstrated that such a failure occurred at Pelican Bay, and that it substantially contributed to the development and persistence of the pattern of excessive force. Finally, the evidence shows that the pattern of excessive force, and the lack of adequate systems to control it, are not simply the result of inadvertence, genuine mistakes in judgment, or good faith efforts gone awry. Rather, they are attributable — not only to defendants’ deliberate indifference— but also their knowing willingness that harm occur. The Court agrees with Nathan’s observation that “the use of unnecessary and excessive force at [Pelican Bay] appears to be open, acknowledged, tolerated, and sometimes expressly approved.” Nathan Deck at 12.

We divide our factual findings concerning the use of force at Pelican Bay into three parts: (1) findings regarding the pattern of excessive force at Pelican Bay, (2) findings regarding the lack of adequate systems to regulate and control the use of force, and (3) findings regarding defendants’ state of mind.

1. Pattern of Excessive Force

a. Sampling of Evidence of Use of Excessive Force

The evidence pertaining to excessive force was not limited to one area of prison life. Rather, the record shows that excessive force was used in a variety of circumstances and settings, from staff assaults on inmates to punitive cagings under harsh conditions. To gether, these strands of evidence weave a picture that reflects not just isolated indiscretions, but a pattern and practice of excessive force. The Court does not attempt to address every incident of excessive force that was raised at trial, in plaintiffs’ experts’ declarations, or in the documentary evidence; rather what follows is a sampling of the evidence concerning the use of excessive force in different contexts. 19

(1) Staff Assaults on Inmates

Inmate Castillo

On January 31, 1991, Arturo Castillo refused to return his food tray in protest against a correctional officer (“officer”) who had called him and other inmates derogatory names. After leaving the tray near the front of the cell, Castillo retreated to the back and covered himself with his mattress for protection, in anticipation of a cell extraction. 20 It is undisputed that Castillo, who is small in stature, made no verbal threats or aggressive gestures. Nor did he possess, or pretend to possess, any kind of weapon. Shortly thereafter, Sergeant Avila- warned Castillo that if he did not give up his food tray, it was going to be very painful. Castillo refused to hand Avila the tray, stating that if they wanted the tray, they would have to come and get it. The supervising lieutenant then authorized his sergeants to forcibly remove Castillo from the cell.

To accomplish this removal, two rounds from a 38 millimeter gas gun were fired into the cell. A taser gun was also fired, striking Castillo in the chest and stomach. 21 Then, without attempting to retrieve the tray (which remained near the front of the cell), some number of officers entered the cell, walked past the tray, and advanced toward Castillo. Castillo testified that one of the officers then hit him on the top of his head with the butt of the gas gun, knocking him unconscious. When he regained consciousness, he was on the floor with his face down. An officer was stepping on his hands and hitting him on his calves with a baton, at which point Castillo passed out a second time. When he regained consciousness again, he was dragged out of the cell face down; his head was bleeding, and a piece of his scalp had been detached or peeled back. At that point, it became clear that Castillo had been seriously injured, and he was taken to the infirmary and then to the hospital by ambulance. 22

According to the incident report, Castillo sustained his head injury when he fell and accidentally hit his head on the toilet during the incident. Trial Exh. P-1100 at 4099. Lieutenant Trujillo, who was present at the time, also testified that he saw Castillo “falling forward” and heard a “loud bang” and “somebody saying that he hit the toilet.” Tr. 21-3638.

We do not, however, find defendants’ explanation of the injury credible. First, Trujillo’s testimony loses much of its force since he never actually saw Castillo’s head hit the toilet even though he was “looking into the cell during the entire cell extraction.” Tr. 21-3638-41, 3667-68. Nor did he recall seeing any blood on the toilet. Second, Castillo’s credible testimony was unequivocally corroborated by Sergeant Cox, who observed the entire episode. Cox, who Trujillo admits “had a clear view” of the extraction, Tr. 21-3646, testified that he witnessed another sergeant “hit [Castillo] in the head with a 38 millimeter gun, by the butt of the gun.” Tr. 15-2330. Cox further testified that he was “basically ... ordered to keep my mouth shut and leave the area.” Tr. 15-2334. In addition, plaintiffs’ medical expert, Dr. Ar-mond Start, gave unrebutted testimony that Castillo’s head laceration was more likely the result of a high-velocity accelerated blow to the head than of a collision with a blunt, stationary object, particularly given that the injury occurred on the top of the head. Start Deel. at 242.

The record contains no evidence that would support the conclusion that striking Castillo on the head with a gun, with enough force to render him unconscious, was needed to retrieve the food tray, restore order, or otherwise protect the integrity of the institutional mission. Indeed, the fact that the supervising officer made no attempt to ascertain whether the food tray could simply be taken from the front of the cell, but instead immediately resorted to gas guns and tasers, reflects a pattern of using the maximum, rather than the minimum, amount of force necessary to accomplish a goal. Finally, the fact that officers continued to beat Castillo after he was subdued and unconscious further supports the Court’s inescapable conclusion that Castillo was subjected to the use of excessive force that was imposed, not in a good faith effort to restore order or maintain security, but maliciously for the purpose of causing pain and inflicting punishment. 23

Inmate Richard

Richard, a general population prisoner, was working in the prison’s optical lab in October 1991 when a disturbance erupted between some of the other inmates in the lab. There is no dispute that Richard, of slight build, was a victim rather than an aggressor in this incident. He received some minor injuries from another inmate, including a laceration on his left cheek from a scissors, screwdriver or similar instrument. It is undisputed that Richard did not assault any staff in this incident.

When he was subsequently placed in a holding cage, an officer informed him that he was rumored to have assaulted another officer during the incident. Richard denied the rumor and told him that other staff could verify that he was only defending himself from attack. Shortly thereafter, Officer Bray entered Richard’s cell, grabbed his handcuffs, and took him to an adjacent counselor’s office. There, he shoved Richard into a large table, on which Richard struck his face. Richard testified that while he was bent over the table, with his hands cuffed behind his back, Bray repeatedly struck his head, particularly on the side of the face where he already had the laceration. It is undisputed that Richard was not resisting at this time.

Bray then dragged Richard onto the floor where Bray continued the assault while another officer held Richard’s legs. Richard testified that, despite his repeated denials that he had attacked anyone, Bray continued punching him about the head and neck, and continued to assert that Richard had assaulted another officer (who also happened to be a friend of Bray’s). At one point, blood started shooting out of Richard’s mouth, but the punches continued. He was then taken back to the holding cell, after the blood was washed from his face. During this incident, other officers watched but made no effort to intervene. After experiencing two days of intense pain and inability to chew, Richard was allowed to go to the infirmary, and he was subsequently informed that he had a fractured jaw. Five days later, he was taken to a hospital, and he spent the next six weeks with a wired jaw in the Pelican Bay infirmary.

Prison officials later suggested that Richard had fractured his jaw in the optical lab incident and not at the hands of Officer Bray. The Court again finds that this version of events entirely lacks credibility. First, immediately following the lab incident, Richard was taken to the infirmary and given an entire body check for wounds, which included an examination of the inside of his mouth; however, the medical report indicates only that Richard sustained a laceration on his cheek, a scrape on the top of his left thigh and a small abrasion on his right elbow. Trial Exh. P-3084 at 79958. Second, the prison’s Chief Medical Officer later expressed the opinion to investigators that an attack with scissors or other similar instrument, like the one suffered by Richard in the optical lab, would not likely have caused a fractured jaw. Trial Exh. P-3084 at 79885. Further, Captain Jenkins testified that, in the course of investigating the incident, the officers involved made inconsistent statements which led him to believe that they were not being truthful about the incident. Tr. 3-360-61. He expressed concern that he was facing a “code of silence” problem, particularly because he felt that the correctional officers’ union was suggesting that the officers “play their cards pretty close to the vest.” Tr. 3-361-2.

Based on the above, the Court finds that Richard was beaten about his head and neck, and suffered a broken jaw, as punishment for an officer’s belief that Richard had assaulted another member of the staff. 24 Given the severity of the injury, and the undisputed evidence that Richard neither provoked nor resisted Bray, the Court finds that Bray’s use of force was not only excessive, but also completely unnecessary and inflicted for the purpose of causing pain rather than in a good faith effort to restore order or maintain security.

Inmate Martinez

This incident was precipitated when Martinez spit at Officer Parson and threw his meal tray through the narrow food port, striking Parson. A short time later, correctional officers returned and ordered Martinez to submit to mechanical restraints before being removed from his cell. When Martinez refused, and shielded himself with a mattress and/or blanket, the officers fired tear gas and nine 38 millimeter gas gun rounds. When this did not subdue Martinez because of the shields, a team of officers, including Parson, 25 then entered the cell. Although Martinez was combative, kicking and striking with fists, he was ultimately subdued.

Louie Lopez, another inmate, was able to observe subsequent events. He credibly testified that he observed Martinez, then handcuffed, emerging from the cell and being thrown against the wall. At that point, “he was out [i.e. seemingly unconscious]. He wasn’t moving.” Tr. 1-68. Lopez then observed officers kicking Martinez’ head, face, neck and shoulders, and saw a lot of blood. One of the control booth officers on duty also told the investigating officer that once Martinez was out of the cell, he was restrained and under control, yet an officer whom the booth officer could not identify had kicked Martinez in his upper shoulder or lower head area, after which someone said, “that’s enough, knock it off!” Trial Exh. P-8083 at 79053. He also confirmed that other inmates on the tier could see what was happening and that they were yelling, “you can’t kick a guy in the head like that.” Trial Exh. P-3083 at 79052. As a result of this incident, Martinez lost four teeth, received a 1.5 inch laceration to the back of his head, and suffered abrasions to the head, face, back, neck, chest and both legs.

The Court finds that the level of force used against Martinez was not motivated by a good faith effort to restore order or maintain the security of the prison. Rather, the extraction was prompted by the desire to punish or retaliate against Martinez for spitting on an officer and striking him with a food tray. Moreover, even assuming that Martinez strenuously resisted during the initial part of the extraction, this furnishes no justification for the continued use of force after Martinez was subdued and no longer resisting. Such use of completely gratuitous force evidences a malicious intent to inflict injury rather than a good faith effort to restore order.

Inmate Ward

Kenneth Ward testified that one morning he refused lunch, which led to a heated verbal exchange with a female correctional officer (Officer Reynoso). Ward, who used profanity during the exchange, continued to be verbally abusive to Reynoso at other points during the day. The next morning, Ward was awakened by Officer Kelly who had returned with Reynoso. Kelly told Ward he was being moved. Ward did not at first understand what Kelly had said, so he stood up. He saw Kelly standing with a “smile on his face, and [saying] ‘you like disrespecting my old lady, huh?”’ Ward swore at both officers. Kelly responded that he would move Ward to a cell “in [Kelly’s] block” with “somebody your [own] size.” Ward Tr. 2-219.

When Ward protested, Kelly returned with Sergeant Rowland and three other officers. Rowland threatened Ward with a cell extraction and ordered him to “cuff up.” Ward Tr. 2-220. Ward eventually turned his back to the cell door and put his hands through the small food slot (approximately four to five inches high and 12 inches long) to be cuffed up. Ward’s trial testimony and the officers’ deposition testimony about what happened next dramatically conflict.

According to Ward, one of the officers grabbed his arm and twisted his wrist upward, causing Ward enormous pain. When he couldn’t take the pain any longer, he jerked his hands back inside the cell and accused them of trying to break his wrist, to which an officer again responded with a threat of a cell extraction. Ward then again placed his hands through the food port, at which point Kelly immediately jerked his left arm through the small slot all the way up to his tricep while Rowland grabbed the other arm. Ward then pulled his right hand back into the cell and bent down to try ,and ease the pain. At this point, Kelly repeatedly threw his body weight against Ward’s left arm. It is undisputed that as a result, Ward’s upper left arm snapped and broke. Ward immediately became lightheaded and he felt a lot of pain. He recalls that he felt as if he were in shock and was unable to move. As a result of this injury, Ward has suffered recurring problems with nerves in arms, including numbness and spasms.

Officers Rowland and Kelly did not testify at trial, but the Court has been provided with their deposition transcripts. Officer Rowland does not dispute that he heard a loud snap, after which Ward stopped struggling immediately, although Officer Kelly, who actually broke Ward’s arm, denies hearing any snap. Both officers do, however, state that the amount of force used was necessary and appropriate because Ward was resisting being cuffed up, and was attempting to pull an officers’ arm into the cell. Officer Kelly testified that when he tried to cuff Ward up the first time, Ward pulled on his thumb, which resulted in a sprain to his thumb and his wrist and some abrasions. Officer Rowland stated that it was better to gain control of the inmate in the manner they did than to have to send in a cell extraction team.

Even assuming that Ward was resisting in some fashion, we are convinced, by the weight of the evidence, that the force used was employed, at least in substantial part, for the purpose of inflicting pain and not for good faith security-related reasons. Several factors inform this finding. First, it is un-contradicted that the sheer amount of force required to break Ward’s arm was “enormous,” in the words of plaintiffs medical expert Dr. Armond Start. The Court itself observed that Ward has a well-developed physique. The sheer amount of force that would be needed to break Ward’s arm suggests that excessive force was used.

The context also suggests that such force was probably unnecessary. The officers were not facing the threat of any immediate, serious injury, given that Ward was confined inside his cell, with his back to the cell door and his hands through a narrow food port. Even if Ward offered some struggle, there is no indication that the officers made any effort to end the incident by less violent means. In short, there is no convincing basis for concluding that the degree of force used was necessary to protect either the officers involved or any other person.

There are certainly instances where, in the heat of the moment, officers may use more force than intended. Such a case would suggest that the officers did not act with punishment in mind. Here, however, the manner in which Kelly and Rowland executed the cuff-up procedure — severely twisting Ward’s arm and using enormous pressure — and their openly hostile attitude demonstrate an intent to do Ward some harm. Thus, although they may not have known that the precise level of force used would be sufficient to break Ward’s arm, they clearly used unnecessary force with a knowing willingness that harm occur. It is also significant that the circumstances leading up to the incident provided Kelly with a motive for retaliating against Ward. All too often, the evidence showed that inmates suffered serious injuries after providing an officer with some provocation, such as spitting, name calling, or refusal to obey a minor order.

Inmate Cooper

In this incident, an officer struck inmate Cooper twice with a closed fist in the head at a time when he was offering no resistance. Cooper was being escorted, in handcuffs and leg irons, by several officers from the Facility D Program hallway. When Cooper began resisting and kicking, Officers Plumlee and Bettencourt leaned against Cooper’s legs and forced him against the wall. At this point, two other officers both observed that Cooper was under control and unable to kick, but that Officer Plumlee nonetheless proceeded to punch Cooper twice in the face with a closed fist. An internal investigation concluded that Plumlee had hit Cooper twice on the left side of his face with a closed fist, which is also consistent with a subsequent medical report, which indicated contusions to the left chin, left cheek, and mid forehead, a black eye and a small cut above the eye. Trial Exh. P-3087 at 77542. Although Officer Plumlee defended his punches as being simply “reflexive,” the Court finds that the record, overall, does not support his position, given the eyewitness reports of the other officers and the fact that there were two punches thrown. We further find that punching Cooper at a time when he was restrained and under control constituted a gratuitous and unnecessary use of force that was imposed not for any security related purpose but for the purpose of inflicting pain.

Inmate Dortch

Vaughn Dortch, a mentally ill inmate, suffered second-and third-degree burns over one-third of his body when he was given a bath in scalding water in the prison infirmary. The week before the incident Dortch bit an officer. Dortch had also created a nuisance by smearing himself and his cell with his own fecal matter. Although there was a shower near Dortch’s cell, which would have provided a more efficient method of cleaning Dortch than a bath (even assuming Dortch was uncooperative), the officers instead forcibly escorted Dortch to a bathtub in the SHU infirmary, located some distance away in an other complex. 26

According to Barbara Kuroda, the nurse on duty at the infirmary, a Medical Technical Assistant arrived shortly before Dortch, and was asked if he “want[ed] part of this bath,” to which he responded, yes, he would take some of the “brush end,” referring to a hard-bristle brush which is wrapped in a towel and used to clean an inmate. Tr. 1-144. Five or six correctional officers then arrived with Dortch. Although a nurse would normally run the water for a therapeutic bath, Dortch’s bath was managed solely by correctional staff.

Kuroda later observed, from her nurse’s station, that Dortch was in the bathtub with his hands cuffed behind his back, with an officer pushing down on his shoulder and holding his arms in place. Subsequently, another officer came into the nurse’s station and made a call. Kuroda’s unrebutted testimony is that she overheard the officer say about Dortch, who is African-American, that it “looks like we’re going to have a white boy before this is through, that his skin is so dirty and so rotten, it’s all fallen off.” Tr. 1-154. Concerned by this .remark, Kuroda walked over toward the tub, and saw Dortch standing with his back to her. She testified that, from just below the buttocks down, his skin had peeled off and was hanging in large clumps around his legs, which had turned white with some redness. Even then, in a shocking show of indifference, the officers made no effort to seek any medical assistance or advice. Instead, it appeared to Kuroda that the officers were simply dressing Dortch to return him to his cell. When Kuroda told them they could not return him in that condition, Officer Williams responded, in a manner described by Kuroda as disparaging and challenging, that Dortch had been living in his own feces and urine for three months, and if he was going to get infected, he would have been already. Williams added, however, that if Kuroda wanted to admit him, she could do the paperwork. Dortch then either fell, or began falling, to the floor from weakness, at which point Kuroda had Dortch taken to the emergency room. Although Dortch was not evidencing any pain at this point, Kuroda testified that this did not surprise her. Because severe burns destroy the surrounding nerve endings, the victim does not experience any pain until the nerves began to mend. Dortch was ultimately transported to a hospital bum center for treatment.

Based on the record before us, we can not say that any of the staff involved in the incident specifically intended the severity of the burns inflicted upon Dortch. It is unclear whether the officers knew the actual temperature of the water or the full extent of the burns that were being inflicted. 27 Nor did Dortch yell out in pain to alert the officers. 28 On the other hand, officers were observed holding Dortch down in the tub, and the burns he was experiencing must have been visible.

Although we assume, for purposes of this ease, that those involved did not intend to inflict third-degree burns, it is nonetheless clear, from all of the surrounding circumstances, that Dortch was given the bath primarily as a punitive measure and for the purpose of inflicting some degree of pain, in retaliation for, and perhaps out of frustration with, his prior offensive conduct.

Inmate Brown

In this incident, John Brown was taken from his cell in full restraints when a staff member observed, and later reported to in ternal investigators, that Brown was kicked in the face by another officer in the presence of a Lieutenant, Sergeant, and acting Program Administrator. There is no evidence that he was struggling or resisting at the time. 29 In an interview with investigators, Brown also stated that he had been kicked in the head three or four times while being carried in restraints. Trial Exh. P-3085 at 776601-02. The medical report, which is consistent with being kicked in the face, notes an abrasion on the right side of his face and a % inch cut on his chin. Id. at 77638. The fact that Brown was kicked in the face for no apparent reason, while he was fully restrained, leads us to conclude that this force was applied maliciously for the purpose of causing harm rather than to restore or maintain security.

(2) Use of Fetal Restraints

The fetal restraint, also known as “restraint control status” or “hog-tying,” is a particular type of in-cell restraint. Utilized numerous times until late 1991 or mid 1992, the fetal restraint procedure involves handcuffing an inmate’s hands at the front of his body, placing him in leg irons, and then drawing a chain between the handcuffs and legs until only a few inches separate the bound wrists and ankles. At least one officer, however, handcuffed inmates in the back, so that the inmate’s arms were behind his back and his ankles were up around his handcuffs. The fetal restraint was applied most commonly in response to an inmate kicking his cell door, although it was utilized on other occasions as well.

Plaintiffs’ medical expert confirmed inmate testimony that being in this position without the ability to stretch one’s legs or arms would, over time, likely cause considerable pain, 30 and could pose a serious health risk to inmates with respiratory ailments. Thus, unlike four- or five-point restraints, which completely prevent any disruptive movement without imposing pain or health risks, fetal restraints can inflict significant pain and yet not fully secure the inmate.

Notably, no expert at trial defended the use of fetal restraints. Plaintiffs’ expert described such restraints as a painful, repugnant, humiliating punishment, and termed their level of use at Pelican Bay “unprecedented” in modem corrections. 31 Martin Tr. 18-1351 Defendants’ expert Daniel McCarthy testified that he had never previously used or seen anyone use a fetal restraint in his forty years in the California Department of Corrections, and did not believe that it would be an acceptable technique. Defendants’ other expert, Larry DuBois, also stated that he had never used a fetal restraint and that he had not been asked to express an opinion regarding its use.

The testimony of Mark Jones, which was not refuted at trial, provides but one example of an inmate’s experience with fetal restraints at Pelican Bay. On one occasion, Jones repeatedly kicked his cell door to get the attention of the control booth officer, in an effort to get him to close the door to the exercise pod, which was letting in cold air. In response, he was put in fetal restraints for five hours. Another time, Jones attempted to call up to the control booth officer to find out why the television in his cell was not working, but was ignored. When he tried again later that day, and was still ignored, he started kicking the cell door to get attention. About 10 or 15 minutes later, a sergeant and a few more officers went to his cell. Upon admitting that he had been kicking the door, he was put in his bunk in fetal restraint chains for approximately nine hours, from 1:00 p.m. to about 10:00 p.m. He was in pain after a while but eventually fell asleep. When he awoke, parts of his body were asleep. He attempted to spin around, but, unable to control his movements, fell off his bunk, hitting his back, head and shoulder on the floor. 32 Officers who later' arrived to deliver Jones his regularly-scheduled medicine refused to remove the restraints, instead taunting him that he should “get up and get [his medicine].” In response, Jones got “kind of verbal.” Jones Tr. 3-521. Approximately one hour later, the fetal restraints were removed by another officer.

On another occasion, following a verbal altercation between Jones and an officer over a book that had been confiscated, two officers handcuffed Jones and escorted him down the hallway, pulling him sideways at times. When Jones complained, one of the officers slammed Jones’ head into the wall, chipping his teeth. Jones was then taken to a holding cell where he was put in fetal restraints for four horn’s. In none of these instances is there any evidence that the fetal restraints, particularly for periods of four to nine hours, were necessary for security reasons. Rather, their usage evinces an intent to punish and inflict pain.

Inmates restrained in fetal restraints were at times also chained to toilets or other fixed objects, particularly during program administrator Rippetoe’s tenure in the SHU. Although there is directly conflicting testimony regarding the extent of this practice, the Court finds that it was more than merely an occasional occurrence. Sergeant Cox testified credibly that during the limited period that he was working overtime in the SHU, he personally observed, over the course of different shifts, ten to twelve inmates who were in fetal restraints and chained to stationary objects. Some staff, including Sergeant Cox, objected to this practice, and one SHU program administrator testified that he “never ha[s] and never would” authorize such a practice. Lopez Tr. 14-2196. Sergeant Cox testified, however, that his objections were dismissed out of hand. When he asked Rip-petoe why they were engaging in this practice, he responded “because we can do it.” Tr. 15-2345. When he raised the issue with Associate Warden Garcia, he was told “this is Pelican Bay State Prison, and if you don’t like it, get out '... [W]e’re going to do it our way.” Tr. 15-2347.

Prison records indicate that fetal restraints were used in dozens of instances between January 1990 and August 1992. 33 Such restraints were imposed from anywhere between a few minutes to 24 hours, -with most instances falling in the three to six hour range. Current written SHU policies permit use of fetal restraints for up to 12.hours without obtaining consent of the Warden. Whether an inmate would be freed from the restraints in order to eat or use the toilet was left to the discretion of individual staff.

At trial, defendants did not attempt to justify the fetal restraint as an appropriate response to the kicking of cell doors per se, although current SHU policy permits the use of fetal restraints for cell door kicking. Trial Exh. D-49 at 18307. 34 Indeed, although cell door kicking is a common occurrence, no lay or expert witness was able to identify any other prison that resorted to fetal restraints in response.

Rather, defendants testified that the use of fetal restraints for kicking doors was necessary because a weakness in the metal in the door retainer at the bottom of each cell made cell doors vulnerable to metal fatigue in the event of continued kicking. Various prison officials testified that they became aware of this problem when the prison opened (in December 1989), and that the practice of using fetal restraints stopped once the doors were fixed in December 1991.

Integrity of cell doors is, of course, a critical security concern. However, we are not persuaded by the record that repairs to address the metal fatigue problem were not undertaken until late 1991. Although defendant Peetz testified that repair work on the cell doors occurred between approximately June and December 1991, it appears that these repairs primarily concerned another door problem involving the pneumatic locking mechanism, which was not discovered until sometime in 1991.

Prison officials were extremely vague as to what was done and when to respond specifically ' to the metal fatigue problem. One SHU program administrator testified that work was done to reinforce the holding cells in the SHU (where inmates kicking their cells could be placed), and that this work probably took a “week or so.” Lopez Tr. 14-2199. He did not identify any time frame for this work. It is highly unlikely, however, that it was not done until 1991. As stated above, prison officials discovered the metal fatigue problem as early as December 1989 or January 1990. It is not credible that they would have waited until December 1991, two years later, to reinforce the holding cells, if in fact, metal fatigue presented a serious security concern. 35

More fundamentally, however, we are not persuaded that the use of fetal restraints was necessary or primarily prompted by legitimate penological purposes. The “D” wing of the SHU and the “C” wing of - the SHU suffered from the same metal defect, and housed roughly the same number of inmates (physically the two wings are mirror images of each other). Yet, during the period that inmates in C-SHU were fetally restrained on dozens of occasions, fetal restraints were used in D-SHU less than 5 times. Given that defendants offered no basis for concluding that the inmates in D-SHU would be less likely to kick their cell doors than in C-SHU, this discrepancy in numbers is a compelling indication that the utilization" of fetal restraints was not necessary to maintain security.

Moreover, fetal restraints did not even effectively prevent continued kicking of cell doors. As Captain Scribner agreed, it was “clear that if an inmate wanted to kick the door while he had his [fetal restraint] chains on ... he could do so,” and in fact it appeared to him that an inmate might end up kicking the door even harder after he was restrained. Scribner also testified that such an inmate could still get “enough of a kick” that he could compromise the integrity of the door. Scribner Tr. 7-1241, 6-1120. Indeed, it is clear that the continued ability to kick only escalated the need for more restraints, such that there was a practice, albeit relatively short-lived, of locking an already fetally-restrained inmate to his toilet — a practice which Lieutenant Carl agreed was in violation of governing California regulations. Carl Depo. at 291-92.

The use of fetal restraints in response to the kicking of doors becomes particularly suspect when it is considered that a far more effective, yet less painful, alternative was available. As plaintiffs’ expert Steve Martin emphasized without contradiction, from a custody standpoint, the most effective means of preventing an inmate from kicking a cell door is to place him in full restraints: “you put a man in a four-point, five-point restraint, he’s not going to kick a door.” Tr. 8-1350. And, as noted above, full restraints accomplish this result without inflicting discomfort and pain. Nonetheless, there is no indication in the record that defendants ever considered full restraints as an alternative at any point between December 1989 and December 1991.

In short, it is undisputed that the fetal restraint, which requires an obvious eontor tion of normal body position, can cause considerable pain over time. It is also undisputed that this restraint failed to effectively prevent the kicking of doors, and that other more effective and less painful alternatives were available and known to defendants. Nonetheless, in one section of the SHU, fetal restraints were often the response of first resort to cell door kicking for a period of at least two years.

We do not, and need not, find that every application of the fetal restraint at Pelican Bay was punitive in nature. Nor do we address the facial validity of the prison’s fetal restraint rule. However, the record and particular circumstances presented here convince us that there was a practice of using fetal restraints at Pelican Bay for solely punitive rather than good faith security purposes.

(3) Cagings

Another use of force at Pelican Bay that is punitive in character is the confinement of naked or partially dressed inmates in outdoor holding cages during inclement weather. These cages, approximately the size of a telephone booth, and constructed of weave mesh metal, are designed to provide a temporary holding place for an inmate, and are positioned at various locations around the prison. Inmates confined in the cages are exposed to the elements as well as public view.

Violet Baker, a former educational program supervisor at Pelican Bay, gave a frank and credible account of one such incident. She testified that one day in late January or early February, she was walking from her office toward another facility. It was very cold (she was wearing gloves and a heavy jacket), and it was pouring rain. She observed two African-American inmates being held naked in two cages. When she passed by again one hour later, one inmate was still there, and she observed that he was covered with goose bumps. He said he was freezing, and asked her to request a pair of shorts and a T-shirt. She then saw an officer coming in her direction. When she looked at him, he looked back and just shrugged his shoulders, saying it was “Lieutenant’s order.” When she determined that it was Lieutenant Slay-ton on duty, she let the matter drop. Although the incident upset her, Slayton had a reputation for causing problems if crossed, and she did not want her educational program or teachers to suffer by her interference in this matter.

In another such incident, inmate Johnny Barnes testified credibly that he was caged naked in one of the outdoor holding cages on a “misty” day. Although he was bleeding from his nose and mouth after a physical altercation with several correctional officers, Barnes was held in the outdoor cage for an hour and a half without receiving medical attention. In public view of whoever passed by, Barnes recalled that he felt like he was “just a[n] animal or something.” Tr. 10-1529.

Lieutenant Slayton at first denied that there was “ever any occasion” when an inmate was held in a holding cell completely nude. Tr. 20-3363. However, he later testified that there were instances where inmates were briefly held naked in cages, but insisted that they were dressed as soon as possible. Providing inmates with clothes was a priority, he testified, because of the inclement weather, and because “it’s just a common dignity.” Tr. 20-3364.

Clearly, there are times when prison officials will need to take an inmate’s clothes, as potential evidence after an incident, or for other justifiable purposes. And we agree with Lieutenant Slayton that providing substitute clothes is not only a matter of health and safety in inclement weather, but a matter of common dignity, given the public placement of the cages and the routine presence of female staff. However, his testimony that the inmates were never caged naked for more than brief periods lacks credibility in light of his inconsistent testimony on this point, as well as the credible testimony of Violet Baker and Johnny Barnes.

Moreover, some of these cages are visible from the main administrative offices for the yard (including the Lieutenant’s office), and are in full view of anyone who crosses the yard. Thus, it is apparent that such naked cagings would be known to, and thus implicitly, if not explicitly, condoned by supervisory staff. Indeed, the incident to which Violet Baker testified was clearly ordered by mid-level supervisory staff, as opposed to a renegade officer.

Such incidents may be relatively infrequent. Baker testified that she had never seen anyone naked in the cages except for that one day. 36 However, as Martin observed, this is the type of incident that is not typically memorialized in reports, making it difficult to determine how often it occurs. Notably, neither of the above incidents were documented, which leads the Court to conclude that there likely have been other undocumented instances as well. The reaction of the correctional officer to Baker’s inquiry suggests that this was not considered an extraordinary or unique event.

Leaving inmates in outdoor cages for any significant period — as if animals in a zoo — offends even the most elementary notions of common decency and dignity. It also fails to serve any legitimate penological purpose in any kind of weather, much less cold and rainy weather. The fact that it occurred at all exhibits a callous and malicious intent to inflict gratuitous humiliation and punishment.

(4) Cell Extractions

(i) Overview

The forcible removal of an inmate from his cell — also known as a “cell extraction” — is indisputably an essential tool in maintaining security in any prison. There will clearly be occasions when security concerns mandate that an inmate be removed from his cell against his will, such as where the inmate is suspected of harboring contraband, or has had an altercation with a cellmate. Such a forcible removal can be accomplished by various means. Staff who are completely unarmed, or armed only with a mattress or shield, can use sheer weight and numbers to overwhelm the inmate. 37 At the other extreme, staff can be highly armed, even during routine extractions, as is the case at Pelican Bay.

Indeed, the cell extraction process at Pelican Bay is an undeniably violent maneuver which can involve several weapons, including 38 millimeter gas guns, tasers, short metal batons, and mace. It also results in frequent injuries and infliction of pain. As Chief Deputy Warden Peetz summed it up, “cell extractions are a very, very violent maneuver ... Inmates get hurt and staff get hurt, and it’s just the nature of the thing.” Tr. 20-3316. As such, witnesses for both sides agree that cell extractions should be performed only when necessary. Indeed, under normal circumstances, an inmate should not be extracted absent an imminent risk to the safety and security of the institution. Scribner Tr. 7-1211.

(ii) Unnecessary use of cell extractions

Notwithstanding the above, for over two years (until February 1992), the number of extractions at Pelican Bay prison was clearly greater than necessary to meet safety or security needs. Prison records show that for the period through February 1992, 205 inmates were eell extracted in 26 months, at a rate of approximately 94.5 per year. Fenton viewed such numbers as “seriously excessive,” Fenton Tr. 5-792, while defendants’ experts also acknowledged that the number of extractions the first two years was “relatively high,” DuBois Deck at 4-5, or “a little on the high side.” McCarthy Tr. 15-2472.

A comparison with practices at the prison since February 1992 is particularly revealing. From that date through May 1993, a period of 16 months, only 26 extractions were performed, at a rate of approximately 1.6 per month or 19.5 per year. This constitutes a drop of almost 500 percent from the previous 16 months (October of 1990 through January of 1992), when 120 inmates were cell extracted at a rate of 7.5 per month or 90 per year.

Defendants’ experts testified that the high rate of cell extractions prior to February 1992 simply reflected the normal starting pains of a new prison, as well as the need to remove inmates who were kicking their doors, given the security concerns regarding cell doors. If the number of cell extractions had declined gradually between January 1990 and May 1993, this first explanation would carry significant weight; the record, however, demonstrates that the number of cell extractions was fairly consistent over 1990 and 1991 and then dropped off significantly in February 1992. Nor are we persuaded that the door problems explain the high frequency of cell extractions. Only a small number of the total extractions were reportedly door-related, and defendants testified that door kicking was often addressed through the use of fetal restraints.

Rather, the substantial reduction in extractions after February 1992 more likely reflected a change in policy. The Court further finds no indication that the safety or security of the institution was compromised by this change. As such, we agree with plaintiffs’ experts that the previous volume of extractions was unnecessarily high.

This conclusion is confirmed by the fact that prior to February 1992, cell extractions — using the full arsenal of force described below — were frequently employed, not to address imminent threats to security, but to respond to relatively minor infractions. For example, inmates were subjected to full scale cell extractions, often resulting in significant injuries, for not promptly giving up a meal tray, a jumpsuit, a pair of tennis shoes, or a skull cap. Martin Decl. at 40-57. 38 Many of these cell extractions were performed without any indication that the situation presented an imminent security risk. Clearly, defendants are entitled to enforce prison rules through normal disciplinary channels; inmates must be made to know that prison rules can not be ignored without consequence. However, in this case, it is clear that minor infractions were used as a pretextual justification for inflicting physical, and often brutal, punishment. 39

(iii) High degree of force

The mere fact that staff employs a procedure more often than necessary, may not, in and of itself, warrant serious concern. In this case, however, the unnecessarily high number of cell extractions in 1990 and 1991 takes on a much more troubling and ominous character given that these extractions routinely involved a strikingly high degree of force, and resulted in numerous injuries that were too often left unexplained by official incident reports.

Based on the evidence and written policies, cell extractions would generally proceed as follows: First, a supervising officer must ap prove the cell extraction. 40 Then a team of four correctional officers is assembled, wearing special protective gear, including helmets with visors. Each team member carries one of the following: a large plexiglás shield, a small metal baton, leg irons, or hand cuffs. The two officers carrying the shield and baton are generally referred to as the “shield man” and the “baton man.” A Medical Technical Assistant (“MTA”) must also be present in case of injuries.

The supervising sergeant and/or the lieutenant then typically fire some combination of the following weapons, any one of which can potentially cause serious injury: a 38 millimeter gas gun (which ejects high velocity rounds of rubber blocks), mace (a chemical agent that causes a burning sensation and tearing of the eyes) and a taser (which temporarily incapacitates an inmate by way of electrical shock). The four member team then enters the cell.

According to SHU regulations, the shield man enters the cell and rushes the inmate, striking him in the chest area with the shield and forcing him against a wall. If the inmate resists, the second team member, armed with the baton, should provide protective coverage by “rapidly and repetitiously” striking the inmate’s shoulder and clavicle with the baton. “As the inmate is pinned against the wall by the shield officer, and busy blocking the blows of the baton officer,” the third and fourth team members enter the cell, help subdue the inmate, and apply handcuffs and leg restraints. Trial Exh. D-49 at 1892-94. The inmate is then removed.

Not every weapon described above is used in every cell extraction; however, most cell extractions involved multiple weapons, and each weapon was often used more than once. 41 It is an undeniably massive combination of weapons - and manpower, especially considering that the target is almost always an unarmed inmate 42 already behind cell doors.

The taser, for example, which was routinely used until late 1992, inflicts significant pain, 43 and is described by one lieutenant as resembling being hit on the back with a “four-by-four” by Arnold Schwarzenegger. Scribner Depo. at 44-46. When activated, the taser ejects two thin wires with darts that pierce the inmates’ skin and transmit an electrical shock of 40,000 to 50,000 volts, temporarily paralyzing the body’s large muscles. Between July 1990 and July 1992, ta-sers were used in approximately 70 percent of all extractions. 44 And although the CDC’s main training academy suggests that maintaining the voltage 2 or 3 seconds should be sufficient, numerous inmates received discharges of 5 to 7 seconds, double the suggested length. Notably, there is no evidence that discontinuation of the taser in September 1992 adversely affected either the security of the prison or the safety and efficiency with which cell extractions were performed. 45

The 38 millimeter gas gun can also cause serious pain and injury. Generally fired into the cell through the narrow food port, it ejects high speed rounds of rubber blocks (approximately 1 and ¡é inches across) which ricochet in an unpredictable pattern around the cell. Given the small space of the cell, the ricochet has sufficient velocity to inflict significant pain or injury if it hits an inmate. As Captain Scribner testified, if such a ricochet hit an inmate’s head, it could possibly cause “great bodily injury.” Tr. 6-1113. Captain Jenkins similarly testified that it was possible that a ricochet from a gas gun round could strike an inmate in the face with sufficient force during a cell extraction to knock out his teeth. Jenkins Tr. 3-408. And should an inmate be hit directly at close range, the result could be serious injury and possibly death. 46 Former Warden Fenton emphasized the serious nature of using a gas gun in cell extractions: “The pellet [gas] gun ... If it hits in the wrong place, it can result in seriouss;injury or death.... Why would you fire that into a cell where the maximum distance doesn’t exceed eight feet if, in your own opinion, it’s capable of killing somebody?” Fenton Tr. 5-780 (referring to policy discussed in note 46, supra).

Nonetheless, gas guns were routinely discharged during cell extractions. Although no one has yet been killed, inmates have suffered injuries from shots that have hit the inmate directly or by ricochet. Notably, the evidence does not show that any consideration has ever been given to tempering the frequency or the manner in which the gas gun is used at the prison.

Use of the short baton has also resulted in numerous injuries. Although Pelican Bay policy provides that its purpose is to strike an inmate on the shoulders if he is still struggling or reaching around the shield, there is little evidence that the baton is limited to this use. Not only did staff testify that it was used to strike inmates on the feet or ankles, but the medical reports reflect that inmates received baton welts on their backs and other parts of their bodies during cell extractions. 47 The Court has no doubt that the baton has frequently been used against inmates in ways beyond those authorized by formal cell extraction procedures.

Both of plaintiffs’ experts were clearly amazed and disturbed by the level of force that was employed in cell extractions on a routine basis. 48 Martin unequivocally testified that the amount of force used in Pelican Bay cell extractions, as a routine practice, has been grossly excessive, utterly unbelievable, and without parallel in present-day American corrections. Decl. at 7; Tr. 8-1322. Former Warden Fenton stated that he could only come up with two explanations for the methods of carrying out cell extractions at Pelican Bay: “either they absolutely don’t know what they’re doing, or they’re deliberately inflicting pain.” Tr. 5-792. Neither was aware of any other institution which even approached the level of force in cell extractions that was routinely applied at Pelican Bay. Nor did defendants’ expert Dubois offer a defense of the routine use of multiple weapons which characterized actual cell extraction practices at Pelican Bay. 49

Martin also testified that the deployment of the extensive weaponry at Pelican Bay does not, in many cases, effectively advance the purpose of an extraction, which is to secure control of and move the inmate. Tr. 8-1323. Fenton expressed the same point in his testimony, stating that the routine use of all of these weapons “strikes me as a ritual of inflicting punishment. Why in the world would you over and over and over again do all these things?” Tr. 5-779-80.

Most significantly, the high level of force employed in cell extractions has resulted, time and time again, in significant injuries, many of which were indicative of beatings by staff. Martin’s declaration details many such instances (mostly unrebutted) which are too numerous to summarize here; we include one as an example:

Luis Fierro Extraction: According to the incident report ... Mr. Fierro ... refused to return his dinner tray. He gave as his reason the fact that he had been denied the opportunity to get a haircut that evening ... Ten minutes later ... an extraction team [arrived]_ Mr. Fierro did not have a weapon, and he is listed as 5'8" and 130 pounds. According to the incident report, Mr. Fierro “verbally challenged staff to enter his cell.” Sgt. Upton states that Mr. Fierro then began “banging on the cell door with his food tray.” CO [Correctional Officer] Davis described Mr. Fierro’s conduct as an attempt “to block the food tray slot with his tray.” According to Lt. Arneson, Sgt. Avila then “discharged two multiple baton rounds from the Federal Gas Gun to dislodge the tray from Fierro’s hands.” Sgt. Avila reports that- he fired one of these rounds off the cell floor, but he does not say the same for the other round. This omission is potentially significant inasmuch as Mr. Fierro was hit in the groin by one of the Sgt. Avila’s shots. If Mr. Fierro was standing close enough to the door to bang on it with the tray, or to attempt to block the food tray slot, it is difficult to imagine how a shot fired off the floor could ricochet up an hit him in the groin. Based upon Lt. Arneson’s description, it appears to be as likely that Mr. Fierro was hit directly with a shot to the groin....

After [the two gas gun] rounds were fired, Sgt. Upton states that he dispensed two bursts of mace which struck Mr. Fier-ro in the upper chest area, and these bursts “appeared to [a]ffect his movements.” Sgt. Upton then fired “two taser cartridges striking Fierro in the upper right leg, lower left leg, right upper chest and left arm,” which cartridges Mr. Fierro is supposed to have removed “during [the] administration of current.” The extraction team finally entered the cell and “restrained Fierro after a brief struggle.” Mr. Fierro is described as being “taken to the floor and placed in restraints without hitting any staff.” None of the 8 reporting officers describes in any further detail what steps were taken during the extraction itself in placing Mr. Fierro into restraints.

At the conclusion of this retrieval of Mr. Fierro’s meal tray, therefore, Mr. Fierro had been hit in the groin (possibly directly) with a gas gun round; struck with two bursts of mace; and been subjected to two jolts from the taser. In addition he suffered what are described in the report ... as “minor abrasions and bruises.” These “minor” injuries are described [in the medical report as]: (1) multiple areas of bright red bruising on back, especially upper back, (2) scratches on left neck, left cheek and right cheek, (3) abrasion and bruising of left shoulder; (4) bruising and swelling about both eyes at edges of eye brow, (5) small scratch on right calf ... (6) large abrasion [on] right chin, (7) large areas of bruising around both ankles, (8) bruising on left upper outer thigh, (9) abrasion on left shin, and (10) abrasion right calf.

Very few of these injuries can be accounted for by the officers’ version of events. Most notable are the injuries to Mr. Fierro’s face and back, which are the areas that are most commonly a “red flag” that an inmate has been assaulted. None of these injuries [described above] can be explained by the gas gun, the taser or the mace. Nor can they be explained by the process of “takfing Mr. Fierro] to the floor.” Common sense suggests that no matter how forcefully someone is slammed against the floor, it is impossible simultaneously to strike the area above both eyes and also to strike the chin. Moreover, Mr. Fierro clearly struck the floor (or was otherwise struck on the face) with great force: the bruises above his eyes are described ... as' ‘hematomas,’ which are more serious than (and require more force to inflict than) bruises. Finally, Mr. Fier-ro obviously cannot simultaneously have been taken to the floor face-first and back-first, and therefore the back injuries remain unexplained. In my experience, they conform precisely to the description of injuries of an inmate who has been beaten, most probably by a baton. Based on these unexplained injuries, I believe that Mr. Fierro was subjected to an extensive beating to the head and back after the extraction team entered the cell_ In conclusion, from defendants’ own reports it appears almost certain that Mr. Fierro was subjected to blatant and brutal punishment for disobeying an order — in this case, the order to turn over the dinner tray before discussing his ability to get a hair cut.

Martin Decl. at 19-22.

In another instance, an' inmate was extracted after failing to submit to a cell search. Afterwards, he was taken to the medical clinic for treatment for abrasions. He was in a very agitated state and dared officers to “do this again.” Instead of giving him time to calm down in a holding cell, he was immediately replaced in his own cell, still agitated. He immediately began strenuously kicking the cell door, which set off an electronic sensor in the control booth. He was then almost immediately extracted again, within about 25 minutes of the first extraction. 50 The second cell extraction team included the one officer who had received an injury (a cut) in the first cell extraction. As a result of the second extraction, the inmate was left unconscious for one minute, and suffered a contusion to the forehead and several abrasions. Trial Exh. P-1199 at 6371; Martin Decl. at 61-62.

As the above incident reflects, there was a practice for some period of time of including, as a member of the cell extraction team, the officer who had been either previously injured by the inmate or involved in whatever incident precipitated the cell extraction. Lieutenant Carl told investigators that, for a time at Pelican Bay, it was “normal” to include the previously involved officer and that he had to go to “real lengths” to get that practice changed. Trial Exh. P-3111 (tape interviews). Similarly, another officer told investigators that it was his understanding that whatever officers were “working the floor” were included on the cell extraction team, even if they had been a “victim” of the inmate being extracted. Id. It is generally understood, however, that such a practice should be avoided to allow the situation to defuse and preclude the potential for retaliation. As Lieutenant Mallory explained: “you may have an incident where the cop in the first cell extraction got bit, spit on, or whatever, and he may be a little mad. You don’t want to put him back in there and make him ... do something that he would not normally do.” Mallory Depo. at 83. The fact that Pelican Bay tolerated such a procedure further supports the conclusion that cell extractions were utilized as a vehicle for inflicting punishment.

As the Court has previously emphasized, cell extractions are an essential part of effective prison management, given that there will inevitably be circumstances where security concerns require the forced removal of an inmate from his cell. Nor do we doubt that each of the cell extraction weapons discussed above may have an appropriate use in certain situations. There may even be certain situations where it would not be excessive to use all of the above-described weaponry, either in some combined or sequential form. We need not, and do not, find that any particular weaponry or cell extraction strategy constitutes a per se use of excessive force.

What the record does reveal, however, is the disturbing pattern — an apparent modus operandi — of routinely using the same extremely high level of force, no matter the level of threat posed or the particularities of the situation. Not surprisingly, it is a pattern that has caused the substantial infliction of pain and left behind a string of injuries— injuries that are too often left unexplained and unjustified in official reports.

Viewed separately, the high level of force deployed as a routine practice, the string of significant injuries, and the unnecessarily high number of cell extractions, could each raise a legitimate concern. Combined, however, they are potent evidence that cell extractions at Pelican Bay have too often been considered, not as tools to be used sparingly in response to threats to prison security, but as opportunities to punish, and inflict pain upon, the inmate population for what were often minor rules violations. The evasive and cursory nature of incident reports, discussed supra, further reinforces this conclusion.

(5) Lethal force

The California Department of Corrections has for many years integrated firearms into its system of maintaining security, both inside and outside of housing units. 51 Pelican Bay is no exception. It employs a variety of firearms to maintain control inside the general population maximum security housing units and exercise yards and inside the SHU. 52 The general population yards are observed by an armed officer in a central tower overlooking the entire yard, as well as by armed officers in control booths in each housing unit, who oversee daily activities through windows that look onto the yard. 53 The housing units in both the maximum security general population units and the SHU are monitored by armed officers in the control booths. In the former, the armed officers have a direct line of sight into the cells, while in the SHU, armed control booth officers have a direct line of sight into the common pod area outside of the cells but not into the cells themselves.

Notably, other large prison systems, such as New York, Texas, Ohio, and the Federal Bureau of Prisons, manage their prisons (except for the perimeters) without the use of any firearms. Indeed, reliance on firearms in housing units (either general population or security housing units) and exercise yards to maintain control and break up incidents is unusual. Defendant Gomez was aware of only one other state, Nevada, which employs firearms inside housing units.

The record does not, however, support a finding that the decision to deploy firearms at Pelican Bay in and of itself constitutes a policy or practice of excessive force. Whether firearms should be integrated into a prison’s security system is a matter best left to the sound discretion of prison administrators. However, given that every use of a firearm creates the potential for death or serious bodily injury — not only for the intended victim but for others- nearby as well — a policy of arming prison staff can easily lead to the application of excessive force.

Recognizing this, governing regulations prohibit the use of firearms except when “absolutely necessary,” i.e. “only as a.last resort after other reasonable and available resources have been considered and exhausted or are determined to be clearly inappropriate in view of the immediate need to use armed force.” Cal.Code Regs. § 3276(b); see also Trial Exh. D-37, California Department of Corrections Operations Manual (referred to as “DOM”) at 55050.8 (“Employees shall not discharge a firearm ... except under [certain specified] circumstances and only after all other reasonable means fail”). The evidence showed, however, that staff have resorted to firearms too quickly — before any life threatening situation has developed — rather than reserving such lethal force as “the force of last resort.” 54

Based on the evidence presented, we conclude that firearms at Pelican Bay have been used unnecessarily, and in some cases, reek- lessly. 55 However, in contrast to the instances of excessive force discussed in the sections above, the record before us does not demonstrate that lethal force has been applied maliciously for the purpose of causing harm, rather than in a good faith effort to restore order, on more than the isolated occasion. Rather, it appears that, in many instances, officers resorted first to lethal force because prison administrators failed to supply them with alternative weapons or because prison policies promoted the use of lethal force. See, e.g., notes 67, 64 and accompanying text, infra. These facts, however, explain more than just the actions of staff. They also reveal that defendants are strikingly unconcerned that lethal force may be applied even when a lesser degree of force would be sufficient.

b. Existence of Pattern

As defendants emphasize, Pelican Bay is a large facility that houses as many as 3,900 inmates at one time. However, the size of an institution does not mandate that plaintiffs prove some particular number of incidents to demonstrate the presence of a pattern of excessive force. Indeed, given the code of silence, the lack of specificity in many incident reports, and the fact that some number of incidents go unreported by staff and inmates, it is surely impossible to determine conclusively the number of times that excessive force has been used against inmates at Pelican Bay. 56

Instead, plaintiffs must prove to the satisfaction of the Court, and by a preponderance of the evidence, that the instances of excessive force presented to the Court reflect a pattern, rather than isolated, aberrational events. In this case, plaintiffs’ experts were firmly convinced that they had discovered the former rather than the latter. Martin concluded that the pattern of excessive and unnecessary force at Pelican Bay was “clear and unmistakable” and that there “are circumstances in which force is routinely employed as a method of punishing, of inflicting physical pain and discomfort on [ ] the population.” Tr. 8-1303-4; Martin Decl. at 157. Nathan agreed, finding Pelican Bay “unique” in his experience.

I have simply never observed the level of violence ... the overt reliance upon violence. I’ve just never seen anything like it. I don’t think you’ll find anything like it. I don’t think you’ll find it at Marion, Illinois. I know you won’t find it at Georgia State Prison, which is the maximum security prison for the state of Georgia. I know you won’t find it in the state of Texas. I know you won’t find it in the state of Ohio. There just isn’t any parallel.... It is a place in which officers exercise unfettered discretion to take whatever physical action they deem appropriate or necessary to control, to punish, to accomplish whatever objective it is they wish to accomplish.

Tr. 13-2050-51. Defendants’ expert expressed a general opinion that Pelican Bay was “well run,” but did not specifically address the existence of a pattern or practice of excessive force. DuBois Tr. 29-4689.

Based on the evidence regarding specific incidents and practices, and the opinions of prison experts, the Court is convinced that the instances of force being used excessively and for the purpose of causing harm are of sufficient scope, variety, and number to constitute a pattern. Plaintiffs have convincingly documented a staggering number of instances in which prison personnel applied unjustifiably high levels of force, both pursuant to, and in contravention of, official prison policies. Simply put, the evidence before the Court is proof of the most powerful, unambiguous kind that a pattern of excessive force has become an undeniable reality at Pelican Bay. 57

2. Inadequacies in the Systems for Regulating the Use of Force

There is no dispute among the parties that the use of force must be carefully regulated and controlled in order to prevent abuses against inmates, as well as to ensure the overall safety and security of the prison. As Fenton explained:

In running a penitentiary, a high security prison, force is a necessary element. It must be used when it’s appropriate and when it’s required. But it is the most dangerous and potentially destructive function that takes place in a prison. Not only can it destroy or seriously injure the people it’s directed against, but it can do enormous damage to ... the spirit and morale of the inmate body [and] to the staff.... The use of force can be enormously dangerous. For that reason, it ought to be absolutely seriously monitored.

Tr. 5-766. Indeed, there was no evidence suggesting that the failure to adequately monitor or regulate the use of force would serve any legitimate penological purpose or otherwise advance the security of a prison.

A system that adequately monitors and regulates the use of force consists of five components: (a) written policies that clearly identify for line staff when and how much force is appropriate under different circumstances; (b) training of correctional officers regarding the proper use of force; (c) supervision of the use of force to ensure that it is consonant with departmental and institutional policies and procedures; (d) investigation of possible misuses of force; and (e) officer discipline for the misuse of force. Nathan Tr. 13-1999-2000; Nathan Deel. at 15; see also Fisher v. Koehler, 692 F.Supp. 1519, 1551 (S.D.N.Y.1988), aff'd, 902 F.2d 2 (2nd Cir.1990).

Each of these interrelated components builds upon and reinforces the others. Thus, adequate written policies provide the necessary framework for properly training staff and evaluating subsequent conduct. Yet, written policies alone serve little purpose unless staff are trained as to their content. Adequate supervision and investigation are necessary to ensure that, in practice, staff are properly implementing written policies and principles learned through training. Finally, a meaningful disciplinary system is essential, for if there are no sanctions imposed for misconduct, the prison’s “policies and procedures ... become a dead letter.” Nathan Tr. 13-2003.

The evidence shows that the system for controlling use of force at Pelican Bay suffers from’ serious deficiencies, particularly with respect to the supervisory and investigatory components described above. The Court also finds that these deficiencies, known and tolerated by defendants, are a significant cause of the misuse of force at Pelican Bay.

a. Written Policies Regarding the Use of Force

At trial, no one disputed the importance of having written policies and procedures regulating the use of force. As Nathan observed, clear and consistent written policies and procedures are the “cornerstone for the development of ... a mechanism ... to control force in high security prisons.” Nathan Tr. 13-2010. Defendants’ expert DuBois agreed that written policies and procedures were “important” in the administration of a correctional facility, and that post-orders, which guide a staff member through his or her specific duties, were “very important” from a correctional officer’s point of view. Tr. 29-4742-43. Indeed, the significance of written policies concerning use of force is self-evident. Not only do they give direction to staff regarding when, and what kind of, force should be used, but they also provide an important basis for developing training programs, foster equal treatment of inmates within the prison, and furnish a consistent benchmark for evaluating staff conduct in the event allegations of misconduct arise.

As defendants point out, there is a general policy regarding use of force set forth in Title 15 of the California Code of Regulations; however, because it is exceedingly general, this policy fails to provide any meaningful guidance regarding use of force. 58 In addition to Title 15, there are departmental-level policies set forth in the DOM, institution-level policies issued by Pelican Bay, and finally, “post orders” which guide a staff member through his or her specific duties. After reviewing these materials, Nathan con-eluded, and we agree, that the web of written policies relating to use of force at Pelican Bay have often lacked the necessary clarity and consistency to provide meaningful guidance on the use of force in certain key areas such as cell extractions and use of lethal force. 59 It is also clear that the absence of authoritative written guidelines allows policy to shift according to the predilections of individual mid-level staff. 60

The use of tasers presents one example where the written policies fail to provide a consistent framework. Title 15 itself does not address tasers. The DOM authorizes their use in cell extractions but adds the following caution: “Situations involving the removal of an inmate from a cell, in the absence of immediate danger of self-inflicted injury, injury to another person, or substantial property damage, shall be carefully evaluated piior to the use of the taser.” DOM §§ 55050.18. Pelican Bay written policy however, appears to dispense with this caution, virtually mandating the use of tasers in a large number of extractions. In the SHU, for example, the cell extraction team shall “only enter the cell [of an assaultive inmate] if the inmate [has] been immobilized by use of the taser.” Trial Exh. D-49 at 18292. In the general population units, the taser is “the weapon of choice” in forced relocations of an inmate. Trial Exh. D-45 at 3. These kinds of discordances contribute to divergent understandings among staff regarding the proper use of the taser. 61

There is also a conflict between the written general population procedures and actual training materials regarding the length of time the taser’s electrical current should be discharged. Training materials that were used suggest a 2-8 second discharge to knock down and subdue the average man for 20-30 seconds, 62 while the policy governing general operation procedures sanctions a discharge of over twice that duration (5-10 seconds) to immobilize an inmate for a few minutes. Trial Exh. D-45 at 18096.' Given the substantial pain inflicted by the taser, and the still uncertain health risks, see note 43, supra, clear written policy on use of the taser is critical, particularly since the taser itself can not be pre-programmed to regulate or register the length of the charge. The taser guidelines also furnish an example of an area in which written policy has not kept pace with changing practice. Although use of the taser in cell extractions informally stopped in the Fall of 1992, this change was never incorporated into any written policy.

Another problematic area, particularly during Pelican Bay’s first two years, concerns the written policies governing the use of lethal force. The reliance on direct gun coverage of cells in the general housing population, and the frequency of in-cell assaults, make it essential that there be clear written rules regarding the handling of cell fights. As the Chief Deputy Warden acknowledged, because of the “seriousness of the consequences,” it is “extremely important” for staff “to understand precisely when it is and when it is not appropriate” to fire at inmates. Peetz Tr. 20-3300.

Again, Title 15 and DOM provide general statements regarding the use of lethal force, but they are far too broad to provide any specific guidance in handling recurrent situations, such as cell or yard fights. 63 Nonetheless, as of June 25, 1991 (a year and a half after the prison opened), a statewide CDC task force found that “PBSP [Pelican Bay] does not have a written policy specifically designed for the handling of in-cell assaults,” Trial Exh. D-59 at 51741, an omission which Nathan found “shock[ing].” Nathan Deel. at 38. As a result, staff resorted to lethal force in the general population units before exhausting lesser alternatives. For example, if an assault continued after the control booth officer opened the cell door to allow one inmate a chance to exit, the control booth officer would immediately attempt to control the assault with a firearm, by firing either a warning shot or a shot for effect. 64

Another critical firearms issue that suffered from a lack of authoritative written rules was the use of warning shots inside the SHU units. Some post orders, as well as Pelican Bay Addendum to DOM § 55050, provided that officers could fire a warning shot inside the SHU before firing at inmates. Other post orders, however, prohibited the firing of warning shots, explaining that such shots could not be safely fired inside the SHU because the areas are small and composed of concrete or steel. As a result, there was no clear consensus regarding the prison’s warning shot policy. As the Chief Deputy Warden acknowledged, “many people in the staff’ believed that no warning shot could be fired. Peetz Tr. 20-3300-3301. One Associate Warden, for example, stated that during her tenure there were no warning shots permitted in the SHU. “That means that if you resort to violence, we can shoot [to injure].” Garcia Depo. at 93-94. However, as Peetz testified, the correct policy has always been that warning shots are permitted in the SHU. Tr. 20-3300; see also Carl Depo. at 269-70. Yet, Robert Bark, the Associate Warden in charge of the SHU facility, stated that from what he could recall, the warning shot policy had “changed back and forth a couple of times” during his tenure, and that when he left, a warning shot was allowed but “discouraged.” Bark Depo. at 113.

Given that the intentional firing of a gun, even as a warning shot, appears to constitute a use of “deadly force,” see DOM § 55050.3, Nathan found the contradictory policies and confusion regarding a subject of “this magnitude” to be “intolerable.” Nathan Deck at 43-44. Nonetheless, it does not appear that the problem was swiftly resolved. Only after the issue had come up “several different times” was the matter definitively addressed. Peetz Tr. 19-3223. As Officer Brodeur testified, one of those times was a review of a shooting in which he was involved; however, the post orders were not changed until “about three months later.” Tr. 24-3992.

Finally, we note that although cell extractions represent one of the most significant and recurrent uses of force, written cell extraction policies are less than clear as to whether or not any effort should be made to relate the degree of force used to the behavior of the inmate or the particular circumstances presented (including whether the inmate is armed). Thus, for example, the cell extraction policy for the SHU simply provides that the lieutenant will decide whether an extraction is necessary when an inmate has refused to cooperate in a cell move or submit to restraints, and lists the different types of equipment and weapons to be utilized in the process. Trial Exh. D-49 at 18290-91. Similarly, one officer observed that while there has always been the general guideline that staff should only use the minimum force necessary, there are no specific guidelines regarding application of that standard in different situations. Van de Hey Depo. at 112.

Perhaps of most concern, the Court finds the deficiencies in certain written policies, described above, symptomatic of a more general disregard for the importance of written policies. The Court notes that, as of the date of trial, a number of officially sanctioned policy changes still had not been memorialized in written form, thus creating a schism between the prison’s written policies and its actual operating procedure. 65 Notably, the current written operation policy for the SHU has remained “under revision” since 1990. Scribner Tr. 7-1214-1215. Although this means that senior staff sometimes find themselves directing line officers to rely on “unapproved” revised policy, see, e.g., Helsel Tr. at 21-3565-66, no prison staff at trial evinced any concern over the fact that the SHU was operating under an outdated written policy with no apparent date set for the issuance of a current version.

As might be expected, the lack of completeness and consistency in written policies relating to the use of force, and the lack of importance ascribed to written policies in general, have also served to .undermine the legitimacy of those written policies that are in effect. It was not uncommon for staff to testify that they were unaware of written policies, 66 or for the evidence to show that a written policy was simply not followed, whether by design or because of a simple lack of familiarity. 67

b. Training in the Use of Force

Correctional officers receive formal training in the use of force during their basic training at the state-wide R.A. McGee Correctional Training Center (known as “the Academy”). Additional formal instruction is also provided at individual institutions as part of “in-service training.” 68 In 1991, the Department of Corrections conducted an audit of Pelican Bay’s in-service training department and found the facility in full compliance.

Plaintiffs contend that defendants have failed to provide adequate instruction to Pelican Bay officers regarding the appropriate use of force. None of plaintiffs’ experts, however, engaged in any systematic qualitative or quantitative analysis of the use of force training that is provided by way of the Academy and in-service programs. 69 As such, the Court can not meaningfully assess the adequacy of the formal training that has been provided to officers currently employed at Pelican Bay.

Plaintiffs’ experts suggest that, given the evidence that force is frequently misused, we may properly infer that Pelican Bay officers are not receiving adequate use of force training. We agree that this evidence raises legitimate questions as to the adequacy of the training that is provided on the subject of force. 70 However, we can not on this basis alone draw the sweeping conclusion that the training regarding use of force is inadequate. Indeed, there are circumstances in which a pattern of excessive force could persist despite adequate formal training (for example, where officers receive adequate formal training which is subsequently undermined by improper supervision). 71 We further note that, if plaintiffs’ “bootstrapping” approach were accepted, the Court could conclude, based solely on a finding of a pattern of excessive force and without specific supporting evidence, that each and every one of the five above-described components of a system for regulating use of force was necessarily inadequate. This we decline to do.

c. Supervision of the Use of Force

Adequate supervision is probably the most critical component of any system that regulates the use of force: not only does it serve as an immediate cheek on any abuses, but it also creates an atmosphere that encourages responsible conduct. At the same time, adequate supervision ensures that sufficient force is used to maintain security for staff and inmates.

At trial, defendants’ witnesses readily acknowledged the importance of providing effective supervision over the use of force. The evidence, however, shows that senior prison administrators have, for the most part, abdicated their responsibility in this crucial area. Indeed, Pelican Bay’s approach to the use of force is often so passive that plaintiffs’ expert concluded that there is a “near total absence of meaningful supervision” of the application of force at Pelican Bay. Nathan Tr. 13-2039.

Because the use of lethal force, i.e. firearms, is supervised through a separate administrative mechanism than other uses of force are, the Court discusses separately the supervision of the use of non-lethal and lethal force.

(1) Supervision of the Use of Non-Lethal Force

Given limited time resources, prison administrators necessarily rely on written inci dent reports as well as informal feedback from staff to monitor and regulate staff conduct involving use of force. These reports generally consist of a cover form (No. 837) which contains a synopsis of the incident, a statement from each officer involved in the incident, and any pertinent medical reports or reports of chemical agent usage. All such reports are reviewed by a senior prison administrator (either the Warden, Chief Deputy Warden or Associate Warden), and then forwarded to the Director of the Department of Corrections under the Warden’s signature. At Pelican Bay, the reports take on added importance as a supervisory tool because the Warden does not permit videotaping of taser use or cell extractions, as is the practice in some prisons. 72

The Court finds that supervision of the use of non-lethal force at Pelican Bay is strikingly deficient. The breakdown in supervision reveals itself in a number of ways. First, senior administrators permit, or even encourage, officers to submit overly general incident reports, a practice which both parties’ experts criticized as making it impossible to evaluate the propriety of staff conduct. As defendants’ expert stated, “[the incident reports] appear to be generic in nature. They’re — they’re not reports that I, as a Commissioner in Massachusetts ... looked at as ... covering the whole event but more in terms of describing what happened very generically.” DuBois Tr. 29-4713. It is not unusual for a report to gloss over events and inmate injuries by reporting little more than the team “gained control” and “applied mechanical restraints.” See, e.g., Trial Exh. P-4925 at 3208. Other reports provide more detail but still lack sufficient information to enable a supervisor to determine what oc-eurred. See Martin Deck at 64-65,158 (pattern of sanitized reports is “simply undeniable.”). In still other cases, the reports of different officers are suspiciously identical. 73

The fact that such reports are routinely accepted leaves the clear inference that senior prison administrators not only have little concern as to what actually occurred, but that they affirmatively approve of such reports. Notably, one sergeant testified that his supervisor wanted reports to be “vague and non-specific,” because otherwise those higher up the chain of command would not accept them. Cox Tr. 15-2348-49.

Second, it is not unusual for prison administrators to turn a blind eye when an incident report clearly calls for further inquiry, such as when it indicates that an inmate sustained serious injuries that are either unexplained or suspiciously explained. A review of the following representative reports, which were accepted by supervisors, without comment, illustrates this problem.

Peter Moreno Incident Report

On April 29, 1991, Peter Moreno was extracted from his cell. The medical report attached to the full incident report showed that, among other injuries to his face and legs, Moreno’s back and upper arms were covered with 12 welting abrasions which the MTA described as “probable baton markings.” The report submitted by the officer using the baton, however, does not explain the markings, stating only that he was required to strike several blows with the baton around the feet and ankles because Moreno was resisting the imposition of leg irons. Trial Exh. P-1142 at 4975. Nor does the incident report otherwise explain the welts. At trial, the baton officer testified that he still could not say how Moreno’s injuries were caused. Owens Tr. 28-4588. Although the Chief Deputy Warden agreed that “it would certainly be extraordinary that covering an inmate’s ’ back with baton markings could be proper,” Tr. 20-3276-77, 74 and the incident report provided no explanation, no follow-up inquiries were ever made. 75 Defendants’ experts both agreed that the incident should have precipitated further investigation. 76

Jesse Calhoun Incident Report

Jesse Calhoun was extracted from his cell on October 25, 1990. Again, the medical injury report showed eight baton welts in the middle of Mr. Calhoun’s back. Trial Ex. P-4925 at 3218. The officer equipped with the baton, however, did not report using it at all, although he does state that Calhoun kicked him in the knee before being restrained. 77 Nor does the report otherwise explain the welts. As Martin concluded, these facts were “obvious red flags” indicating that an investigation was warranted. Decl. at 700. Nonetheless, the baton officer involved was never questioned about the incident by any superior. Franklin Tr. 24-4070-4071.

Luis Fierro Incident Report

According to this incident report, Luis Fi-erro was restrained after “a brief struggle.” P-1102 at 4165. This description is not easily reconciled with the rather extensive injuries suffered by Fierro including “multiple areas of bright red bruising on back,” bruising and swelling above both eyes, and scratches and abrasions on Mr. Fierro’s chin, cheeks and neck. Defendants did not call any witnesses who were present during this incident, but their expert agreed that, as a reviewing supervisor, he would not consider the report a satisfactory explanation for the injuries. DuBois Tr. 29-4812.

Julius Dunn Incident Report

According to the medical report accompanying the incident report, Julius Dunn was rendered unconscious for approximately one minute at the end of his cell extraction. None of the officers’ reports, however, mentions this fact or describes any blow to Dunn’s head. Injuries to the head and face are a clear “red flag” that force may have been misused, since, under standard procedure, head and face injuries are to be avoided where possible. In this case, the red flag was especially vivid because the head injury occurred during the second of two back-to-back extractions, and the second extraction team included an injured member of the first team. Trial Exh. P-1199 at 6376 (“R. Williams” member of both teams). Both of defendants’ experts agreed that, based on the incident report, an investigation should have been initiated. Chief Deputy Warden Peetz, by contrast, testified that there was nothing in the incident report that concerned him enough to order an investigation. Tr. 20-3277. The supervisory lieutenant confirmed that no one had ever asked how Dunn had been rendered unconscious or otherwise inquired into the incident.

Incident Report of 20-inmate cell extraction

In January 1991, prison staff extracted 20 inmates, seven of whom received significant injuries, primarily lacerations to the scalp, head or face. The generic and essentially identical reports submitted by the officers involved did not explain the injuries and gave supervisors little basis for reviewing what had occurred. Trial Exh. P-1098; Martin Decl. at 49-52. The substantial head injuries sustained by seven of the inmates were particularly notable given the fact that, while all 20 inmates were originally charged with serious rules violations, those charges were subsequently dropped against all but three of the 20 inmates. Lopez Tr. 14-2157-59, 14-2241-45.

Daniel Molano Incident Report

When Daniel Molano suffered substantial facial injuries (a laceration to the bridge of his nose, swelling and bruising above both eyes, and a swollen lower lip), the incident report explained that he had been dropped face down on the stairs after a cell extraction. Trial Ex. P-1142 at 4965; Martin Decl. at 65-66. According to the testimony of an officer who was present, at the time Molano was “dropped,” he was being carried by “at least four” officers head first with his arms cuffed behind his back. Owens Tr. 28-4576, 4592. Plaintiffs’ expert testified that, in his experience, it is common to see accidents such as “stair-dropping” used to mask prior injuries. Martin Decl. at 66. Defendants’ expert did not disagree, and testified that he would have initiated an internal affairs inquiry based on the report that the inmate had been dropped, even without the presence of facial injuries. DuBois Tr. 29-4814-15.

Arturo Castillo Incident Report

Similarly, when Arturo Castillo suffered a serious injury to the top of his scalp during a cell extraction, the reports explained that he had accidentally struck his head against his toilet while falling. The seriousness of the injury (which required hospitalization), the unusual explanation, and inconsistencies in the reports should have caused this incident, in the words of plaintiffs’ expert, to be “investigated up one side and down the other.” Martin Decl. at 74. Yet, Chief Deputy Warden Peetz testified that he saw no reason to inquire into the incident. Tr. 20-3273; see also id. at 20-3269-70 (Q: ‘You weren’t concerned about a report that said the very top of the inmate’s head was wounded by falling down upon a toilet?” A; “No I was not.”). 78

Finally, senior prison administrators offer little incentive for officers to come forward with eyewitness reports of misuse of force. Indeed, it is apparent that officers who report misuse of force against inmates risk reprisals not only from other staff but from prison management as well. See note 4, supra. Such reports are also implicitly discouraged because they rarely lead to results. As discussed infra, even when an officer reports a misuse of force, it is unlikely that any wrongdoing will be found or discipline imposed, a pattern which plaintiffs’ expert described as an “extremely disturbing” indication that the system for checking the use of force is “entirely dysfunctional.” Martin Decl. at 90.

The Warden’s lax attitude toward supervising the use of force is signaled to his staff in other ways as well. Defendants’ expert McCarthy testified that, during his tenure as Warden, he personally reviewed all incident reports because they were an important tool in helping him supervise the use of force. Warden Marshall, however, delegated review of incident reports, save for the cover sheet, to subordinates sometime at the end of 1990 or 1991, although he resumed the practice of personal review one month before testifying in this action. 79 Nor has Warden Marshall ever observed a cell extraction or the use of a gas gun or taser at Pelican Bay. Rather, cell extractions are supervised or observed by mid-level staff (lieutenants and sergeants). As Fenton testified, if the Warden personally attends scheduled uses of force (e.g., a cell extraction) on a consistent basis, such a practice encourages staff “not lightly [to] recommend this use of force.” Fenton Decl. at 10. “I am amazed,” he testified, “by the manner in which senior management has absented itself from the use of force at Pelican Bay.” Id. at 9. While it may not be practicable for the Warden to attend every cell extraction, given the size of Pelican Bay, the complete non-involvement of the Warden and other high-ranking administrators is indicative of the fact that supervision of the use of force is a tragically low priority at Pelican Bay.

(2) Supervision of the Use of Lethal Force

Given Pelican Bay’s substantial reliance on firearms, and the fact that every firing of these weapons potentially inflicts serious injury or death, effective supervision over the use of firearms is particularly critical. The evidence shows, however, that, despite a facially complex system for reviewing the use of firearms, the lax attitude toward the use of non-lethal force, described above, is equally evident in the area of lethal force. Indeed, meaningful firearm supervision of the kind that actually protects human beings is almost non-existent.

As an initial matter, we note that a significant number of shootings go unreviewed altogether. Department regulations require that all firearm discharges be reviewed to determine whether staff actions comply with policy guidelines governing the use of firearms. When the shooting incident results in serious injury or death, the review must be conducted by a departmental Shooting Review Board (“SRB”). Shootings that do not result in serious injury or death must be reviewed by an institutional Shooting Review Team (“SRT”). DOM § 55050.13. 80 Prison records show, however, that at least 24 rifle shots in 19 separate incidents (between December 1989 and March 31, 1993) were never reviewed at all. Trial Exh. P-5571. Some of these shots were shots “for effect” (i.e. shots intended to hit a person) or shots resulting in injury. 81 In an additional 17 incidents, involving 30 shots, a shooting review number was assigned to the incident, but there is no evidence that the shooting was ever actually reviewed. Trial Exh. P-5571. Given defendants’ failure to refute the apparent lack of review in these incidents, we conclude that no such review occurred.

Nor have regulations concerning the composition of SRTs been adhered to. Pursuant to the DOM, SRTs must consist of a chairperson plus three officers from different correctional ranks. DOM § 55050.13.1. Yet, until three months before trial, SRTs at Pelican Bay inexplicably consisted of only one administrator. Not only does a one person “team” clearly defeat the very purpose behind the group approach to shooting reviews, but it also signals that such shootings are not considered serious enough to warrant review by more than one person.

Of most concern, however, is that the actual review process has been rendered a mockery of its intended purpose. The shooting officer’s incident report is typically taken at face value and given little scrutiny, even where it fails to identify any facts that would justify use of lethal force. 82 One administrator candidly expressed the prevailing deferential attitude toward incident reports: “I can’t second-guess the officer_ The only person who can make the determination on whether to fire or not is the officer at the time of the incident.” Lopez Tr. 14-2223. Nor is it a matter of practice to interview persons who either witnessed or were directly involved in the incident. And although reviewers are charged with determining whether a shooting was in complete compliance with relevant policies and procedures, they are not always aware of what those policies are. Consequently, shooting reviews at Pelican Bay are little more than a perfunctory validation of the incident report itself. 83

A notable illustration of the lack of meaningful review is provided by the administration’s response to officer claims of “stabbing motions” to justify the use of lethal force. As plaintiffs’ expert observed, “nobody ever makes a stabbing motion if they don’t have a weapon,” yet officers at Pelican Bay repeatedly attribute such motions to inmates to explain shooting incidents when after the fact no weapon is found and no one has been cut. Fenton Tr. 5-759. Such a claim suggests that the officer has either made an honest mistake or is engaged in after-the-fact justification. Under either circumstance, some supervisory action is warranted (further training in the former, or training and discipline for lack of candor in the latter). There is no evidence in the record, however, that such action .ever took place; on the contrary, a statement that an officer saw “stabbing motions” appears to automatically sanction the shooting.

Plaintiffs’ expert Nathan joined in Fen-ton’s condemnation of the shooting review process, calling it a “farce.” Tr. 13-2038. Defendants’ expert also had little positive to say about the shooting review process, and agreed that shootings “could stand more scrutiny” at Pelican Bay. DuBois Tr. 29-4766-4767. 84 The Warden, however, ex pressed no dissatisfaction with shooting review practices at the prison. Although Warden Marshall receives a copy of all shooting reviews, he could not recall a single review that he had found unsatisfactory. Tr. 22-3815.

d. Investigations

Investigations into allegations of misconduct by Pelican Bay employees are conducted primarily through the Internal Affairs Division (“IAD”) of the Investigative Services Unit (“ISU”). 85 The responsibilities of the IAD are twofold. The first is to identify allegations of misconduct that merit further investigation and obtain approval from the Warden to initiate an investigation. The IAD opens approximately ten to fifteen excessive force investigations annually, which constitutes about 20 to 35 percent of the investigations into staff conduct each year. 86 The second is to conduct the investigation (through interviews and review of medical reports and other documentary evidence) and prepare an Internal Affairs report. With respect to each allegation, the report can make one of four findings: (1) that the allegation is “unfounded” because the alleged conduct did not occur, (2) that the allegation is “not sustained” because there is insufficient evidence to prove the allegation, (3) that the alleged conduct did in fact occur, but that it was fully justified and therefore the officer should be “exonerated,” or (4) that the alleged conduct occurred and was not justified, and therefore the allegation should be “sustained.” Alternatively, the IAD investigation can result in “no finding.” The report is subsequently submitted to the Warden for his review and approval. It does not appear that the Warden has ever withheld his approval because of dissatisfaction with some aspect of the report.

It is clear to the Court that while the IAD goes through the necessary motions, it is invariably a counterfeit investigation pursued with one outcome in mind:' to avoid finding officer misconduct as often as possible. As described below, not only are all presumptions in favor of the officer, but evidence is routinely strained, twisted or ignored to reached the desired result. The consequence, as Nathan testified, “is to reinforce an already clear message to line staff that unnecessary and excessive force will be tolerated, if not actively encouraged.” Nathan Decl. at 79. Notably, the evidence showed that, in the nearly four years since the prison opened, only one officer was found to have engaged in what could be described as a relatively major misuse of force (punching a restrained inmate twice in the face with a closed fist), and those charges were later dismissed by the Warden. 87 Marshall Tr. 22-3745-8.

(1) Initiation of Investigations

The IAD can not initiate an investigation without first obtaining approval from the Warden or Chief Deputy Warden, who have complete discretion on this matter. The IAD can recommend that a matter be investigated based on grievances or letters received from inmates, informal verbal reports from staff or written incident reports (which are routinely received within a week of the incident). Although the record is replete with incident reports that clearly should have triggered further inquiry, see section II(A)(2)(c)(l), su- pro, the IAD rarely, if ever, sought approval for an investigation based on an incident report. As Deputy Chief Warden Peetz testified, “[v]ery few [investigations] would come from [ISU Captain] Jenkins as a result of Jenkins reading an [incident report] and coming to me and saying, ‘hey, we need to open a case.’ ” Peetz Tr. 20-3264. Rather, when IAD did seek approval for an investigation it was invariably because an officer came forward with a verbal report or because an inmate submitted a grievance or complaint. 88

(2) Investigative Process

In most cases, the fact-finding process of interviewing witnesses and obtaining pertinent medical reports functions satisfactorily, although the record does contain instances in which obvious witnesses were left unquestioned. It is in the evaluation of the information obtained, however, that the process loses all integrity.

First, the IAD applies standards more consistent with criminal than civil or administrative proceedings. Defendants’ witnesses testified that an inmate allegation of excessive force will only be sustained if the wrongdoing was “clearly prove[d] with certainty,” or “beyond a reasonable doubt.” Long Tr. 17-2801; Beckwith Tr. 17-2764. Suspicions that officers are withholding information are ignored unless such misconduct can be “absolutely prove[d].” Beckwith Tr. 17-2752-53. As Nathan observed, “If the inmate must establish the misuse of force ‘conclusively’ and by evidence that excludes every ‘possibility' other than officer misconduct, he will never prevail.” Nathan Deck at 88.

Second, not only are the above standards exacting on their 'face, but the manner in which they are applied at Pelican Bay makes them almost impossible to meet. Internal Affairs routinely minimizes or ignores evidence adverse to staff, and strains to find explanations (however implausible) that can be used to reject allegations of excessive force. Thus, as long as some theoretically possible version of events exculpates the officer, it will be relied upon to avoid a finding of culpability, even though it may be highly improbable and lack any credible basis in the record.

We by no means intend to suggest that inmate allegations of staff misconduct should be accepted without careful scrutiny. Some inmates may have serious credibility problems; others may be prompted by improper motives. However, as the examples below illustrate, Pelican Bay has, for the most part, abandoned any notion of reasoned analysis in favor of absolving prison officers of any fault.

Investigation re: use of force against Inmate Calhoun

In March of 1992, Officers Rader and Hle-bo approached Jesse Calhoun’s cell and ordered him to cuff up. 89 Calhoun at first refused, but then agreed to comply and offered no resistance. During the escort, Officer Rader and Calhoun became involved in a heated verbal exchange and Rader called Calhoun a child molester and a punk. Officer Hlebo, the other escorting officer, later gave an account of what happened next which can be summarized as follows: Rader grabbed Calhoun’s head “and drove the side of his face into the wall approximately three times.” Calhoun’s eyeglasses fell to the floor and broke. Hearing noise, Sergeant Avila approached the scene and asked what the problem was. When Rader responded that Calhoun doesn’t “want to do what he’s told,” Avila clutched Calhoun’s throat with one hand and said “you’re going to do what my officers tell you to do, do I make my self clear?” Calhoun responded, “Do what you’ve got to do.” Still clutching Calhoun’s throat, Avila said, “Now where are you going to go?” Avila placed his hands over Calhoun’s thumb area and applied pressure, asking “Does that feel good?” and escorted Calhoun to the back dock entrance. He maintained pressure on the thumb for 5 to 8 minutes waiting for the ambulance to arrive. He then motioned for Rader “to take over,” which he did. When Calhoun kicked backward at Rader two times, five other staff forced Calhoun to the floor into a prone position. While several staff were lying on top of Calhoun, with his legs crossed and pinned down, Avila “slapped Calhoun’s face approximately two times and said, ‘shut up.’ ” Leg restraints were then applied. When Calhoun subsequently looked at Avila, Avila said “go f— yourself.” Trial Ex. P-3095 at 83301-02. 90

The Captain in charge of IAD acknowledged that he had no reason to doubt the veracity of Hlebo’s account. Jenkins Tr. 3-379-380. When investigators interviewed Calhoun, he also substantially confirmed Hle-bo’s account. Trial Exh. P-3095 at 83228-29; Jenkins Tr. 3-383-84. Nonetheless, Avila was not found to have misused force in any part of this incident. Jenkins Tr. 3-379-384. Regarding the hpld on Calhoun’s thumb, the report concluded that it could not be determined “how much” pressure was applied. Trial Exh. P-3095 at 83236. Calhoun’s statement that his thumb “was yanked real hard,” the investigating lieutenant testified, was discredited as uncorroborated. Long Tr. 17-2828. Hlebo’s statement that pressure was applied was not considered corroboration because “you can’t see ... pressure being applied.” Long Tr. 17-2829. Nor did he consider Avila’s statement “does that feel good?” in assessing whether excessive force was applied. Long Tr. 17-2829.

Investigators did find that Avila slapped Calhoun, but concluded that it was fully justified by the situation because Calhoun “may not have been completely under control at the time.” The only evidence pointed to in support of this possibility is the fact that “they were still on the ground at the time of occurrence.” Trial Exh. 3095 at 83236. In reaching this conclusion, the report simply ignores the contrary description in Hlebo’s written report, which indicates that Calhoun not only was handcuffed, but also had “several staff lying on top” of him and his “legs were crossed and pinned down by someone’s feet” at the time of the incident. It also ignores Hlebo’s account that the two kicks by Calhoun before he was taken down were the only aggressive actions he made. Id. at 83255. The report also notes that Avila’s conduct “may have been” a “reflex” without offering any support for this conclusion. Even Avila had not offered this explanation; rather, he denied the conduct altogether. Defendants’ expert agreed that there seemed to be no support for the ■ conclusion that Avila’s action in slapping Calhoun might be a “reflex,” and stated that he would not have been satisfied if he had received this report. DuBois Tr. 29-4795-6. 91

Investigation re: Use of Force against Inmate Martinez

In October 1991, Antonio Martinez was extracted from his cell shortly after he threw his meal tray through the food port, hitting Officer Parson. As discussed earlier in these findings, Martinez suffered significant injuries, including the loss of four teeth and a 1.5 inch laceration to the back of his head. Trial Exh. P-3083 at 79081; P-1178 at 5598. Five months later, an investigation was initiated when Martinez filed a complaint that he had been kicked in the face during the extraction.

In the course of the investigation, Officer Nietschke, the control booth officer with a clear view of the pod area outside the Martinez cell, confirmed that Martinez had been kicked in the head or shoulder area after being restrained. Specifically, the investigative report states that “[Nietschke] observed an officer kick Inmate Martinez” once he had been taken from the cell, and recounts Nietschke’s opinion that the officer had “defi nitely used excessive force due to Martinez being restrained and under control at that time.” Nietsehke also indicated that he was unsure if the kick was to Inmate Martinez’ “upper shoulder or lower head area,” and that the inmates were yelling “you can’t kick a guy in the head like that.” Trial Ex. P-3083 at 79052, 70957-58.

The investigation concludes, however, that “no reasonable cause for misconduct” could be established with respect to any aspect of the incident. Id. at 79060. Nietschke’s statement that an officer had kicked a restrained inmate in the head or shoulder area is noted but then seemingly discounted because Nietsehke “offered that Sergeant Cox had a much closer viewpoint.” 92 Id. at 79058. Thus, regardless whether the force used inside the cell was justified, the IAD’s readiness to conclude that Nietsehke’s observations of excessive force should be ignored raises serious questions about the integrity of the internal affairs process.

Equally disturbing is the IAD’s conclusion that Officer Parson — at whom Martinez threw the tray — could not have been culpable because “all evidence indicates that Officer Parson did not participate in the subject cell extraction nor was he present inside the Unit Pod when the extraction occurred.” Trial Ex. P-3083 at 79057 (emphasis added). Yet, the available evidence overwhelmingly demonstrates that Parsons did, in fact, participate in the cell extraction: (1) the incident report lists Parson as a participant, (2) Lieutenant Carl, who supervised the extraction, told investigators during his taped interview that Parson participated in the extraction, (3) Sergeant Miller, also interviewed on tape, said that he was “sure” that Parson participated in the extraction, and (4) the MTA’s medical report reflects that Parson informed the MTA that he was involved in the cell extraction and was hit in the face by some mace. 93 None of these items are mentioned in the Internal Affairs report. • In reviewing the report, defendants’ expert concluded that the matter should have been investigated further, and that “the omission of the incident report and the omission of the MTA report raised a question in [his] mind regarding the integrity” of the investigation report. DuBois Tr. 29-4799-4800.

Finally, we note that despite a full-blown investigation and a plethora of eyewitnesses, the LAD was unable to explain how Martinez came to lose four teeth. While the report’s conclusion “speculates” that the shield caused this injury, Trial Exh. P-3083 at 79060, this speculation is not supported by the record. The shield officer states that, while he hit Martinez with the shield, there was no sign of blood in Martinez’ mouth at the time, and that he did not believe that the shield caused the injury. Another officer offered his opinion that the shield caused the injury, but admitted that he did not see the shield hit Martinez; nor did he observe blood on Martinez’ face until later. Trial Exh. P-3083 at 79046-47, 50.

Clearly there are instances in which responsible administrators may, despite reasonable efforts, be unable to determine exactly what transpired due to inconsistent recollections or genuine confusion regarding the events in question. However, at Pelican Bay, the failure by IAD to do so more likely reflects not only the code of silence, but also a general indifference to acts of excessive force, shared by officials at all levels of the prison. For example, the supervising lieutenant in the Martinez cell extraction candidly stated that “I was not really concerned about how they [the teeth] got knocked out ...” Trial Exh. p-3111 (Carl 6/26/92 tape recording). Nor, according to the lieutenant, was the matter ever raised with him “by management.” Id. 94

e. Discipline

In the event that the IAD sustains an allegation of misconduct, the Warden has a number of disciplinary measures at his discretion. The most lenient form of adverse action is an official letter of reprimand which remains in the officer’s file for three years and may affect opportunities relating to assignments and other matters. Other more severe adverse actions include a percentage reduction in salary for some period of time, suspension, and dismissal. 95

Given the lapses in supervision, and the unlikelihood that the IAD will find a misuse of force in any given case, officers rarely face the prospect of discipline for using excessive force against inmates. As noted above, the record indicates that there have only been four instances since Pelican Bay opened in December 1989 in which the IAD concluded that excessive force had been used. In three of these cases the Warden exercised his discretion to minimize or eliminate any adverse action. A review of two of them illuminates the approach to discipline for misuse of force at Pelican Bay.

Officer Plumlee

Based on the eyewitness reports of two other officers, the IAD found that Officer Plumlee had struck an inmate twice in the face with a closed fist while the inmate was already restrained. This December 1990 incident is the only matter brought to the Court’s attention where the IAD found that an officer had engaged in a relatively major misuse of force. It is also clear that Plumlee was less than candid during the investigation. In his taped interviews, which Warden Marshall reviewed, Plumlee admitted only to hitting the inmate somewhere in the upper body or upper left bony portion of his head as a “reflexive” action, accused the other officers of falsely reporting what happened, and suggested that the inmate inflicted the injuries on himself. Trial Ex. P-3103 (tape recording).

Warden Marshall initially recommended a five percent salary reduction for 12 months. However, after meeting with Plumlee during the course of the disciplinary proceedings, the Warden dismissed all adverse action in the “interest of justice.” Trial Ex. P-3087 at 77460. The Warden explained that this decision was based on the fact that Plumlee had shown remorse (though he still contended that the incident did not occur as stated in the IAD report), and had expressed concern about his career and his upcoming participation in Operation Desert Storm as a reserve member of the Marine Corps.

As plaintiffs’ expert Nathan observed, where an officer is “caught red-handed in a serious assault of a prisoner” and “openly lied about it during the course of the investigation, ... any warden attempting to maintain lawful conditions within a prison has no choice but to impose maximum disciplinary sanctions under these circumstances.” Nathan Decl. at 89. Defendant Gomez also agreed that “aggressive and strong action” should be taken when excessive force occurs. Tr. 28-4651. Yet, all adverse action against the officer was dropped. In so doing, “the warden of [Pelican Bay] sent a dramatic and unmistakable message that excessive force, even in those rare cases in which it is proved to the Internal Affairs investigator’s satisfaction, will not lead to punishment.” Nathan Decl. at 89. 96

Officer Rader

Based primarily on the eyewitness report of Officer Hlebo, Internal Affairs sustained the allegation that Officer Rader used excessive force against inmate Jesse Calhoun by slamming his head into the wall during an escort:

Although by talking and turning his head during the escort inmate Calhoun failed to abide by SHU escort guidelines, it also appears that he was somewhat provoked into doing so by Officer Rader. Officer Rader’s reference to Calhoun as “a f — king child molester” and his apparent willingness to exchange derogatory terms throughout the escort, did nothing to encourage cooperative behavior from Calhoun and is inconsistent with professional expectations. It does not appear that inmate Calhoun demonstrated any aggressive movement which would have justified slamming his face into the wall as witnessed by officer Hlebo_ The allegation that Officer Rader applied excessive, force against inmate Calhoun is therefore “SUSTAINED.”.

Trial Exh. P-3095 at 83256. Notwithstanding the above, the IAD concluded only that Rader “may be” in violation of a Directors Rule requiring courteous and professional dealings with inmates and a government code section which prohibits “other failure of good behavior ... [that] causes discredit to the agency for which an employee works.” Id. at 83256-7. The Warden exercised his discretion to sanction Officer Rader only for his unprofessional verbal behavior, and only by issuing a “letter of reprimand,” the lowest level of disciplinary action. Marshall Tr. 22-3743; Trial Exh. D-ll. The Warden did not recall having considered a more serious disciplinary action. Marshall Tr. 22-3744;

The failure'to discipline Rader for his misuse of force was unexplained by defendants at trial, leaving the clear impression that it was not considered improper or objectionable. Notably, defendant Gomez testified that “generally, I am comfortable with what is happening at Pelican Bay” with respect to discipline. Tr. 28-4653. The clear signal to staff is that misuse of force against inmates will not yield significant adverse consequences. 97

The Court does not intend to suggest that special circumstances may not warrant leniency in a particular case or that a Warden should not be permitted to exercise his or her discretion in this regard. What the rec,.ord reflects at Pelican Bay, however, is an institution that lacks serious commitment to disciplining or controlling the behavior of staff who misuse force against inmates.

3. Defendants’ State of Mind

As found in sections 11(A)(1), and (2), supra, a breakdown in the systems necessary to control the use of force has allowed a pattern of excessive force to develop and persist at Pelican Bay. In assessing defendants’ knowledge and state of mind, with respect to these findings, we take into account the fact that Pelican Bay is a new prison. As in any new undertaking of this scale and complexity — and particularly in the pressurized atmosphere of a high security prison — some trouble spots and errors are simply unavoidable. As defendant Gomez testified, “procedures are typically ragged when we start a new institution.” Tr. 28-4626. In the case of Pelican Bay, this factor was exacerbated because the prison opened three or four months ahead of schedule due to population pressures within the prison system. There may also be a tendency on the part of prisoners to “test” a facility when it first opens.

However the fact that a prison may be new does not excuse defendants’ obligation to operate it in a constitutionally acceptable manner, an obligation which defendant Gomez testified he understood. Tr. 28-4629. Thus, we have attempted to discern to what extent the pattern of excessive force and breakdown in the systems for controlling use of force can be attributed to the “growing pains” of a new facility (that is, good faith errors or mistakes), and to what extent defendants were not only aware of the problems and the consequences, but deliberately chose to ignore them. We conclude that while the newness of the facility may explain some of the problems identified in the findings above, defendants are largely culpable for the pattern of excessive force at Pelican Bay.

Having carefully reviewed the record, it is clear that defendants were aware that there were serious problems concerning excessive force at Pelican Bay. 98 As Martin found, “the incident reports, internal affairs reports, and other reports and memoranda received by defendants clearly reveal the existence and extent of [the] misuse of force. Defendants accordingly knew of [the] violence....” Martin Decl. at 5. Nathan also testified that “[t]he reports the defendants receive reveal that staff use unnecessary and grossly excessive force against inmates on a regular and frequent basis,” and that “the prevalence of the misuse of force at [Pelican Bay] would have been apparent to any even marginally competent prison administrator who reviewed the documents and information the defendants receive.” Nathan Decl. at 12-13. We also find that the continuing and substantial risk of serious injury to inmates in a prison where misuse of force is prevalent is so obvious that defendants did, in fact, know of this risk.

The record also demonstrates that this risk was consciously disregarded, evincing, at the very least, an attitude of deliberate indifference. Although defendants have ceased some of the practices complained of by plaintiffs, 99 such changes, which post-date the fil ing of this class action, were likely motivated by this litigation, and at least as of the time of trial, had not been cemented in any formal written policy. As such, they may well be transitory in nature, and the Court is not persuaded that such changes would not be undone in the absence of court intervention. Nor has any prison official ever suggested that such changes were made to address problems concerning the use of excessive force. On the contrary, defendants never acknowledged that there was a genuine problem to be addressed and always offered other reasons to explain these changes in practice. Moreover, defendants never offered the Court any firm or clear assurances that such changes would be permanent. Accordingly, the Court is not convinced that these recent changes represent a serious commitment by defendants to end the pattern of excessive force.

Rather, the great weight of the evidence indicates that the misuse of force against inmates was something that prison administrators preferred to disregard or ignore. Although defendants acknowledged that regulation of the use of force is important if abuses are to be minimized or avoided, they made no serious effort to operate the prison in a manner that would effectively regulate and control the use of force. As detailed in section 11(A)(2) above, defendants failed to provide (1) clear and authoritative use of force policies, (2) any meaningful supervision of the use of force, (3) a bona fide investigatory process into allegations of misuse of force, or (4) consistent imposition of discipline in those cases in which misuse of force was found. While a failure in one area might not raise any particular inference, the glaring deficiencies in all of the above areas convinces us that such deficiencies are not accidental but the result of deliberate indifference. 100

This indifference was underscored when senior administrators were questioned about particular incidents. For example, when the Warden was questioned about the circumstances surrounding an inmate who suffered a broken jaw, he evinced no concern that neither he nor the investigative report could satisfactorily explain why there had been “blood on the floor.” See note 94, supra. Similarly, when the Chief Deputy Warden was questioned about an incident report that all experts agreed should be investigated, he saw nothing of concern. See section II(A)(2)((c)(l), supra. Given all of the above, we readily find that defendants were deliberately indifferent to the risk of serious injury to inmates.

Plaintiffs’ experts also convincingly testified that the degree of excessive force found at Pelican Bay and the deficient systems to control the use of force reflect a management practice that is designed to inflict unnecessary pain and suffering. Fenton, for example, stated that “this is the first I had ever heard of an administrative organization where prisoners were, on a fairly systematic basis, cruelly treated as an administrative device. I’ve never seen ... that before.” Tr. 5-734. Similarly, Nathan described Pelican Bay as “a lawless, violent place” where “defendants have knowingly allowed grossly inappropriate use of force to occur as a deliberate management policy.” Nathan Tr. at 13-2051; Decl. at 13. See also Martin Tr. 8-1377 (“they’ve got some folks out there obsessed ... with pain and with sending this ... message to the rest of the system: that you will hurt and you will experience sheer pain if you come to Pelican Bay”). 101

We agreé that the extent to which force is misused at Pelican Bay, combined with the flagrant and pervasive failures in defendants’ systems for controlling the use of force reveal more than just deliberate indifference: they reveal an affirmative management strategy to permit the use of excessive force for the purposes of punishment and deterrence. For example, when defendants manifest no concern that the SHU has no current official operating policy, when they fail to explain why SHU control booth officers are not provided with gas guns as a non-lethal alternative to rifles, when they let highly suspicious incident and investigative reports go unchallenged, and when they promote the code of silence by failing to support those who come forward, they lead us to conclude that they have implicitly sanctioned the misuse of force and acted with a knowing willingness that harm occur. Of course, these points only touch on some of the evidence discussed in the Court’s findings that bears on this point. All together, it paints a picture of a prison that all too often uses force, not only in good faith efforts to restore and maintain order, but also for the very purpose of inflicting punishment and pain.

B. MEDICAL HEALTH CARE

Plaintiffs contend that Pelican Bay has a constitutionally inadequate system for delivering medical care. At trial, the Court heard testimony from two expert witnesses. Dr. .Armond Start, testifying for the plaintiffs, based his opinion on an extensive examination of Pelican Bay’s medical care system. In addition to touring the prison and interviewing both inmates and prison personnel, he has reviewed over 130 prisoner medical records 102 and a random sample of 3000 sick call slips. He also examined the depositions of Department of Corrections staff and read numerous other documents such as grievance forms and records from hospitals and other third party health care providers. In all, Dr. Start spent more than 300 hours evaluating the health care system at Pelican Bay. Start Tr. 11-1703.

Dr. Jay Harness, the defendants’ expert, evaluated Pelican Bay’s delivery of medical care by touring the facility, reviewing documents and depositions, including that of Dr. Start, and reviewing eight prisoner medical records. Dr. Harness testified that his entire evaluation took around thirty or thirty-one hours. Harness Tr. 19-3089-90. 103

The Court also heard testimony from Na-dim Khoury, the Assistant Deputy Director for Health Care Services for the California Department of Corrections, Kyle McKinsey, Deputy Director for Health Care Services with the Department of Corrections, and Pelican Bay physician Dr. David Cooper. Nurses, the head Medical Technical Assistant, and several inmates also presented testimony, and the Court received deposition testimony from other health care providers at Pelican Bay. The Court also carefully reviewed well over a thousand pages of documentary evidence submitted by both parties.

The evidence before the Court compels us to find that the medical care system at Pelican Bay does not meet minimum constitutional standards. We agree with plaintiffs’ expert, Dr. Start, who concluded that “the entire system is grossly inadequate and unsatisfactory in meeting the health care needs of the inmate population. Indeed, ... [it is] deplorably inadequate.” Start Deck at 4. 104 As described below, the record reveals systemic, unremedied deficiencies in the system for delivering health care at Pelican Bay which render that system incapable of meeting inmates’ serious medical needs. Moreover, we find that the evidence reflects defendants’ deliberate indifference to those needs.

1. Serious Need for Medical Services

There is no doubt that inmates at Pelican Bay have serious medical needs. Like the population at large, prisoners entering the facility suffer from diseases such as asthma, hypertension, epilepsy, diabetes, tuberculosis and lupus. Once at Pelican Bay, inmates experience the full spectrum of medical problems, ranging from the routine to the life-threatening, including loss of hearing, abdominal pains, fractures, kidney stones, lacerations and gunshot wounds. Dr. Cooper, a physician and surgeon at Pelican Bay, stated that inmates experience more serious levels of illness than patients he sees in private practice. Cooper Tr. 14-2259. In addition, many serious health disorders are overrepresented in the prison population. Start Decl. at 11 (seizure disorders, diabetes, asthma, chronic obstructive pulmonary disease, trauma, hypertension and cardiac problems are disproportionally present, as are risk factors for communicable diseases). In addition, inmates in the SHU generally need more medical care than those in the general prison population. Khoury Tr. 10-1588. Inmates clearly have medical needs that are genuine, frequent, and serious.

2. Systemic Deficiencies in the Delivery of Health Care

a. Staffing Levels

Both sides agree that the presence of sufficient, qualified medical staff is indispensable to the provision of adequate medical care. However, Pelican Bay has from its opening operated without enough doctors and properly trained and supervised medical personnel to meet the needs of the inmate population. This numerical inadequacy contributes significantly to the failure of the medical system as a whole.

Dr. Nadim Khoury, the former Chief of Medical Services for the CDC, was called as an adverse witness by the plaintiffs. He confirmed that CDC policy is to require a ratio of 1 physician for every 550 inmates. Khoury Tr. 10-1587. Notwithstanding this established policy, of which the Warden was well aware 105 , defendants elected to open the prison in December of 1989 without a single physician on staff. Instead, when Pelican Bay opened, the only physician available to inmates was a Crescent City emergency room physician who worked at the facility one day per week. Astorga Depo. at 28. Since then, the number of physicians and other medical personnel on staff has grown, but has not kept pace with the needs of the inmates. For example, by the end of January 1990, one M.D., five registered nurses, and sixteen Medical Technical Assistants (“MTAs”) 106 had been hired for a prison population of approximately 1,300, a physician/inmate ratio of 1:1,300. By January of 1991, the staff had increased moderately to include three physicians, six nurses, and 24 MTAs, but the inmate population had grown to over 3,500 inmates. Thus the grossly inadequate physician/inmate ratio remained almost constant at approximately 1:1,166. These conditions amounted to an “extreme shortage of staff,” such that medical personnel were “spread too thin to really be able to give very much individual attention to inmates.” Lara Depo. at 35, 73. Even the defendants’ medical expert, Dr. Harness, admitted in exquisite understatement that the initial staffing levels were “incomplete.” Harness Decl. at 11.

Support staff repeatedly voiced to supervisors their concerns about insufficient staffing; one MTA testified that she remembers asking “[e]verybody [ — ] [sjenior MTAs, doctors [ — ]” for more MTAs to be hired. Golli-har Depo. at 63. A supervising nurse testified that in 1992 she needed and requested more nurses “to deal with the increased acuity of the inmates,” but her request was denied because of a hiring freeze. S. Bliesner Tr. 26-4262-63. In fact, physicians openly referred to staffing shortages to justify inadequate care. For instance, after one inmate complained after a seven month delay in removing his cysts, Dr. Astorga answered that the facility was “125% short of doctors.” Start Decl., Exh. E at 7091. The lack of staff has had predictable effects, from delays in medical treatment, discussed below, to tragic oversights, as when the medical staff “missed” a inmate’s ruptured appendix. Ruble Depo. at 63.

By January 1, 1993, over three years after the prison opened, there were five doctors, seven nurses (and one open position), and 26 MTAs providing medical care for 3898 prisoners, a doctor/inmate ratio of approximately 1:780. By the time of trial another doctor had been added, yet the medical staff is still unable to serve the sheer number of inmates who need medical-attention. As one MTA notes, it is “almost impossible to keep up with the demand of patient care services.” Simmons Depo. at 49. MTAs and doctors, including Dr. Astorga, agree that Pelican Bay is still “short'of physicians,” Astorga Depo. at 100, and some doctors have expressed a wish for more doctors, more nurses, and more MTAs. As Dr. Start concluded, Pelican Bay “continues to have ... an inadequate number of health care staff to provide necessary services to the inmate population there.” Start Deel. at 5. In combination with Pelican Bay’s other medical shortcomings, understaffing clearly contributes to the systemic failure of medical services,

b. Inadequate Training and Supervision

What medical staff does exist must be properly supervised and trained in order to be effective, but “medical training at Pelican Bay is virtually non-existent and supervision is woefully deficient.” Start Decl. at 92. Dr. Start testified to the importance of prison staff staying updated on changes in health care management. Although facilities can offer in-house (“in-service”) programs on topics such as management of emergencies or tuberculosis, “[t]here is none of that in existence at Pelican Bay.” Start Tr. 11-1742. Pelican Bay medical personnel at all levels have identified this lack of training as a problem. MTAs have repeatedly requested additional training, one even going so far as to file a grievance protesting the lack of continuing medical education. Carter Depo. at 182-83. Dr. Gard, a physician at Pelican Bay, recognized a ongoing need to train MTAs to handle emergency situations. Gard Depo. at 71.

Particularly noteworthy is an absence of “drills to practice emergency care and. [a lack of] instruction given for basic emergency procedures which is particularly disturbing because of the frequency of trauma.” Start Deck at 96. The need for training in emergency procedures is especially clear in light of instances in which MTAs have mishandled emergency situations. Dr. Start highlighted several such examples; in one instance, an MTA waited until inmate Roger Hernandez was carried to a clinic on a gurney before CPR was initiated. Carter Depo. at 95-96; Trial Exh. P-3053 at 32679. In several other eases, MTAs improperly treated inmates who were in shock. See Start Deck at 98-102.

Even more troubling than the absence of training programs is the basic lack of physician supervision of MTAs. MTAs play a critical role in inmates’ medical treatment by performing initial triage. 107 When inmates need medical assistance, they fill out a sick call slip which is collected from a central location once a day by an MTA. 108 The MTA then reviews the slips to determine whether the illness requires emergency treatment or can be treated at a later time. Unbelievably, MTAs perform this vital triage function without an organized form of supervision. 109 Dr. Cooper, a physician and surgeon at Pelican Bay, testified that he was unaware of any protocol that “specifies any way to review or supervise the performance .of an MTA in triaging medical slips.” Cooper Tr. 14-2308. In fact, there is very little supervision in general: Dr. Start concluded that there is “no evidence that the physicians who are ultimately responsible for what kind of care is delivered ... are supervising the nurses and, more importantly, the correctional medical technicians, the MTAs.” Start Tr. 11-1742-43.

Clearly, there is a need to supervise medical staff; even defendant’s expert Dr. Harness agreed that “physicians need to be monitoring what MTAs are doing.” Harness Tr. 19-3102-03. The record is replete with instances in which MTAs inappropriately re fused to refer inmates to doctors or exceeded the scope of their competence. 110 For instance, one inmate complained to an MTA of ear pain and hearing loss. Despite being unqualified to do so, the MTA nevertheless examined the inmate’s ears, Start Decl. at 373, and noted that one ear canal was completely occluded with wax. However, she did not refer the inmate to a physician or for treatment. When the inmate finally saw a physician more than a week later, the physician was unable to examine the eardrum because the ear canal was . still occluded with wax. Trial Exh. P-637 at 27624. Another inmate was told by two MTAs that he needed no further treatment for a facial fracture when, in fact, the fracture actually required surgery. Trial Exh. P-430 at 2968-69, 2947. In another case, when an inmate exhausted his asthma inhaler before his prescription ran out (usually a sign that asthma is worsening and an indication that the patient should be seen by a physician), an MTA chose to “counsel” the inmate simply not to use the inhaler so frequently. Start Decl. at 200; Trial Exh. P-535 at 16118, 16124. There is no supervisory process in place to correct these errors or prevent them in the future.

c. Medical Records

The medical records system at Pelican Bay is nothing short of disastrous. Accurate and complete medical records are essential to adequate medical care. Providers must know the patient’s medical history, allergies, medications, and past courses of therapy in order to properly diagnose and treat current problems. Without accurate and thorough records, providers continually run the risk of prescribing contraindicated medications, failing to notice ongoing illnesses, or ordering inappropriate or even dangerous courses of treatment. Despite these dangers, and defendants’ knowledge of them, the Pelican Bay medical records system “is outrageously disorganized, making it almost impossible to understand what is happening to the patient, which in turn prevents the inmate from obtaining health care.” 111 Start Decl. at 89.

Several problems contribute to the utter failure of the medical records system. First, recordkeeping personnel at Pelican Bay are both too few and insufficiently trained. Even though the task of maintaining medical records is onerous and complex, records staff receive no specialized instruction beyond on-the-job training. In addition, there are simply too few people on staff to oversee the records of over 3,500 inmates. 112 Even Dr. Astorga stated that he was unsatisfied with recordkeeping personnel levels. Astorga Depo. at 135.

Second, patient records are stored in a central recordkeeping area separate from where inmates are examined in satellite clinics. As a result, records are often delayed. 113 Providers are often forced to risk treating patients without consulting their medical records at all — a practice, in the words of one MTA, tantamount to “flying by the seat of your pants.” Carter Depo. at 182. 114

Third, and most important, the notes that have been made in patient records by physicians and medical support staff are disorganized, incomplete, sometimes contradictory, and inadequate. As Dr. Khoury acknowledged, it is a basic, fundamental principle of medical practice to document everything the provider does. Khoury Tr. 10-1636. Nevertheless, the record is replete with examples of charts without medical histories, with no record of examinations, no management plan, orders for tests with no record of results, test results with no record of why, when, or by whom the test was ordered, and so forth. 115 Even aside from the shortcomings of each individual entry, the entire system “ought to be better organized,” as defendants’ expert admitted. Harness Tr. 19-3110. There is no uniform note taking format, no system for correlating physicians’ orders and progress notes, and no auditing of medical records, despite a warning in a 1991 audit that recordkeeping audits and better record managing were needed. Trial Exh. P-3334 at 32533. These ongoing problems led Dr. Start to declare that he was “impressed over and over again with the gross deficiency of the • records system.” Start Decl. at 89.

d. Screening

By examining inmates as they enter the facility, providers can identify those patients who need uninterrupted medication, catch prisoners’ previously unnoticed medical problems early on, and discover potential medical emergencies among newly-arrived inmates. Providers can also prevent from being admitted to the prison’s general population those who pose a threat to the health and safety of others (such as inmates with communicable diseases).

Despite the importance of initial health screenings, Pelican Bay has failed to provide consistent or meaningful screening of incoming prisoners. First, physicians are not involved at all in initial screenings; nurses examine the medical records of arriving inmates if the records are available, and only MTAs screen inmates in person. Astorga Depo. at 99-100. Usually inmates simply answer questions without being actually examined. Dr. Astorga admitted that there is no practice of conducting routine physical examinations at the prison. Astorga Depo. at 99-100. Dr. Start found telling evidence of'inadequate screening when he surveyed 130 patient files 116 : only ten percent of inmates transferring into Pelican Bay had an adequate medical history taken 117 , thirty-four percent had no intake history performed at all, and a mere four percent had any kind of physical examination or assessment. Start Deel. at 20-21. Such superficial intake screenings plainly cannot elicit accurate or complete medical profiles of incoming inmates.

In addition, Pelican Bay fails to provide adequate screening for communicable diseases commonly found in prison populations. For instance, tuberculosis (“TB”) is particularly dangerous in a prison environment, where overcrowding and poor ventilation can hasten the spread of this airborne disease, infecting prisoners, health care workers, correctional staff, visitors, and the surrounding community. Screening for TB is a rudimentary public health measure, and can be done with a simple skin test. 118 However, there was no screening for TB prior to May of 1992, when the Department of Corrections finally mandated testing in all California prisons. Even after Pelican Bay was forced to conduct TB screening, it did so in a slipshod fashion, and then failed to follow up on the results of testing. For example, the staff did not consult medical records to determine whether inmates had previously tested posi-five. 119 Thirty-two percent of the files examined by Dr. Start showed no evidence of TB testing, even though all the inmates in question were at Pelican Bay during the mandated screening. Start Deck at 24. Of those prisoners in Dr. Start’s sample who tested positive, over half were never treated either for active TB or with preventative antibiotics, the standard treatment for latent tuberculosis. 120 Start Tr. 11-1708-10; Trial Exh. P-5614.

Defendants also fail to test for syphilis, another communicable disease, and do nothing to sponsor educational outreach to prisoners about AIDS, or to encourage voluntary HIV testing, despite the high incidence of AIDS in the prison population. 121 . Seemingly the only improvement in this regard has been the addition of an infection control nurse, who supervises hepatitis B screening and immunization for the staff and yearly TB screening. In sum, defendants’ screening for TB and other communicable diseases has been and still is grossly inadequate,

e. Access to Medical Care

Inmates must be afforded access in a timely fashion to medical providers who are qualified to treat their illnesses. However, prisoners at Pelican Bay often experience significant and unnecessary delays in obtaining access to physicians. In many instances, they are denied access altogether. For in mates with serious or painful symptoms, delays lasting days or even weeks can cause unnecessary suffering, exacerbate illness, and have life-threatening medical consequences.

As discussed above, inmates who want medical care submit sick call slips, which are then read and analyzed by MTAs. MTAs determine whether and when the inmate will be allowed to see a physician — often solely on the basis of what is written on the sick call slip. See E. Thayer Tr. 25-4204. If the MTA feels that the inmate should see a physician, the inmate is placed on a “doctor’s line,” the rough equivalent of having an appointment. Thus, MTAs function as a “gatekeeper” through which inmates must pass before they can have access to a doctor. Yet, as discussed above in section 11(B)(2)(b), supra, MTAs have insufficient training and supervision to perform this vital function. Moreover, Pelican Bay has no written protocol or triage training to help MTAs determine who needs- to be evaluated by a nurse or physician or how urgently care is needed. As one MTA put it, the decision whether to send an inmate to a clinic is “pretty much ... left to our judgment.” Griffin Depo. at 16.

The record shows that, over and over, MTAs have inappropriately used that judgment to deny prisoners access to medical care. Prisoners complaining of symptoms as serious as chest pain, severe abdominal pain, coughing up blood, and seizures are often made to wait for regular appointments or denied access to a physician altogether. For instance, MTA Griffin, who described her own position at Pelican Bay as “sort of a glorified delivery person,” Griffin Depo. at 17, recounted what MTAs do when prisoners have seizures: “We monitor them and if we feel that they need to be sent, then they’re sent. Otherwise, if they’re alert, oriented and their vital signs are stable and they haven’t voided on themselves, then we just let them be and tell them to get in touch with us if they have any problems.” Id. at 41. 122

An example of MTA failure to refer seriously ill inmates to a physician is the case of Ralph Burke. At 2:00 a.m. on November' 1, 1992, Burke notified an MTA that his back hurt and he was having trouble breathing. The MTA gave him some ibuprofin, an over-the-counter pain reliever, but refused to take him to the infirmary. At 4:80 a.m., Burke told the MTA that his neck hurt and that he could not move. The MTA still refused to take him to the clinic. Half an hour later the MTA noted that the inmate was “sleeping,” but at 5:45 a.m. noticed that Burke was “breathing in a snorting mode” and took him to the infirmary for evaluation. Trial Exh. P-405 at 36840. Although Burke was semiconscious and paralyzed, repeatedly blurting out “help me,” MTAs and the infirmary nurse were convinced that he was “faking it.” Id. at 36838. When Burke was finally taken to the hospital after 7:00 a.m., he was diagnosed with an intercranial hemorrhage; he entered a coma and died shortly thereafter. Diagnosing Burke’s symptoms exceeded the bounds of MTA expertise, and by refusing to refer Burke to the infirmary the MTA effectively denied him appropriate medical care.

Sergeant Cox testified about another disturbing instance in which an MTA effectively denied an inmate timely access to appropriate medical care. Cox responded to an alarm for an inmate fight and arrived to see an inmate bleeding profusely with bruises on his neck. It took an MTA 18 minutes to arrive on the scene after being called on the radio. Cox suspected that one of the inmates had been raped; he testified that he told the MTA it was “obvious” that “the guy’s probably been raped.” Cox Tr. 18-3004. Despite Cox’s protestations, the MTA refused to examine the inmate or refer him to a doctor and merely wiped up the inmate’s blood. Sergeant Cox testified that he went to the watch commander: “I explained to the watch commander, hey I even did a — performed an unclothed body search on this man. There was fluids coming out of his rectum that wasn’t supposed to be. And all she did was put the inmates in ad seg [administrative segregation].” Cox Tr. 18-3005.

Even when inmates presenting serious medical problems are put on the doctor’s fine by MTAs, prisoners experience delays ranging from significant to appalling before they actually see a physician. Understaffing has created a constant backlog of inmates vying for appointments. For instance,- in 1991, when Pelican Bay was particularly short of doctors, inmates waited to see a doctor for as long as four to six weeks. Inmate Arturo Castillo’s experience exemplifies the outrageous delays typical of Pelican Bay’s early years. After suffering a serious scalp laceration, Castillo was treated with surgical staples at Sutter Coast Hospital and then returned to his cell after a week’s recuperation in the prison infirmary. Castillo subsequently told an MTA that his wound had become painful, dirty, and itchy, and even filed a grievance, but the MTA merely told him he could see a doctor in two weeks. Castillo received no medical attention at all until weeks after he complained, when a piece of his scalp finally became so severely infected that it fell off. Castillo Tr. 1-102-04; Trial Exh. P-667 at 4407-08, 4415.

Although improved staffing levels have reduced delays in access to physicians, such delays still pose a significant problem. MTA Ruble testified that by 1992 medical staff had “got it down to the point where we were running two weeks and sometimes one week” for an appointment. Ruble Depo. at 59; see also Elliott Depo. at 57 (one to two week wait in December 1992). As late as July 26,1992, there were 242 prisoners on the waiting list to see a doctor. Start Deck, Exh. U at 8905. While it is impossible to discern from the record how long the average delay in treatment was at the time of trial, 123 the record is still filled with examples of unacceptable- delays in access to physicians and treatment. 124

For instance, although inmate Zeke Cooper’s jaw was broken on October 31,1992, an MTA did an assessment and simply sent him back to his cell. The next day he complained of pain and again an MTA refused to refer him to a doctor. He was not seen by a doctor or X-rayed until two days after his jaw was broken. Start Deck at 244, Trial Exh. P-430 at 2968-9, 3014. In another typical case, inmate Louie Lopez testified that he waited approximately three weeks to see a physician about his bleeding hemorrhoids. L. Lopez Tr. 1-58. The examples above are not isolated instánces. Rather, the record overwhelmingly demonstrates that Pelican Bay has simply and utterly failed to provide a system in which serious medical problems are regularly treated in a timely fashion.

Inmate access to appropriate treatment is even further impeded by delays in lab testing. The record is rife with examples of lab tests that are ordered but never performed, performed only after unexplained and lengthy delays, or performed and never reported. A particularly notable example of delayed lab testing was observed by Dr. Start. Of the eight inmates who had positive tuberculosis tests in the sample of records he examined, 125 the average amount of time before performance of a chest X-ray (which would indicate whether inmates had active TB and were contagious) was 47 days. Start Tr. 11-1710. This was so even though Pelican Bay had an X-ray machine on site. Id. Diagnosis and treatment of inmates’ conditions is pointlessly deferred because of the prison’s inadequate system for ordering, performing, and reporting lab tests.

Finally, prisoners’ access to emergency treatment is impeded by both lack of expertise on the part of medical staff and custody concerns. As Dr. Start observed, “[o]ne principle of basic emergency medicine is that the difference between whether one saves or loses the patient depends on what treatment is given quickly at the scene of the accident.” Start Decl. at 98. However, as discussed above, there are no protocols for handling emergencies at Pelican Bay, and MTAs receive virtually no training in emergency techniques or handling trauma. 126 In addition, access to emergency treatment is often delayed for significant periods because a transportation .team or chase car is unavailable to follow ambulances leaving Pelican Bay. 127 Thus the ambulance transporting inmate Ricky Hurtado, who had been stabbed in the neck and shoulder, was delayed for twenty minutes before it could leave Pelican Bay. Start Decl., Exh. G at 1158; Ray Norris, having sustained a severe head injury, endured a twenty-eight minute delay; id. at 57; Roman Davis, experiencing respiratory difficulties, suffered a forty minute delay while waiting for a transportation team and paperwork; id. at 969, 971; and so forth. These life-threatening delays in delivering emergency services because custody staff is disorganized or understaffed are constitutionally unacceptable.

f. Lack of Quality Control Procedures

Although the quality of medical care provided to inmates at Pelican Bay often falls dramatically below community standards, medical staff and administrators have taken no effective steps to systematically review the care provided or to supervise the physicians providing it.

As Dr. Khoury noted, peer review 128 is a “very important” way for practitioners to “review and upgrade care they provide.” Khoury Tr. 10-1625-26. However, at the time of trial Pelican Bay had no formal peer review process through which physicians could review and discuss each others’ work. Although defendants insist that they have a “peer review committee,” this committee merely performs a function known as “utilization review” — a process of approving or denying requests for particular medical procedures, such as consultation or surgery. 129 See Cooper Tr. 14-2279-2282; McKinsey Tr. 26-4291-92. This review, in essence a cost/benefit analysis, is a far cry from the mutual evaluation and learning process ordinarily called peer review. It is no wonder, then, that defendants’ own expert, Dr. Harness, could not “recall” a peer review process at Pelican Bay. Harness Tr. 19-3117.

Another basic procedure that helps medical staff learn from experience and avoid fatal mistakes is the performance of a “death review,” an investigation and report on each death that occurs in custody. However, the medical staff at Pelican Bay does not conduct death reviews. This is the case even though Dr. Astorga, the Chief Medical Officer, thinks death reviews would be a “good idea,” and testified that he saw no reason, administrative or budgetary, why they could not be performed. Astorga Depo. at 721.

Of most concern is the fact that Pelican Bay has no formal quality assurance program. 130 Dr. Harness agreed that quality assurance is “standard practice in virtually any health care facility in the country” and a “fundamental part” of the provision of health care. Harness Tr. 19-3117. Although the Pelican Bay medical staff has organized a Quality Control committee, Dr. Cooper admitted that at the time of trial it had not yet met. Cooper Tr. 14-2282. Defendants argue that quality control measures are planned for the future, and offered the testimony of Kyle MeKinsey 131 to that effect. McKinsey testified that a Quality Program Unit will be charged with three tasks, one of which is “to insure that in the field over time we have quality assurance programs in place in all of our institutions.” 132 MeKinsey Tr. 26-4291. McKinsey projected that it would take until 1995 for the Quality Program Unit merely to develop formal standards against which medical care can be evaluated, and even longer to actually implement the program. Id. at 4317-18. In addition, McKin-sey stated that he hoped to have formal peer review instituted as part of quality assurance “eventually, resources being available.” Id. at 4292. Thus, Pelican Bay has yet to implement quality assurance within the facility itself, and the Department of Corrections has barely set in motion the machinery that may or may not someday yield effective quality assurance programs.

Failure to institute quality control procedures has had predictable consequences: grossly inadequate care is neither disciplined nor redressed. For instance, one physician was reprimanded by the Medical Board of California, which stated in a 1992 letter that the history and physical examination he performed on one inmate “were of such brevity as to not demonstrate a level of care that is considered within the community standard in the State of California.” Trial Exh. P-553 at 6890. Although Dr. Astorga testified he did not recall receiving a copy of the letter, he did remember one of several complaints by Pelican Bay staff that the doctor appeared to be intoxicated on the job. Astorga Depo. at 597-98. Dr. Astorga took no disciplinary action other than talking to the physician.

Similarly, a system for review of the numerous avoidable inmate illnesses, as well as inmate deaths, would have underscored the systemic deficiencies in the Pelican Bay health care system. For example, the care received by Tyler Henderson displayed, in Dr. Start’s words, “a long, and well-documented history of neglect, inappropriate evaluations, and sub-standard care” that led to his death at age 24. Start Deck at 53. When he arrived at Pelican Bay in August of 1990, Henderson did not receive his seizure medication for several days, even though he had a cyst on his brain and a significant seizure disorder. Dr. Start characterized Henderson’s treatment as reflected in his chart:

Seizures are destructive to the brain, and, except for very rare circumstances, preventable and controllable. If a treatable spell of seizures continues uncontrolled, the patient dies. Even very frequent, but self-limited, seizures have serious repercussions — black-outs, trauma to the patient, the risk of aspiration pneumonia, and short- and long-term memory loss. No one knows exactly how many seizures is too many....

In the case of Tyler Henderson, this patient was having prolonged periods of as many documented seizures as two a day, or three or four a week ... [h]e was having frequent, recognized seizures that were not being treated. He was often just left in his cell after a seizure, or taken to the infirmary just until he woke up, then' returned to his cell. In multiple instances, he was only evaluated by an MTA, and no drug levels were drawn. [The] dosing of his medications and cheeking of blood level was haphazard, at best ... Poignantly, he had appealed for transfer on the basis of inadequate medical care July 29, 1991— citing “occasional blackouts.” His appeal was denied....

Patients who have difficult to control seizures may need more than one medicine, higher levels of a specific medicine, or may be resistant to one class of medicine. Tyler’s medication dosing when he came to PBSP suggested that he could have difficult to control seizures because he was on two types of medicines, and one of the two was a less-commonly-used type of seizure medication. If a doctor knows that a patient has difficult to control seizures, this information pr.ompts aggressive therapy and evaluation when a seizure occurs, close monitoring of blood levels and care in changing a proven regimen. None of these basic elements of care are present in the “care” this inmate received at Pelican Bay. In addition, had the physicians obtained his old medical records (and there is no evidence of any attempt to do so), they may have found that he was resistant to one of the other types of medications. Furthermore, negligence in follow-up and monitoring of his drug levels resulted in Tyler receiving toxic doses of seizure medicine (Trial Ex. P-488 at 17052). Physician notes referred to “probable malingering” -(Trial Ex. P-488 at 17036) in spite of overwhelming evidence to the contrary.

Start Decl. at 54-56. Tyler Henderson died in his cell on March 15, 1992 of probable cerebral anoxia due to epileptic seizures. As defendants’ expert conceded, Mr. Henderson’s case raises serious concerns about “physician involvement in the care” of the patient. Harness Tr. 19-3102. Again, review of this file would have accentuated the urgent need for organized files, adequate staffing, competent medication management, and closer supervision of MTAs, and thus helped to avoid similar problems in the future.

g. Treatment Provided

Predictably, the systemic deficiencies described above in Pelican Bay’s provision of medical care have given rise to a distinct pattern of substandard care. Plaintiffs’ expert opined that Pelican Bay’s health care system is “not merely in difficulty, or even in crisis, but ... has failed entirely in the regular provision of health care services to. inmates.” Start Deck at 5. Both parties agree that Pelican Bay must be able to provide decent primary care — that is, according to Dr. Harness, to care for" inmates with “acute and chronic illnesses that are typically cared for by all primary care physicians,” such as diabetes, hypertension, seizure disorders, asthma, TB, and the complications of HIV positivity. Harness Deck at 11, 14. However, Pelican Bay has failed to produce a health care system in which even these basic needs are consistently met. As Dr. Start noted, “in addressing the known and foreseeable health care needs of the inmates, Pelican Bay ranks among the very worst, if not the worst, of the many prisons I have evaluated.” Start Deck at 4.

In part Pelican Bay’s grossly inadequate provision of primary care stems from a lack of established protocols for dealing with chronic illnesses. There are no chronic disease clinics at Pelican Bay, and the facility has no established protocols for dealing with common illnesses such as diabetes or hypertension. 133 Even defendants’ medical expert agreed that, as a result, there have been inmates who did not receive “appropriate” care. Harness Tr. 19-3098. However, genteel generalizations can not convey the horrifying inadequacy of care that inmates at Pelican Bay have received.

Inmate David Evans died in August of 1992 of pneumocystis pneumonia. This inmate, who had a history of asthma and was a documented diabetic, was admitted to the infirmary in July of 1992 “gasping for air,” with his nostrils flaring, a temperature of 102.4 degrees, and abnormal lung sounds. Tidal Exh. P-683 at 29436, 29466. Dr. Start described the treatment Evans received in the infirmary:

The orders for the inmate written by Dr. Astorga are shockingly inadequate. Although the patient was a chronic asthmatic in distress, no asthma medicines were ordered! No chest X-ray, or sinus films were performed, and there was no assessment of oxygenation. No laboratory tests were done, nor initial assessment of glucose or adjustment of insulin dose, despite the setting of an acute infection. Only-finger sticks were done (which vary from the low range — 79—to 219[)]. Vital signs were only taken twice during stay[,] and his respiratory rate was very abnormal both times. He was given Propranolol, which is contraindicated with asthma and diabetes — a fact that should be known by any physician, and was also clearly noted in the chart by Dr. Winslow when it was discontinued on 5/24/92.

Start Decl. at 50 (citations to Trial Exh. P-683 omitted). Evans was discharged from the infirmary on August 1, 1992 without a physical exam; he was prescribed antibiotics (which it appears he never received) which were not therapeutic for pneumocystis pneumonia.

Evans’s death is particularly tragic because “survival in episodes of pneumocystis is related to how early the infection is detected and treated.” Start Decl. at 52. Yet Evans was never diagnosed at Pelican Bay, despite the fact that his symptoms were “classic” for pneumocystis. Moreover, medical staff may have been more likely to identify Evans’s pneumocystis, an AIDS-related pneumonia, had they known he was HIV positive; however, he was never tested, despite his history of weight loss, lymphadenopathy, and narcotic drug abuse. 134

During the period from August 1 to August 11, Evans refused food and insulin six times, and several MTA records document his acute shortness of breath. Trial Exh. P-451 at 29074, 29077-78. He also refused medical care during this period. If a diabetic, ill patient refuses insulin, food, or medical treatment, investigation by a physician is warranted; each of these refusals should have, prompted an evaluation by a physician. Start Decl. at 51. An MTA put Evans on a doctor’s line on August 8, and when he was finally seen by a doctor on August 11, the physician merely clarified his medication dosage. Between August 12 and August 20, there are seven more documented refusals of food or insulin and medical treatment. Evans never again saw a doctor at Pelican Bay, despite the fact that he “was clearly lying in his cell dying” in the weeks following his release from the infirmary. Start Decl. at 51. When he finally asked to see an R.N. on August 20, he was transferred to Sutter Coast Emergency Room because there were no physicians on the grounds. He died at Sutter Coast that night.

This inmate’s case illustrates Pelican Bay’s grossly inadequate treatment of diabetes and asthma, diseases that appear frequently and foreseeably in the prison population. However, as Dr. Start commented, even given the substandard diagnosis and treatment Evans received, “perhaps the most egregious failure in the case history is the ‘do nothing’ posture of the Pelican Bay physicians during the three weeks the inmáte languished in his cell before dying.” Start Decl. at 48-49.

Examples of inadequate treatment of other types of illness abound in the record. As Dr. Start documented in pages 140-316 of his declaration, scores of inmates received treatment that confirms Pelican Bay’s failure to provide adequate care, adequate recordkeep-ing, access to care, and appropriate training and supervision of the staff. 135 For example, inmate Bernard Hughes saw an MTA on September 2, 1990 after complaining of abdominal pain. The next day he again saw an MTA, who noted “rebound tenderness”— “one of the most convincing signs of appendicitis and [one which] suggests an acute emergency.” Start Decl. at 58. The MTA contacted a nurse, who simply ordered the inmate placed on the doctor’s line. Trial Exh. P-501. When, on September 8, Hughes was finally taken to Sutter Coast Hospital, surgeons there found that Hughes had a ruptured appendix that had become gangrenous.

Inmate Raul Mendoza presents another example:

This inmate had a dislocation of his right shoulder on 6/92 and was left immobilized, in a sling, without adequate hygiene until an abscess with fever developed on 7/6/92. He was seen by an MTA who noticed pus coming from armpit and was left in the cell until the next day. The nurses’ notes say the inmate had fevers 4 days before being admitted to infirmary. There is no physician history or adequate physical exam available. Of concern is the fact that the inmate had been left in a shoulder immobilizer with no long-term plan, physical therapy or follow-up. Patients who do not use their joint develop frozen shoulder syndrome and lose all function of the joint. Gradual range-of-motion and gentle physical therapy are essential in order to ensure that the treatment does not cause a worse outcome than the initial problem. None of that was done here.”

Start Decl. at 250-51. (citations to Trial Exh. P-709 omitted).

Perhaps the most graphic example of inadequate medical care is that received by inmate Vaughn Dortch. The scalding of Dortch in the infirmary tub is discussed in section II(A)(l)(a)(l), supra. Dortch had received second- and third-degree burns that eventually required skin grafts on his legs and buttocks, surgical excision of part of his scrotum, and extensive physical therapy. Start Decl. at 42. However, despite patently obvious indications that Dortch was burned, Dr. Astorga and Dr. Gard attempted to minimize or deny the full extent of his injuries, saying that Dortch merely had “dead skin,” Trial Exh. P-1219 at 29848, or “exfoliation.” Trial Exh. P-444 at 258. His transfer to the hospital was delayed over an hour, until he went into shock and his blood pressure became dangerously low because medical staff had not started fluid resuscitation. Kuroda Depo. at 69-73, Start Decl. at 47. However, in a memorandum written the day after the scalding, Dr. Gard states that medical staff immediately recognized Dortch’s burns and sent him to the hospital. 136 Trial Exh. P-444 at 352.

The Court agrees with Dr. Start’s report that “there is a rampant pattern of improper or inadequate care that nearly defies belief.” Start Decl. at 5. Not only has each discrete deficiency discussed above (inadequate recordkeeping, lack of supervision, and so forth) created unnecessary pain and suffering, but the deficiencies compound each other to render the provision of adequate care nearly impossible. 137 The many instances of grossly inadequate care are the utterly predictable result of systemic failures in Pelican Bay’s medical services.

The record amply demonstrates defendants’ unresponsiveness to the health needs of inmates at Pelican Bay. Some of defendants’ comments, actions, and policies show such disregard for inmates’ pain and suffering that they shock the conscience. For instance, Pelican Bay has an informal policy of treating inmates engaging in hunger strikes simply by weighing them once a week until they lose 20% of their body weight. See Griffin Depo. at 74-75. MTA logs contain unabashed notations such as “[c]omplains of chest pain. Hah!” Start Decl., Exh. U at 8498, revealing medical staffs often flippant attitudes toward inmates’ pain and suffering. Occurrences like these led Dr. Start to state that “[biased on my experience in eighteen years of correctional health care, I cannot think of a prison that more completely embodies a callous indifference toward inmate health needs.” Start Decl. at 117.

Sheer callousness aside, defendants’ behavior unambiguously evinces a conscious disregard for inmates’ serious medical needs. Defendants knew that the plaintiffs had serious medical needs, knew that the medical system at Pelican Bay was inadequate to serve those needs, and nevertheless failed to remedy the gross and obvious deficiencies of the system.

Defendants’ attitude toward staffing typifies their deliberate indifference to the clear dangers created by Pelican Bay’s medical system. Even though Chief Deputy Warden Peetz was authorized to hire 3.5 physicians when the facility opened, the prison began operation without a physician on staff. Incredibly, Warden Marshall stated that he believed Pelican Bay could provide adequate medical care to the inmates without a doctor on site, despite the fact that he had never worked in a prison without a full time physician. Marshall Tr. 22-3825-26.

The defendants have instituted some changes, but they have often been cosmetic at best. For instance, medical staff knew full well that because of the disorganization of the medication distribution system, inmates often did not receive prescribed medications. Minutes of the prison’s Pharmaceutical Committee record Dr. Cooper’s comments that “too many mistakes were being made, too many medications were not being taken care of routinely, [there were] too many delays, and too many 602’s [inmate grievances] [were] coming through because of inmates not getting medications.” Trial Exh. D-283 at 61283 (meeting minutes for April 1992). The Committee grappled with the problem and “solved” it by abdicating responsibility for medication renewal: Dr. Cooper testified that “[w]e’ve resolved the method of ... the patient achieving his refills. We’ve taken the responsibility from a memory of the medical technician and placed it on the ... individual patient to recognize that they’re running low on their medicines and they have to ask for a refill.” Cooper Tr. 14-2273-74. Even defendants’ expert, Dr. Harness, could not bring himself-to say that this “solution” was acceptable. 138

Other systemic deficiencies of the health care system have remained virtually untouched. For example, the 1991 Department of Corrections audit warned medical staff that the record-keeping system posed a danger to patients. Trial Exh. P-3334 at 32551-52. Doctors were constantly reminded of the problem. See, e.g., Trial Exh. P-404 at 3339 (medical record reading “I have no old chart on this patient to document his problems”). However, there have been no efforts to. reform the system.

Pelican Bay doctors also continue to endanger inmates by testing them for tuberculosis without consulting medical records. 139 Before the 1992 tuberculosis screening, Dr. Johns suggested to medical staff that they consult records before applying tuberculosis skin tests. Dr. Cooper knew it was “improper” to retest inmates who had previously tested positive. Cooper Tr. 14-2302. This danger notwithstanding, inmates were tested without medical records that might reveal previous positive test results.

The absence of quality assurance programs and peer review and the lack of supervision of doctors and support staff bespeaks a striking indifference to the quality of care provided. This indifference is illustrated by Pelican Bay’s reaction when one of its doctors was reprimanded by the state Medical Examination Board: As Dr. Start noted, “a letter from the licensing board identifying a deficiency should create an explosion of corrective action. It appears from the record that no one paid any attention to this letter.” 140 Start Decl. at 77.

We find that defendants had abundant knowledge of the inadequacies of medical care at Pelican Bay. That knowledge is reflected in records of complaints by prisoners and staff, audit reports, and budget requests that allude to the risk of harm (and of litigation) if conditions are not ameliorated. We find that by failing to remedy deficiencies in health care, Pelican Bay medical staff did not merely create a risk of harm to inmates but practically insured that inmates would endure unnecessary pain, suffering, debilitating disease, and even death. We agree with Dr. Start’s opinion that “[t]he fact that a new prison with contemporary medical facilities nevertheless could be so shockingly deficient in its provision of health care is ... a terrible indictment of the defendants, and compellingly illustrates what ... is their stunning indifference to the health care needs of the prisoners at Pelican Bay.” Start Decl. at 4.

C. MENTAL HEALTH CARE

Plaintiffs contend that when Pelican Bay opened — with no psychiatrist on staff — the system for delivering mental health care was grossly inadequate. Although staffing has since improved, plaintiffs argue that continued understaffing, along with other chronic problems, continues to render the delivery of mental health care constitutionally inadequate. They also argue that defendants have been deliberately indifferent to the mental health needs of the Pelican Bay prison population.

At trial, plaintiffs relied upon two expert witnesses who gave testimony relating to mental health care, as well as conditions in the SHU. The first, Dr. Stuart Grassian, spent two weeks at Pelican Bay, one in September 1991 and one in May 1993. During this time, he toured the prison, spoke informally with prison personnel, and conducted 69 interviews with 55 inmates (14 were interviewed twice). He also reviewed the medical files of most of these inmates, depositions of Pelican Bay health professionals, and other documents. At trial, Dr. Grassian estimated that he had reviewed “18 U~Haul boxes [of documents] at last count.” Tr. 12-1862. The second expert, Dr. Craig Haney, visited Pelican Bay on September 16, 1992 and January 6, 1993, at which time he toured the prison and spoke informally with prison personnel. He also separately conducted formal interviews with 65 inmates, reviewed depositions of Pelican Bay mental health professionals, and examined an extensive number of documents and files.

Defendants’ expert, Dr. Joel Dvoskin, visited Pelican Bay for one day in April 1992, and then again for one day in January 1993. On both occasions he toured the facilities and spoke informally with inmates and staff. He also comprehensively reviewed approximately eleven inmate medical files, along with selected parts of other medical files. He also reviewed selected CDC training materials. 141

In addition to the expert witnesses, the Court heard from various mental health professionals presently or formerly employed at Pelican Bay or by the CDC. The parties also submitted into evidence deposition excerpts and extensive documentary evidence. Taken together, this testimonial and documentary evidence amply demonstrates that the mental health care system at Pelican Bay falls dramatically short of minimum constitutional standards.

Plaintiffs’ expert, Dr. Grassian, described the mental health services at Pelican Bay as “grossly inadequate.” Decl. at 5; id. at 158, 166-67 (system of psychiatric care is “manifestly deficient” and fails to meet “the most minimal standards for adequate psychiatric care”). 142 Even defendants’ expert, Dr. Dvoskin, could find no cause to endorse the mental health system at Pelican Bay. He testified that, as of the time he visited Pelican Bay, he could not represent to the Court that the mental health care delivery system was “adequate” or met constitutional standards. Tr. at 27-4431-32. Indeed, as detailed below, the evidence plainly shows that there have been, and continue to be, chronic and pervasive problems with the delivery of mental health care at Pelican Bay. It also reveals defendants’ deliberate indifference to the serious mental health needs of inmates.

1. The Need for Mental Health Services at Pelican Bay

A significant number of inmates at Pelican Bay, in both the SHU and the general population section of the prison, suffer from serious mental health problems. A survey done by Dr. Nadim Khoury’s 143 office in August 1990 estimated that there were at least 208 inmates at Pelican Bay who were either psychotic or psychotic in partial remission. 144 Trial Exh. P-3820. 145 Dr. Sheff estimated that during his tenure at Pelican Bay from April 1992 to February 1993, there were between 200 and 300 mentally ill inmates requiring psychiatric intervention at any given time, and that between five and ten percent of incoming inmates were mentally ill or had some need to continue medication. Tr. 25-4103-4. See also Grassian Decl. at 4-5 (expressing his opinion that he had met numerous acutely psychotic inmates during his interviews in the Pelican Bay SHU); Haney Decl. at 55 n. 20 (finding that out of 40 random interviews in the SHU, close to one-third suffered from what appeared to be psychotic symptoms or had been placed on anti-psychotie medication) 146 ; Ruggles Tr. 17-2905 (senior staff psychologist stating that there are a “considerable” number of inmates at Pelican Bay with organic brain damage).

As Warden Marshall described in an August 1991 budget request, the high incidence of mentally ill inmates at Pelican Bay is predictable because mentally ill inmates frequently exhibit behavioral problems, and inmates with a history of misconduct are often transferred to Pelican Bay:

[A] large number of psychiatrically disabled inmates exhibit violent and problematic behavior as part of their symptoma-tology. Consequently, these inmates present a history of severe disciplinary and assaultive behavior that is considered an endangerment to themselves, other inmates, staff, and prison security_ Pelican Bay was constructed to house those inmates considered to be the most violent and problematic_ By virtue of its mission, Pelican Bay now houses most of the psychiatrically disabled inmates who have a history of violent and assaultive behavior.

Trial Exh. P-4602 at 49197-98; id. at 49198 (“Current departmental security needs dictate that institutions transfer problematic psychiatrically-disabled inmates to Pelican Bay as soon as medically possible, since they cannot be accommodated anywhere else”). See also Dvoskin Tr. at 27-4443-44 (observing that there seemed to be more inmates with mental illness in higher security prisons because of classification, disciplinary, and service delivery practices); Haney Decl. at 68 (inmates unable to manage their psychiatric disorders often incur rules violations).

This same budget request also noted that “[b]ecause of inadequate outpatient services at Pelican Bay, these inmates usually decom-pensate quickly and require intensive psychiatric care and/or readmission to inpatient care.” Trial Exh. P-4602 at 49198. Warden Marshall further observed that a “cursory review” showed that there were 214 inmates on psychotropic medication alone, and that “according to referrals by medical doctors and other staff, there are more than 100 other inmates in immediate need who should be seen and receive ongoing psychiatric treatment.” Id. 147

The need for substantial mental health services at Pelican Bay is heightened by the presence of the SHU, which houses approximately 1,500 persons. As detailed more fully in section 11(D)(2), infra, the conditions in the SHU are sufficiently severe that they lead to serious psychiatric consequences for some inmates. As Dr. Grassian concluded, “[f]or some, SHU confinement has severely exacerbated a previously existing mental condition,” while other inmates developed mental illness symptoms not apparent before confinement in the SHU. Grassian Decl. at 4. Defendants’ expert also acknowledged that there are some people who “can’t handle” segregation in the SHU. “Typically, those are people who have a pre-existing disorder that is called borderline personality disorder, and there — there’s a fair amount of consistent observation that those folks, when they’re locked up [in segregation] may have a tendency to experience some transient psychoses, which means just a brief psychosis that quickly resolves itself when they’re removed from the lockdown [segregation] situation.” Dvoskin Tr. 27-4374-75 (emphasis added). Inmates with chronic longstanding depression, chronic schizophrenia, or any other longstanding, severe mental illness are also at a higher risk of deteriorating in the SHU. Dvoskin Tr. 27-4473-74. Pelican Bay senior staff psychologist Ted Ruggles also observed a connection between placement in the SHU and the mental health of certain inmates: “There was a psychiatric deterioration that occurred in correlation with placement. on SHU [with some inmates], and I’m not altogether certain what caused it.” ■ Rug-gles Tr. 17-2914. A memorandum prepared in September 1989 by the Institutions Division of the CDC also underscored the substantial need for psychiatric services in the SHU, particularly where the prison fails to screen out inmates who may be vulnerable to developing .serious mental disorders. Trial Exh. P-3390.

Given the above, it was manifest that operation of the SHU would require close psychiatric monitoring and substantial psychiatric services. Haney Decl. at 67 (the need for psychiatric screening and monitoring in the SHU can not be overemphasized); see also Dvoskin Decl. at 11 (while “some inmates with mental illness can be adequately treated in the [SHU], if the necessary services are available ... [, t]he most basic need is for observation [by a mental health practitioner] to insure that their mental illness is not being exacerbated by the tighter confinement and more restrictive socialization”) (emphasis added). Significantly, Dr. Grassian found numerous acutely psychotic inmates in the SHU in need of immediate, hospital-level, inpatient treatment. Grassian Decl. at 45-104.

2. Systemic Deficiencies in the Delivery of Mental Health Care

When Pelican Bay began operations in December 1989, it was severely understaffed and ill-equipped to respond to the mental health needs of the inmate population, which quickly grew from 1,287 inmates on January 1, 1990 to over 3,500 inmates by January 1, 1991. From December 1989 until April 1992, the total mental health staff employed at Pelican Bay consisted of either one or two psychologists. The result was nothing short of a mental health care crisis. One of the initial staff psychologists described his day as “running as fast as I can putting out as many fires as I could.” Rose Coleman Depo. at 31 148 . When asked “what was a fire,” Rose responded as follows: ‘‘Well, I mean, these were very serious cases that demanded attention and my time was taken up taking care of crisis intervention ... [which involved] acutely psychotic, acting-out inmates.” Id. at 32. Even with these frantic efforts by Rose, acutely psychotic inmates failed to receive treatment for an “unacceptable period of time” on “a regular basis.” Id. at 34. “[I]f a patient did not [engage in] very flagrant behavior, aggressive violent behavior or suicidal behavior, they could stay in that cell for a long period of time, just nobody pays that much attention to them.... [M]onths maybe.” Id. at 35.

Until April 1992 — almost 2 and ]é years after the prison opened — there was no resident psychiatrist at Pelican Bay with the exception of a psychiatrist who submitted his resignation after working for one month. Trial Exh. P-3121. 149 Instead, defendants attempted to obtain visiting psychiatrists from other institutions for two to five days each month. Id. However, even this sparse coverage was not always obtained. When Dr. Baker, one of the visiting psychiatrists, was asked whether he came to Pelican Bay one week a month he responded: “It was very irregular based upon the need and my ability to extricate myself from [California Medical Facility].” Baker Coleman Depo. at 20. He believes that for at least a year, he was the only visiting psychiatrist. Id. at 21. As senior staff psychologist Ted Ruggles recalled, psychiatrists from other institutions visited only “periodically.” Tr. 17-2904. In November 1990, Warden Marshall sent the Deputy Director of Institutions a memorandum stating that there were currently five federal court eases regarding the lack of psychiatric care at Pelican Bay, and that the system of “borrow[ing], on occasion, psychiatrists from existing institutions” was “generally a hit and miss method.” Trial Exh. P-4737. Finally, in January 1992, psychiatrists from different institutions were assigned to Pelican Bay each week; however, this was still a wholly inadequate, stop-gap measure. Even in July of 1992, with two full-time psychiatrists at Pelican Bay, many mentally ill inmates went without adequate psychiatric treatment.

By the time of trial, and under the pressure of litigation in different courts, the mental health staff at Pelican Bay had slowly climbed to nine clinicians: a chief psychiatrist and two staff psychiatrists, a senior psychologist and three staff psychologists, and two licensed social workers. There are also two MTAs, one office technician, and a medical transcriber.

As experts for both sides agree, however, this level of staffing remains insufficient to provide adequate mental health services for the population at Pelican Bay. Grassian Decl. at 166; Dvoskin Tr. 27-4411 (current staffing “probably not” adequate). At his deposition in January 1993, staff psychologist Dr. Ruggles testified that the provision of services is still primarily crisis-oriented, with emphasis on crisis intervention stabilization in cases where inmates are exhibiting disrup tive, bizarre or aberrant behavior, making suicidal statements or gestures, or experiencing a personal family crisis. Ruggles Coleman Depo. at 36-38. 150 Treatment for seriously ill inmates is primarily limited to medication management through use of antipsy-chotic or psychotropic drugs, 151 and intensive outpatient treatment is not available. The lack of staffing is particularly problematic in the SHU. As Dr. Dvoskin testified, current staffing levels are not sufficient to enable the mental health staff to quickly and effectively respond when inmates exhibit serious mental health problems in the SHU. Tr. 27-4475-76.

Defendants recently approved an additional 6.9 mental health positions for Pelican Bay, which would include a psychiatrist, a psychologist, a registered nurse, a senior MTA, and an additional health records technician. However, at the time of trial, none of these positions had been filled. Moreover, even defendants’ expert would not confirm that this additional staffing, which would increase the mental health staff to 16, would provide an adequate level of care, stating that “I can’t give [a] definitive answer without actually seeing [the staff] in place.... [It] might well be enough or it might not.” Dvoskin Tr. 27-4411-12; id. at 27-4476 (noting that if the 6.9 additional positions are filled, “they’re getting closer, but I would still have questions about it”). 152

Needless to say, the lack of adequate staffing severely impacts the level of care received by inmates. As Dr. Grassian testified, “staffing shortages [at Pelican Bay] have led inexorably to inadequate access to care, inappropriate and shoddy medication management and monitoring, and chaotic record-keeping.” Grassian Decl. at 166. Dr. Gras-sian further concluded that “[t]hese failures, taken together, violate even the most minimal standards for adequate psychiatric care.” Id. at 167.

b. Screening and Referrals

It is important that a mental health care system effectively identify those inmates in need of mental health services, both upon their arrival at the prison and during their incarceration. While mentally competent inmates can be relied upon to self-report most medical ailments, mentally ill prisoners may not seek out help where the nature of their mental illness makes them unable to recognize their illness or ask for assistance. Nor are family or friends usually around to notice developing mental problems or help inmates seek treatment.

For almost three years, Pelican Bay did not have an organized screening system at all. As Dr. Baker described, “[u]ntil the advent of our own MTA system, which we have instituted within the last three months [as of January 1993], there really was not an organized way of picking up on problems and feeding them to us. It happened. We had no control over it. It was a passive rather dependent situation which we really didn’t have anyone out there who was a member of our staff to pick up on problems and follow them through.” Baker Coleman Depo. at 26. Thus, many mentally ill inmates did not receive any mental health care until they were grossly psychotic and/or exhibited flagrant or suicidal behavior. Those who had serious mental illnesses but exhibited less unusual outward behavior might suffer in their cells for weeks or months without detection. See also Trial Exh. P-3161 at 83788 (1991 CDC audit finding that Pelican Bay was “seriously lacking” in the “timely identification” of inmates with psychiatric concerns).

The current system, while a significant improvement, still does not provide for “adequate[ ]” early intervention. Dvoskin Tr. 27-4456. The MTAs who briefly screen incoming inmates typically do not have the necessary training and background to recognize psychiatric illnesses. Staffing shortages also create gaps in the screening process, which are further exacerbated when staff are absent because of illness or vacation. As Dr. Dvoskin observed, “it would ... certainly not be unlikely that [mentally ill] people would be missed upon transfer [from one prison to another].” Tr. 27-4458.

For those inmates already confined at Pelican Bay, the prison relies on referrals from custody staff or the inmate. Mental health staff who participate in classification committee reviews can also initiate referrals, and MTAs have contact with inmates taking psychotropic medication. Staff psychiatrists and psychologists, however, rarely visit the cell-blocks in the SHU.

While custody staff can often provide useful information regarding an inmate’s conduct and are instructed to report unusual behavior, 153 they are not adequately trained to identify mental illness. Nor do all custody staff consistently refer abnormal behavior. Defendants’ expert also observed that the lack of mental health staff leads custody staff to impose a higher referral threshold than appropriate. Dvoskin Tr. 27-4462. As a consequence, custody staff essentially make medical judgments that should be reserved for clinicians, and some inmates are not given appropriate early treatment that could prevent or alleviate a severe psychiatric disorder.

As defendants’ expert noted, the need for effective screening and monitoring in the SHU is particularly critical in order to ensure that inmates suffering from mental illness are not experiencing a deterioration in their condition. Dvoskin Decl. at 11; Tr. at 27-4475 (emphasizing the importance of responding quickly if symptoms begin to emerge). The same holds true with respect to inmates who do not have a demonstrated history of psychiatric illness.

In the New York system administered by Dr. Dvoskin, mental health staff make regular rounds (10 hours per week) in each segregation unit (with between 30-110 inmates) “to identify problems before they become anything remotely like mental illness” and to reduce stress among inmates and staff. Dvoskin Tr. 27-4419-4421; 4466-4468. When Dr. Dvoskin recommended such rounds to the Warden or the Chief Deputy Warden at Pelican Bay, he was told that lack of staff precluded such a program in the SHU. Tr. 27-4468; Dvoskin Decl. at 11. In addition, there is no policy requiring any periodic psychological evaluations of SHU inmates. Peetz. Tr. 20-3330-31.

c. Psychiatric Records

The ability to provide appropriate psychiatric treatment at Pelican Bay is also impeded by the poor condition of inmate psychiatric records. First, the psychiatric records that are forwarded from other institutions are often sketchy, and important information, including prior psychiatric hospitalizations, is sometimes missing. Second, once the records arrive at Pelican Bay, they are poorly maintained. Notes of mental health examinations are often cursory, and documentation of monitoring is “very chaotic and haphazard in many of the cases” reviewed by Dr. Gras-sian. Grassian Tr. 12-1903. Entries sometimes fail to account for prior diagnoses; mental health staff “just put[] in another diagnosis with no comment on the fact that there’s a discrepancy here so that, you know, you see a person five times, he’s got five diagnoses.” Id. at 1904. Also, suicide watch records are made in the infirmary record rather than in the medical record, and psychiatric services staff do not receive these records from the infirmary.

d. Delays in Transfers for Impatient and Outpatient Care

Pelican Bay does not offer psychiatric inpatient or intensive psychiatric outpatient treatment for mentally disturbed inmates. Inmates ill enough to require inpatient care 154 must be transferred to another institution such as the California Medical Facility (“CMF”) in Vacaville or Atascadero State Hospital. Inmates needing intensive outpatient care must be transferred to CMF or the California Men’s Colony (“CMC”). At Pelican Bay, several inmates each month may be referred out for either inpatient or intensive outpatient evaluation and treatment.

Traditionally, there have been exceedingly long delays in the transfer of inmates needing inpatient treatment. In June 1991, defendants’ audit found that “[m]ajor problems exist in the transfer of medical and psychiatric patients from the Pelican Bay State Prison to the California Medical Facility — Vaca-ville.” Trial Exh. P-3161 at 83807. Inmates needing either inpatient or outpatient care could wait up to three months before they were transferred from Pelican Bay, during which time they failed to receive appropriate psychiatric care. According to former staff psychologist Rollin Rose, at one point, the mental health staff was so seldom successful in getting inmates into CMF that they just “gave up [trying] after a while” except in “very extreme” cases. Rose Coleman Depo. at 57-58.

The transfer process has considerably improved for inmates needing inpatient care, a fact that former staff psychiatrist Bruce Baker attributed to the initiation of “these legal actions” and the arrival of Chief Psychiatrist Albert Sheff. Baker Coleman Depo. at 41-42. Inmates needing “immediate” inpatient care are now generally transferred to CMF in three days, although this is sometimes stretched to five or six days. Referrals to CMF for intensive outpatient treatment still take at least a month and sometimes two or three months. While inmates awaiting transfer may continue to be seen periodically by clinicians at Pelican Bay, in most cases they are not receiving the services necessary to provide them with appropriate treatment.

It is not uncommon for inmates transferred to another site for inpatient or intensive outpatient treatment to later be returned to Pelican Bay in essentially the same condition. As staff psychologist Rose described, “[w]ell, of course, they were sent to, referred to CMF in the first place because it was felt that they had serious psychiatric problems that we could not treat at Pelican Bay. And then several months later, we would get them back more or less in the same condition.” Rose Coleman Depo. at 54. Other inmates return from CMF or CMC in an improved condition but then regress. According to Dr. Sheff, about half of the inmates transferred back to Pelican Bay from CMF “[do] not do well.” Tr. 25-4177.

e. Lack of Procedures for Necessary Involuntary Psychiatric Treatment

There are occasions where, in the medical judgment of a psychiatrist, a seriously ill patient is clearly in need of anti-psychotic medication, but is too paranoid and frightened to cooperate in his or her own treatment and thus refuses medication. The Supreme Court recently ruled that a state may, consistent with constitutional guarantees, involuntarily treat an inmate with antipsychotic drugs if the inmate is dangerous to himself or others and the treatment is in the inmate’s medical interest as determined by medical professionals. Washington v. Harper, 494 U.S. 210, 223-29, 110 S.Ct. 1028, 1038-40, 108 L.Ed.2d 178 (1990). The inmate, however, is entitled to a hearing before medical professionals to ensure that the decision to medicate the inmate against his will is neither arbitrary nor erroneous. Id. at 227-29, 110 S.Ct. at 1040; cf. Keyhea v. Rushen, 178 Cal.App.3d 526, 223 Cal.Rptr. 746, 755-56 & n. 3 (1986) (prisoners are entitled to judicial determination of their competency to refuse treatment before they can be subjected to long-term [over 10 days] involuntary psychotropic medication).

At Pelican Bay, there are no protocols or procedures in place for administering involuntary psychiatric medication. Instead, inmates needing involuntary medication must be transferred to CMF for inpatient treatment. However, as noted, this process usually takes three days, and sometimes longer, during which time the inmate is not involuntarily medicated. Thus, inmates in acute distress often suffer for an extended period of time before they receive treatment that should be provided immediately.

There are also inmates who need and would benefit from involuntary medication, but who are not transferred to a facility offering such treatment on account of security concerns. For example, Inmate A 155 is an inmate who was suffering from delusional beliefs and auditory command hallucinations telling him to commit violent acts. When Dr. Grassian interviewed the inmate in September 1992, he found him to be “severely mentally ill, incompetent to appreciate his need for treatment, and a danger to himself and others.” Grassian Deel. at 51; Tr. 12-1900-02. Dr. Grassian was informed, however, that the inmate’s security needs required him to remain at Pelican Bay. Inmates like Inmate A are essentially trapped in a Catch-22: they are too psychotic to consent to treatment, yet their psychosis makes them too “dangerous” for a transfer to a facility where they could receive treatment that would potentially reduce their security risk. See also Astorga Depo. at 753-54 (acknowledging “Catch-22” situation of inmates whose custodial needs preclude transfer).

f. Failure to Involve Mental Health Staff in Housing Decisions

There are instances where it may be critical, from a medical standpoint, to alter an inmate’s housing assignment (e.g., from the SHU to another environment or from double to single cell housing), in order to effectively address an inmate’s serious mental health problems. With respect to the SHU, Dr. Grassian concluded that some inmates in the SHU have experienced a severe exacerbation of existing mental conditions or the onset of mental illness, and that “many of the acute symptoms suffered by these inmates are likely to subside upon termination of SHU confinement.” Grassian Deck at 4. Dr. Dvoskin also testified that, except for “very, very rare exceptions,” inmates who are in acute psychiatric distress or suicidal depressions should not be housed in the SHU. Tr. 27-4473.

Nonetheless, Pelican Bay psychiatrists and psychologists are not, as a practical matter, allowed input into cell housing decisions, even when the inmate is suffering acute symptoms and the mental health staff believe that a change in housing conditions is potentially necessary to the effective treatment of the inmate’s disorder. Defendants’ complete failure to consider the mental health needs of inmates in making housing decisions seriously compromises the ability of the mental health clinicians to effectively and adequately treat their patients. 156

g. Suicide Prevention

While prison staff receive a modicum of suicide prevention training, there is no comprehensive suicide prevention program in place. As part of their basic training, new correctional officers take a three-hour course entitled “Unusual Inmate Behavior,” which includes a short section on how to identify inmates susceptible to suicide and what to do after identifying such an inmate or discovering an attempted suicide. Trial Exh. D-327. In June 1992, a “Suicide Prevention Handbook” was distributed to all Pelican Bay staff, and they were required to read the handbook and complete an accompanying quiz. Trial Exh. D-297. There has also been some additional in-service training; however, it appears to have taken place on a sporadic basis.

h. Quality Assurance

As described in section 11(B)(2)(f), supra, a Quality Assurance program is designed to enable a medical institution or department to review, on an ongoing basis, staff medical decisions and practices in order to assess whether corrective measures are necessary or appropriate. Such a program is considered “standard practice” in virtually every health care facility in the country and is considered a “fundamental part” of a health care operation. Harness Tr. 19-3117.

At least as of trial, however, Pelican Bay, after almost four years of operation, still had not implemented a Quality Assurance • program for its mental health staff. Former staff psychologist Rollin Rose explained that he never tried to discuss this particular point with Dr. Astorga, Pelican Bay’s Chief Medical Officer, because “there are certain issues that just wouldn’t be very fruitful to discuss with Dr. Astorga and that was probably one of them.” Rose Coleman Depo. at 78-79.

i.Treatment Provided

For those inmates that are seen by the mental health staff, 157 the combined effect of staffing shortages and other problems discussed above has inevitably led to numerous instances of grossly deficient treatment. As Warden Marshall stated in a 1991 budget request, “Pelican Bay’s design staffing is not adequate to even identify all of the psychiat-rically disabled inmates, much less provide the mandated mental health treatment.” Trial Exh. P-4602 at 49198. A 1991 GDC audit also concluded that Pelican Bay was “seriously lacking in the ... treatment and tracking of inmates diagnosed as having psychiatric concerns_ The deficiencies exist[ ] particularly in the areas of appropriate follow-up treatment and/or transfer to more suitable housing.” Moreover, serious problems continue to persist, notwithstanding the modest staffing increases. See Dvoskin Tr. at 27-4456 (current system does not provide adequate stabilization and symptom management); Grassian Tr. at 12-1904 (“the followup of psychiatric staff is ... often extremely chaotic [and] sporadic”).

The pain, suffering, and deterioration experienced by inmates who fail to receive appropriate treatment for their mental disorders is substantial. In the case of an inmate suffering a serious psychotic break, the impact can be enormous. As Dr. Grassian described, “I’ve had patients who’ve lived through psychotic breaks of that magnitude [observed at Pelican Bay]. And it is a scarring experience for years, probably for the rest of their lives, to feel that out of control and that agitated and that terrified, to know how absolutely terrified you can be, to know how absolutely out of control you can be. It is a very scarring, frightening experience that people live with, and there are prisoners suffering ... it day after day.” Tr. 12-1973-74.

While the deficiencies in defendants’ system of mental health care are felt prison-wide, the problems are especially severe in the SHU. At least three factors have contributed to this result. The first stems from the very mission of the SHU, which is to house the most dangerous and disruptive inmates. Since, as discussed above, inmates suffering from mental illness are more likely to engage in disruptive conduct, significant numbers of mentally ill inmates within the California prison system are ultimately transferred to the Pelican Bay SHU. Second, the severity of the environment and restrictions in the SHU often cause mentally ill inmates to seriously deteriorate; other inmates who are otherwise able to psychologically cope with normal prison routines may also begin decompensating in the SHU. Third, defendants chose to provide only limited psychiatric services to inmates in the SHU. Aside from the obvious limitations which ensue from lack of staffing, discussed above, defendants have made no effort to provide appropriate treatment for inmates suffering from major mental disorders. The prison is not equipped to provide any inpatient or intensive outpatient treatment or involuntary medication. At the same time, delays prevent urgently needed care from being provided off-site as well. While inpatient care can be provided elsewhere in 3 to 5 days, this does not help the inmate who needs immediate hospitalization or involuntary medication. Similarly, a one to three month delay effectively denies adequate treatment for the seriously ill inmate needing immediate intensive outpatient treatment. And in some cases, security concerns preclude any transfer at all.

In short, defendants created a prison which, given its mission, size, and nature, would necessarily and inevitably result in an extensive demand for mental health services — perhaps more so than any other California facility; yet, at the same time, they scarcely bothered to furnish mental health services at all, and then only at a level more appropriate to a facility much smaller in size and modest in mission.

It is not surprising, then, thát during his 69 interviews with 50 SHU inmates in September of 1992 and May of 1993, Dr. Gras-sian found that 17 of those inmates were acutely psychotic and not receiving appropriate treatment for their condition. As detailed at length in Dr. Grassian’s declaration at pages 25 and 45 through 104, medical staff often failed to conduct adequate mental examinations, ignored past medical history or failed to obtain a proper history, made contradictory or inconsistent diagnoses, and engaged in only a superficial review and diagnosis. A number of inmates were in need of immediate involuntary medication and/or hospital-level inpatient treatment; some were .the most severely ill people Dr. Gras-sian has encountered in his research and observations.

The following examples are illustrative of some of the glaring problems found in the delivery of mental health care at Pelican Bay:'

Inmate B

Inmate B is a 38 year-old white male with a history of childhood sexual abuse, intermittent paranoia, periods of depression, and pri- or psychiatric hospitalization. Grassian Deck at 66. When Dr. Grassian interviewed him in May 1993, he was “quite obviously paranoid and psychotic.” Id. Among other things, he was suffering from auditory and visual hallucinations and delusional ideas. Inmate B reported to Dr. Grassian that:

“the [custody staff] are growing marijuana. They killed someone. They cut each other off in conversation day by day.... I’m on Stelazine (an antipsychotic) because of audio and visual hallucinations. I hear radio transmissions, background hiss. If I have difficulty sleeping it gets worse. I see patterns on the wall, shadows get brighter. The hallucinations get more intense with less sleep.”

Id. at 66-67. Inmate B also believed that his body had been transported by “astral projection” to a place where it was invaded and mutilated. Id. at 67. Although the severity of his illness warranted transfer to a psychiatric hospital, he was simply receiving medication adjustments on “roughly a monthly basis.” Id. at 66-67.

Inmate C

In September 1992, Dr. Grassian found Inmate C to be in an acute catatonic state requiring immediate hospitalization and anti-psychotic medication. He was in a fixed, immobile posture, staring “bug-eyed” at the walls and ceiling, with his posture punctuated by “sudden jerking movements of his eyes and body, giving the clear impression that he was responding defensively to frightening internal (hallucinatory) stimuli.” Grassian Decl. at 46. This type of catatonic posturing is “usually associated with an inner state of absolute abject terror.” Grassian Tr. 12-1908. Over the years, Dr. Grassian has seen many patients in a similar state, especially in his experience with psychiatric inpatients, and has learned to “regard it as a psychiatric emergency of the first magnitude, a living nightmare which even after the acute episode is successfully treated, produces deep lasting emotional scars.” Grassian Decl. at 46-47. Such a patient should be immediately hospitalized and treated with antipsychotic medication under very close supervision.

Inmate C had been on antipsychotic medications continuously since 1991, and had previously been diagnosed as suffering from various mental disorders by several mental health professionals. Dr. Grassian also noted that accurately feigning a state of acute catatonia is something that few, if any, can achieve. Grassian Decl. at 49. Nonetheless, Pelican Bay staff suspected Inmate C of malingering. Grassian Decl. at 47 — 19. When Dr. Grassian returned to Pelican Bay in May 1993, Inmate C was still psychotic and hallucinatory.

Dr. Grassian found that Inmate C’s situation is one that particularly shocks the conscience. “[Tjhere has been no consistency regarding the clinicians who saw him, nor was there adequate supportive psychothera-peutic contact: he appears to have been seen only a handful of times during the entire period. Finally, there was no consistency from visit to visit as to diagnosis. He was at various times diagnosed as suffering from schizophrenia ..., organic hallucinosis, a personality disorder, an organic mental disorder, or to be malingering. There is no continuity in these assessments; it is as though each interview was a unique event unrelated to prior contacts.” Grassian Decl. at 49-50.

Inmate D

Inmate D was housed at Corcoran State Prison prior to his transfer to the Pelican Bay SHU in April 1990. While at Corcoran, he was treated with the antipsychotic medication Mellaril and the mood-stabilizing medication Lithium. In a psychiatric summary dated March 28,1990, just three weeks prior to Inmate D’s transfer to Pelican Bay, it was specifically noted that Mellaril helped stop the prisoner’s auditory hallucinations and his inability to concentrate, and that Lithium helped him control his temper. Trial Exh. P-643 at 69169.

Once at Pelican Bay, however, Inmate D was not provided any psychiatric treatment until September 1990, five months after his transfer. When the inmate complained that he had trouble controlling his anger and aggressive behavior, the staff psychologist noted that the inmate’s file was unavailable for review to check his psychiatric history and psychotropic medications. As a result of the inmate’s complaint, a psychiatrist who apparently never saw Inmate D nor reviewed his records prescribed Lithium and an antidepressant, Elavil. He was not prescribed Mellaril, the other major medication he had received at Corcoran. Five months later, in January 1991, he simply stopped receiving medication, for no reason apparent in the record. Inmate D was not seen again until March 1993, over two years later. At that time, a social worker interviewed him briefly and referred him to a psychiatrist, who wrote in his record: “[unintelligible] and wants legal relief from all the harassment from inmates. No mental disorder. Not asking for treatment. Plan — no treatment — no followup.” Trial Exh. P-643 at 69167. There is no indication that the psychiatrist ever considered Inmate D’s psychiatric history. By May of 1993, Inmate D was in a grossly deteriorated mental state. Just days before he was interviewed by Dr. Grassian, he was placed on Lithium, but not Mellaril or any other antipsychotic medication. Based on his May 1993 interview, Dr. Grassian concluded that Inmate D was seriously ill and suffering from obsessional thoughts of violence, impul sive violence, paranoia, severe anxiety, massive difficulty with concentration, recurrent dissociative eonfusional episodes, and perceptual distortions. Grassian Decl. at 43, 128-32.

Inmate E

Inmate E was previously incarcerated by the California Youth Authority (“CYA”) beginning in 1987. There he underwent extensive psychiatric evaluation and testing for his pedophilia. Although he was described by CYA psychiatrists as “moody, depressed, narcissistic, and very immature,” Trial Exh. P-173 at 24680, and diagnosed as having pedophilie impulses that were out of control, id. at 24681, clinicians found no evidence that he had a psychotic disorder.

However, Inmate E became overtly psychotic and suicidal after being placed in the Pelican Bay SHU in 1991. He was evaluated in April 1992 after he wrote a suicide note in his own blood. Inmate E reported that he was “hearing voices” and the examining doctor described him as “obviously very psychotic.” Trial Exh. P-694 at 3675. Although Inmate E claimed the next day that some of his behavior was “a fake,” he almost immediately thereafter reported “hearing voices,” then “flipped out,” according to an MTA. Id. The inmate was prescribed Mellaril and remained in the infirmary for two weeks. However, he was discharged to custody by Dr. Sheff in early May with the notation, “no meds, no psych problems noted.” Id. at 3676.

Inmate E continued to have psychotic or suicidal episodes; Pelican Bay staff seemingly vacillated between treating his psychotic episodes as such and dismissing them as manipulation. In late May the inmate again stated that he wanted to kill himself, and then later retracted; Dr. Mandel felt he had no psychiatric disorders and recommended that the inmate be disciplined for manipulation of the system. Trial Exh, P-694 at 3662. In July of 1992, Inmate E was found to have multiple superficial lacerations on his forearm and was “talking nonsensically.” Id. at 3648-49. Dr. Baker noted that the inmate was having panic attacks and that voices were telling him to hurt himself. By August he had deteriorated further, and Dr. Baker characterized him as having a “schizophrenic” episode with “disjointed” thinking after he described hearing voices and receiving messages from a computer at the base of his neck. Trial Exh. P-497 at 993. Nevertheless, only five days after Dr. Baker’s evaluation, another staff psychologist diagnosed Inmate E as having no significant mental disorder. Id. at 990.

When Dr. Grassian subsequently interviewed Inmate E, the inmate was still grossly psychotic and incoherent. He told Dr. Grassian,

“I see or hear things. . I have been hypnotized since April 13th by Cybernetics. I can’t even explain it without being hooked up with polygraph tests. It’s like frequency tests. A bunch of people come up to me and talk about why I have to kill myself. Things I’ve thought of, things I’ve seen, animals and stuff like that. It’s frightening. They tried to kill me. They used sounds, send emotions through my body and my body shakes ... I’m tired of people talking in my head. I was mentally clear before ... sometimes I get so confused, I don’t even know what’s going on.”

Grassian Decl. 54. The lack of coherent approach to this clearly psychotic inmate is not atypical.

Undoubtedly, there are some inmates who attempt to feign mental illness, and who are justifiably considered “malingerers.” The Court also acknowledges that the identification of simulated symptoms may sometimes involve difficult judgment calls. It is clear, however, that an overburdened, and sometimes indifferent, mental health staff is far too quick to dismiss an inmate as a “malingerer” and thus deny him needed treatment, a fact that is illustrated by some of the above examples. Indeed, Dr. Grassian found “an almost obsessive preoccupation by staff members with the possibility that an inmate might be manipulating, which significantly impairs their capacity to recognize severe mental illness.” Grassian Decl. at 56; Tr. 12-1979-80. There is also evidence that inmates are labeled malingerers even though the inmate has been prescribed strong anti-psychotic medications, which should not be taken unless medically necessary, given the potentially dangerous side effects. 158

The record in this case reveals a deliberate, and often shocking, disregard for the serious mental health needs of inmates at Pelican Bay.

It is. certainly “known” that there are inmates with serious mental disorders “throughout” the California prison population. MeKinsey Tr. 26-4826. Indeed, the evidence before the Court demonstrates that it would be patently obvious to any experienced prison administrator that operating a maximum security facility with a population of 3,500 inmates, which includes, as everyone concedes, the “worst of the worst,” would create a need for substantial mental health services. Defendants also knew, given Dr. Khoury’s September 1989 memorandum, that the need for substantial mental health services would be particularly acute given the presence of the SHU, which soon housed approximately 1,500 inmates. It would also be equally obvious that the failure to provide such services would cause considerable pain and suffering. Indeed, these facts are so obvious that we find that defendants clearly knew of them.

At the same time, defendants were made aware that there would be “minimal psychiatric services available to ... SHU inmates unless they ... [became] actively psychotic and thus ... [were] eligible for transfer to another prison.” See Trial Exh. P-4220 at 5; Park Tr. 11-1681. 159 They were also aware that when Pelican Bay opened in December 1989, with no psychiatrist on staff, the prison was operating without sufficient mental health care services, see, e.g., Peetz Tr. at 20-3250, and that serious mental health needs were continuing to go unmet in the months and years thereafter.

Associate Warden Peetz, for example, acknowledged that he “was aware that we had inadequate resources to deal with people that were having mental problems.” Tr. 20-3332. This knowledge is also reflected by internal audits and budget requests for additional staff submitted by Warden Marshall, which plainly highlighted a number of serious deficiencies in the delivery of mental health care.

Defendants’ response to the lack of adequate mental health care — and particularly the response of defendant Gomez, who has overall responsibility for the California Department of Corrections — reflects a deplorable, and clearly conscious, disregard for the serious mental health needs of inmates. For example, defendants suggest that, despite lacking a staff psychiatrist — or any semblance of a mental health care program— they were nonetheless justified in opening Pelican Bay, given their “contingency plan” of providing mental health services through periodic visits from psychiatrists at other institutions. However, this plan was so clearly and grossly deficient that it only highlights defendants’ striking indifference to the mental health of thousands of persons in their custody. 160

Prodded by this litigation, as well as litigation in the Coleman case, supra, defendants slowly improved staffing levels over the last two years. However, even this response has been tepid. As discussed above, staffing levels, as of the time of trial, still remained seriously deficient. Nor is it a sufficient response to simply plead that recruitment of doctors is difficult. Defendants certainly knew before Pelican Bay opened that its remote location would present obstacles to attracting professional mental health staff. Moreover, recruitment difficulties do not excuse compliance with constitutional mandates. See Wellman v. Faulkner, 715 F.2d 269, 273 (7th Cir.1988) (failure of prison to fill authorized position for two years weighs “more heavily against the state than for it”). Rather, they simply require that defendants make additional efforts to compensate for their choice of location for the prison — yet those efforts to date can only be described as half-hearted and weak in substance.

In addition to defendants’ slack response to the lack of staffing, defendants have shown little interest in addressing other systemic problems in the delivery of mental health care at Pelican Bay discussed above. For example, defendants have failed to implement a quality assurance program or make serious efforts to provide needed treatment for inmates who, for security reasons, can not be transferred to another institution for inpatient or intensive outpatient treatment.

Defendants emphasize that they have begun plans for initiating a new Health Care Services Division within the Department of Corrections, and that the purpose of this reform is to improve the quality of medical and mental health care and ensure consistent, cost-effective care. While such reform is a step in the right direction, it does not excuse defendants’ deliberate indifference to the mental health needs of inmates at Pelican Bay over the last five years. 161

D. CONDITIONS IN THE SECURITY HOUSING UNIT

The SHU at Pelican Bay is a separate, self-contained complex that operates as the “maxi-SHU” for all of California’s state prisons. Located within the prison’s security perimeter, it is designed to house 1,052 inmates, but has sufficient beds to hold double that number, or 2,104. At the time of trial, the SHU was authorized by the CDC to operate at 150 percent of capacity, bringing the total number of inmates confined in the SHU to approximately 1,500. Roughly two-thirds (or 1,000) of those inmates are double celled, and the remaining 500 inmates are single celled. 162

For the most part, these 1,500 inmates are considered by the CDC to be the most disruptive or potentially dangerous inmates in the California prison system. See Trial Exh. D-130 (designating SHU as housing “of choice” for inmates “who are the greatest threat to prison security and safety”). Roughly half of the 1,500 are inmates who have violated prison rules, usually by possessing a weapon, attempting an escape, or assaulting or participating in an assault on other inmates or staff. They are transferred temporarily to the SHU to serve a set term as punishment for their rule violation(s). The next largest group (numbering around ■ 600) consists of inmates whom the CDC has determined are affiliated with a prison gang. They are assigned to the SHU for indeterminate terms — that is, they will remain in the SHU indefinitely up to the maximum length of their sentence (which, for some prisoners, may mean 10 or 15 years, or the duration of their life). 163 Another sizeable group consists of inmates who are neither gang affiliated nor serving a term for violation of a disciplinary rule; rather they are persons whom prison administrators believe should nonetheless be segregated because of general concerns regarding assaultive or disruptive behavior. These inmates may also remain in the SHU indefinitely. Finally, there are some inmates who are at risk of assault from other inmates and so are housed in the SHU for their own safety. Again, these inmates may remain in the SHU for an indefinite time. 164

Security Housing Units (sometimes referred to as Disciplinary Control Units, Special Management Units, or other similar names) are a common feature in American prisons; their unifying characteristic is that they segregate inmates from other “general population” prisoners and subject them to greater restrictions and fewer privileges. The degree of segregation and restrictions may vary, however, depending on a variety of factors, including penal philosophy and the underlying reason for the inmate’s segregation.

Plaintiffs claim that at Pelican Bay, the degree of segregation is so extreme, and the restrictions so severe, that the conditions in the SHU inflict psychological trauma on inmates confined there, and in some cases, deprive inmates of sanity itself. They further contend that defendants have been deliberately indifferent to the mental health risks posed by the conditions in the SHU. In the remainder of this section, we address: (1) the conditions in the SHU, (2) the impact of SHU conditions on mental health, and (3) defendants’ state of mind.

1. Conditions in the SHU

a. Physical Description

The SHU is a low-level grey structure that roughly resembles a large “X” in shape. There are two separate but physically connected wings which are referred to as the “C” SHU and the “D” SHU. Both wings, which are virtually identical, are divided into “cell blocks”, each of which consists of eight “pods” containing eight cells each. Each pod is divided into two short tiers, with four cells opening onto an upper tier and four cells opening onto a lower tier.

Each cell is 80 square feet and comes equipped with two built-in bunks and a toilet-sink unit. Cell doors are made of heavy gauge perforated metal; this design prevents objects from being thrown through the door but also significantly blocks vision and light. A skylight in each pod does allow some natural light to enter the tier area adjacent to the cells; however, cells are primarily lit with a fluorescent light that can be operated by the inmate. Each cell block is supervised and guarded by a separate control station which is staffed by armed correctional officers and separated from the pods by an electronically controlled metal gate. The officers also electronically control the opening and closing of the cell doors.

Patterned after a “Special Management Unit” in Florence, Arizona (albeit with some modifications), the SHU interior is designed to reduce visual stimulation. See Trial Exh P-3814 at 3955. The cellblocks are marked throughout by a dull sameness in design and color. The cells are windowless; the walls are white concrete. When inside the cell, all one can see through the perforated metal door is another white wall.

A small exercise pen with cement floors and walls is attached to the end of each pod. Because the walls are 20 feet high, they preclude any view of the outside world. The top of the pen is covered partly by a screen and partly by a plastic rain cover, thus pro viding access to some fresh air. However, given their cell-like design and physical attachment to the pod itself, the pens are more suggestive of satellite cells than areas for exercise or recreation.

The overall effect of the SHU is one of stark sterility and unremitting monotony. Inmates can spend years without ever seeing any aspect of the outside world except for a small patch of sky. One inmate fairly described the SHU as being “like a space capsule where one is shot into space and left in isolation.” Lopez Tr. 1-Í9.

b. Social Isolation

Inmates in the SHU can go weeks, months or potentially years with little or no opportunity for normal social contact with other people. Regardless of the reason for their assignment to the SHU, all SHU inmates remain confined to their cells for 22 and f¿ hours of each day. Food trays are passed through a narrow food port in the cell door. Inmates eat all meals in their cells. Opportunities for social interaction with other prisoners or vocational staff are essentially precluded. Inmates are not allowed to participate in prison job opportunities or any other prison recreational or educational programs. Nor is group exercise allowed. Inmates who are single celled exercise alone. Inmates who are double celled exercise with their cellmate or alone if the cellmate chooses not to exercise. No recreational equipment is provided. As the Court observed during its tour of the SHU, some inmates spend the time simply pacing around the edges of the pen; the image created is hauntingly similar to that of caged felines pacing in a zoo. Inmates in adjoining cells can hear but not see each other.

Interaction with correctional staff is kept to an absolute minimum. According to defendants’ expert, the SHU has “attempted to reduce physical contact between inmates and staff to the extent possible, as much probably [as] anyplace I’ve seen in a segregation environment.” Dvoskin Tr. 27-4391. For example, when an inmate leaves his cell to go to the exercise pen, the door is opened automatically by the control booth officer. Once in the tier area, the inmate must strip naked in front of the control booth; the door to the exercise pen is also controlled electronically. In addition, the contact that correctional staff do have with inmates often occurs in a routinized setting while inmates are in handcuffs and waist and ankle chains, such as during an escort from the cell to another point in the prison. As previously found, there is also a pattern of correctional officers using excessive force against inmates. See section 11(A)(1), supra. The resulting tension in the SHU has further limited the ability of inmates and staff to engage in normal and constructive interactions.

The social isolation, however, is not complete. Inmates may leave their pod area on certain specified occasions; however, such opportunities may be infrequent and generally provide only a limited type of interaction. 165 For example, inmates may leave their pod periodically to go to the law library; however, they are assigned to an individual library cell and have little interaction with other inmates or library staff. Inmates may also leave their pod to receive visitors or their attorney; however, all visits are conducted by telephone through a thick glass window, precluding opportunity for human touch. Moreover, because of Pelican Bay’s distance from metropolitan areas, many inmates get either few visitors or none at all. Inmates also attend periodic on-site classification committee meetings, and those who become ill may leave their pod for diagnosis or treatment by the medical or mental health staff. Inmates may also request a counseling, prayer or Bible study visit from a religious volunteer under a program operated by the Pelican Bay chaplain. 166

Roughly two-thirds of the inmates are double celled; however, this does not compensate for the otherwise severe level of social isolation in the SHU. The combination of being in extremely close proximity with one other person, while other avenues for normal social interaction are virtually precluded, often makes any long-term, normal relationship with the cellmate impossible. Instead, two persons housed together in this type of forced, constant intimacy have an “enormously high risk of becoming paranoid, hostile, and potentially violent towards each other.” Grassian Tr. 12-1857; Haney Tr. 6-988-89. The existence of a cellmate is thus unlikely to provide an opportunity for sustained positive or normal social contact.

In sum, those incarcerated in the SHU for any length of time are severely deprived of normal human contact regardless of whether they are single or double celled. As former Warden Fenton testified, conditions in the SHU amount to a “virtual total deprivation, including, insofar as possible, deprivation of human contact.” Tr. 5-808.

c. Privileges

SHU inmates are allowed certain limited privileges which provide a source of environmental stimulation. For the most part, however, they do not involve direct human interaction. Inmates with funds may purchase radios and televisions, and an Arts Film Program is shown on a closed circuit television channel. These televisions and/or radios provide one of the few sources of stimulation or link with the outside world. However, not all inmates possess a television or radio. Inmates may send and receive mail (no phone calls are permitted), read books, and participate in a Bible correspondence class. In recent months, prison administrators have also allowed the mental health staff to provide inmates with reading materials on relaxation techniques. Not all inmates, however, are literate. Inmates may also keep certain personal property in their cells and make purchases through the prison canteen. They are also permitted three showers per week. Other privileges previously mentioned are non-contact visits, participation in the chaplain’s religious visitor program, and an exercise period five times each week.

d. Comparison to Other SHUs

While it is difficult to assess exactly how conditions in the Pelican Bay SHU compare to other security housing units, there is little doubt that, by any measuring stick, the Pelican Bay SHU by design lies on the harsh end of the SHU spectrum. Plaintiffs’ expert Craig Haney, who has toured 20 to 25 segregation units, concluded that inmates at Pelican Bay are more isolated than inmates in any other segregation unit he has experienced. He noted that “[t]he only place that comes close is the federal penitentiary at Marion. But even Marion in some ways is a different and' a less-isolated environment than this one.” Tr. 6-988. Defendants’ expert Dvoskin testified that SHU conditions at Pelican Bay are the conditions “of segregation as they exist in American prisons.” However, he acknowledged that some SHUs provide more “privileges and freedom” than others, and that “Pelican Bay has clearly, on that continuum, decided to err on the side of physical safety rather than ... increased privileges and freedom and increased staff to inmate contact.” Tr. 27-4389-90.

2. Impact of SHU Conditions on Mental Health

Social science and clinical literature have consistently reported that when human beings , are subjected to social isolation and reduced environmental stimulation, they may deteriorate mentally and in some cases develop psychiatric disturbances. These include perceptual distortions, hallucinations, hyper-responsivity to external stimuli, aggressive fantasies, overt paranoia, inability to concentrate, and problems with impulse control. This response has been observed not only in the extreme case where a subject in a clinical setting is completely isolated in a dark soundproofed room or immersed in water, but in a variety of other contexts. For example, similar effects have been observed in hostages, prisoners of war, patients undergoing long-term immobilization in a hospital, and pilots flying long solo flights. While acute symptoms tend to subside after normal stimulation or conditions are returned, some people may sustain long-term effects. This series of symptoms has been discussed using varying terminology; however, one common label is “Reduced Environmental Stimulation,” or “RES.” According to Dr. Gras- sian, the complex of symptoms associated with RES is rarely, if ever, observed in other psychotic syndromes or in humans not subject to RES, a point which defendants did not refute with any specificity.

There is also an ample and growing body of evidence that this phenomenon may occur among persons in solitary or segregated confinement — persons who are, by definition, subject to a significant degree of social isolation and reduced environmental stimulation. Early experiments with complete solitary confinement in American and European penitentiaries in the late 1700’s and 1800’s led to numerous reports of psychiatric disturbances. See Grassian Decl. at 11-16. In 1890, the Supreme Court described the experience with one such facility, the Walnut Street Penitentiary in Philadelphia, in In re Medley, 134 U.S. 160, 10 S.Ct. 384, 33 L.Ed. 835 (1890):

The peculiarities of this system were the complete isolation of the prisoner from all human society, and his confinement in a cell of considerable size, so arranged that he had no direct intercourse with or sight of any human being, and no employment or instruction. Other prisons on the same plan, which were less liberal in the size of their cells and the perfection of their appliances, were erected in Massachusetts, New Jersey, Maryland and some of the other States. But experience demonstrated that there were serious objections to it. A considerable number of the prisoners fell, after even a short confinement, into a semi-fatuous condition, from which it was next to impossible to arouse them, and others became violently insane; others, still, committed suicide; while those who stood the ordeal better were not generally reformed, and in most cases did not recover sufficient mental activity to be of any subsequent service to the community. It became evident that some changes must be made in the system, and the separate system was originated by ■ the Philadelphia Society for Ameliorating the Miseries of Public Prisons, founded in 1787.

Id. at 168, 10 S.Ct. at 386 (emphasis deleted). More recent studies have also documented the potential adverse mental health effects of solitary or segregated confinement. As the Seventh Circuit noted in Davenport v. DeRobertis, 844 F.2d 1310, 1316 (7th Cir.1988), “there is plenty of medical and psychological literature concerning the ill effects of solitary confinement (of which segregation is a variant)” (citing Grassian, Psychopathological Effects of Solitary Confinement, 140 American Journal of Psychiatry 1450. (1983)). 167

Defendants’ expert Dr. Dvoskin acknowledged that it is “possible” that á “syndrome” could be associated with segregated conditions in confinement, although he does not believe there is sufficient data to support “an exact syndrome.” Tr. 27-4373-74. Dr. Dvoskin has, however, used the term “AD SEG [Administrative Segregation] Syndrome” or other terms in his work to describe those people who “can’t handle” segregation or find “segregation intolerable.” Tr. 27 — 4374. Dr. Sheff, the former chief psychiatrist- at Pelican Bay, also testified that he observed prisoners at Pelican Bay demonstrating the RES “symptom complex,” although he did not observe it in-a “large number” of the patients with whom he interacted. 168

Regardless of whether there is an “exact syndrome” associated with incarceration in solitary confinement or security housing units, the Court is well satisfied that a severe reduction in environmental stimulation and social isolation can have serious psychiatric consequences for some people, and that these consequences are typically manifested in the symptoms identified above.

Turning to the case at bar, it is clear that confinement in the Pelican Bay SHU severely deprives inmates of normal human contact and substantially reduces their level of environmental stimulation, as detailed above. It is also clear that there are a significant number of inmates in the Pelican Bay SHU that are suffering from serious mental illness. See section 11(C)(1), supra. At least one Pelican Bay psychologist, Dr. Ruggles, also observed that there was a “psychiatric deterioration that occurred in correlation with placement ... [in the] SHU.” Tr. 17-2914. He did not, however, explain the nature of the deterioration or know the cause. Id. Indeed, the critical question is whether any of the psychiatric problems being experienced by SHU inmates are attributable to conditions in the SHU as opposed to other factors, and if so, the extent and degree of such problems.

To address these issues, Dr. Grassian conducted in-depth interviews with 50 inmates in the SHU over the course of two weeks (in September 1992 and May 1993), and reviewed their medical records. Fourteen inmates were interviewed twice. The inmates were not chosen randomly but were selected because there was some basis to believe that they might be experiencing psychiatric problems. 169

Dr. Grassian concluded that in forty of the fifty inmates, SHU conditions had either massively exacerbated a previous psychiatric illness or precipitated psychiatric symptoms associated with RES conditions. Grassian Tr. 12-1862-63, 1891-92. Of these 40 inmates, 17 were actively psychotic and/or acutely suicidal and in urgent need of inpatient hospital treatment. The other 23 suffered serious psychopathological reactions to the SHU. Grassian Deck at 25. Of the 40 seriously ill inmates, 28 suffered from perceptual disturbances, 35 had problems with concentration, 22 experienced intrusive obsessional thoughts, 29 suffered from paranoia, 28 had impulse dyscontrol, 25 had anxiety/panic disorder, and 24 suffered from overt psychotic disorganization. Ten of the 50 inmates did not appear to be experiencing any significant psychiatric deterioration attributable to the SHU. Grassian Tr. 12-1862-3.

Dr. Grassian concluded that an inmate’s symptoms were attributable to the SHU only where the inmate’s records indicated that the symptoms, or the exacerbation of mental illness, surfaced after confinement in the SHU, and where the inmate was experiencing a constellation of symptoms that is rarely found outside conditions of social isolation and restricted environmental stimulation.

A few examples of Dr. Grassian’s findings are summarized as follows:

Inmate 1 170

Inmate 1, whose records indicate a history of psychiatric illness as an adolescent, was placed in the SHU in November 1990. By April 1992, he was suffering from a paranoid hallucinatory psychosis. He was convinced his food was being poisoned, and was drinking from his toilet and refusing to eat. He reported having auditory and visual hallucinations, claimed that a microphone had been placed in his cell, and was experiencing extreme anxiety. Pelican Bay staff initially asserted that he was malingering, but then also prescribed powerful antipsychotic medicine. A visiting psychiatrist concluded that he had classical symptoms of paranoid schizophrenia and was not being manipulative. On August 28,1992, he was admitted to the infirmary on suicide watch. At that time, a staff psychiatrist diagnosed him as suffering from chronic undifferentiated schizophrenia and recommended that he be transferred to CMF-Vacaville for evaluation and treatment. When Dr. Grassian interviewed him on September 17,1992, Inmate 1 was still in the SHU, actively psychotic and delusionally fearful of being killed. He was eventually transferred to CMF-Vacaville in November, where his clinical state dramatically improved and his psychotic symptoms remitted. The evaluation there indicated that he was: “an immature, needy emotionally underdeveloped young man who simply cannot cope psychologically with the situation that he has made for himself and which he probably never anticipated_ He is genuinely afraid, even panicked, by the Pelican Bay SHU, which seems to have crushed him.” Because his mental state improved at CMF-Vacaville, he was transferred back to the SHU in March of 1993. When Dr. Gras-sian interviewed him a second time in May of 1993, Inmate 1 had again degenerated into a psychotic state; he was agitated, terrified, and hallucinatory. Grassian Decl. at 38-39, 61-64.

Inmate 2

In December 1986, while at the SHU in Folsom Prison, Inmate 2 developed a brief confusional psychosis and saw “little black fuzzy things.” Other than this, his records do not indicate any psychiatric history prior to his incarceration. Prior to his transfer to Pelican Bay, he asked for psychiatric help for his quick and uncontrollable temper and because he had attempted suicide in the past. The examining doctor concluded he did not have a psychiatric problem and recommended no treatment.

Within several weeks of his transfer to the SHU, he had difficulty with insomnia, suicidal and homicidal thoughts, and claustrophobic fears. He was given a low dose of an antidepressant medication. Inmate 2 subsequently developed an overt confusional, paranoid psychosis. At one point he had a severe psychotic episode, during which he became severely confused and hallucinatory; he was eventually cell extracted when he began kicking his cell door in a highly agitated state. 171 He continued to experience a fear and preoccupation with “entities” and “demons.” In July and August of 1992, a staff psychologist assessed him as being in no apparent distress with normal behavior. However, when Dr. Grassian interviewed him in September, he found that Inmate 2 was continuing to suffer from a preoccupation with “entities” and intense fear. In a second interview the following May, Dr. Grassian found that Inmate 2’s thinking was more disorganized than it had been the previous September. He reported “I still have trouble with entities and demons — evil spirits — comic books I read are about the antichrist. I can see them through the walls, black evil. Used to be real heavy. If you pay attention to them, you give in. Mostly it is the devil — no doubt about it_ Got to fight back....” He also described increasing obsessive fears and anxiety about his health: “I fear I’m going to die. I trip on it. I can’t sleep O.K., I can’t relax. My back hurts, my neck hurts_” Grassian Decl. at 56-61.

Inmate S

Inmate 3 has been housed almost continually in the SHU since October 1991. He is unable to read or write and has a history of cognitive difficulties, severe emotional volatility and impulsivity, and wrist cutting. He was psyehiatrically hospitalized in 1987. He is “precisely the type of individual most vulnerable to becoming psyehotically disorganized in [the] SHU.” Grassian Decl. at 110. Once in the SHU his mental state deteriorated. Dr. Grassian found that he was suffering, among other things, from acute psychotic disorganization, perceptual distortions (“like on Television, if things get closer to you, it makes me think I’m going crazy”), and obsessional ruminations. Grassian Decl. at 41, 108-110. He was eventually prescribed a mood stabilizing medication in August of 1992. Id. at 109-10.

Inmate J

Inmate 4 arrived at the Pelican Bay SHU already vulnerable to decompensation. He was institutionalized for much of his childhood and adolescence in state psychiatric hospitals, suffering from developmental disability, a seizure disorder, and behavioral problems, but until his transfer to the SHU in May of 1992, he seemed to have had few behavioral disturbances since 1988.. Given his psychiatric history, however, Dr. Gras-sian noted that “it is not at all surprising that within just a few weeks of his incarceration at Pelican Bay [SHU], he became increasingly psychotic, increasingly agitated, and increasingly out of control.” Deck at 145.

On June 8, 1992, an MTA was called to Inmate 4’s cell after he had ripped the sprinkler head off of the ceiling of his cell and tried to swallow it. He had also attempted to gouge his wrists with a broken plastic spoon. He was, according to Dr. Fulton, “in severe distress, suffering from auditory and visual hallucinations.” Trial Exh. P-480 at 3532. Although the inmate was put on suicide watch in the infirmary, he was later released to the SHU. By the end of July the inmate again felt as if he was “tripping out ... losing it,” and told an MTA that he planned to hurt himself. Id. at 3419. On August 1, 1992, custody officers noticed that he was extremely agitated and tearing up his mattress; on August 13, he was found kicking his cell door in an attempt to escape from “demons”; on August 25, the inmate injured himself by banging his handcuffed hands against the cell wall while trying to hit the “demons”.

The inmate’s psychotic behavior escalated over the next few days. MTAs and correctional officers reported that at times they found him “out of control,” screaming, or incoherent. Trial Exh. P — 480 at 3414, P-158 at 19344. The inmate repeatedly said that he was being attacked by demons and that he would try to kill himself to get away from them. He was observed crying in the comer of his cell on August 29, and an MTA noted that “[the] inmate appears sincere in his suicidal ideation.” Trial Exh. P-158 at 19517. On September 9, he again tried to kill himself by swallowing a piece of the fire sprinkler.

When Dr. Grassian interviewed Inmate 4 less than a week after his last suicide attempt, the inmate was disheveled, despondent, and desperate. He explained that “my heart starts racing, I get dizzy spells, scared, nervous, shaking, crying. I hear voices telling me to tear up my mattress. Demons come out. I see them_ I never saw them before SHU.” Grassian Deck at 148. Inmate 4 was eventually recommended for transfer to the California Medical Facility at Vacaville by the Pelican Bay staff.

In a separate study undertaken by Dr. Haney, 100 randomly chosen SHU inmates were interviewed using a highly structured questionnaire format. 172 Inmates were asked a series of 27 questions, mostly drawn from existing literature relating to RES, which focused on symptoms of psychological distress and the negative effects of prolonged social isolation, including confused thought process, hallucinations, irrational anger, emotional flatness, violent fantasies, social withdrawal, oversensitivity to stimuli, and chronic depression. The study included a control question — whether the inmate had experienced a tingling sensation in the ends of fingers or toes, which is not a symptom of psychological trauma or a psychopathological effect of social isolation. The results showed that a majority of SHU inmates reported a number of the above symptoms, and that many reported “a constellation of symptoms that appears to be related to developing mood or emotional disorders — concerns over emotional flatness or losing ability to feel, swings in emotional responding, and feelings of depressions or sadness that did not easily go away.” Dr. Haney also found that “sizea-ble minorities” reported symptoms that are typically only associated with more extreme forms of psychopathology — hallucinations, perceptual distortions, and thoughts of suicide. Haney Deck at 42. 173

On its face, the study does not indicate that a majority of SHU inmates are experiencing the “more extreme” types of symptoms associated with severe isolation and reduced environmental stimulation. The study also does not specify either the frequency or the degree to which the reported symptoms were experienced; thus, it is difficult to exactly assess the extent of the mental trauma being reported by most inmates. The study also did not review inmate records to compare symptoms reported outside the SHU or otherwise include a control study of non-SHU inmates. The relatively high response to the control question (19 percent) also suggests that the response level may include some exaggeration. 174

Notwithstanding the above, the study is not without some probative value. First, it strongly suggests that many of the symptoms observed by Dr. Grassian are not isolated to the inmates he interviewed but are also likely experienced to some degree by other inmates in the SHU. The study also suggests that the more severe symptoms are only experienced by a minority of the SHU population.

Based on studies undertaken in this case, and the entirety of the record bearing on this claim, the Court finds that many, if not most, inmates in the SHU experience some degree of psychological trauma in reaction to their extreme social isolation and the severely restricted environmental stimulation in the SHU. As one court recently observed in connection with an Illinois state prison, “the record shows, what anyway seems pretty obvious, that isolating a human being from other human beings year after year or even month after month can cause substantial psychological damage, even if the isolation is not total.” Davenport v. DeRobertis, 844 F.2d 1810, 1313 (7th Cir.), cert. denied, 488 U.S. 908, 109 S.Ct. 260, 102 L.Ed.2d 248 (1988)

It is also equally clear that although the SHU conditions are relatively extreme, they do not have a uniform effect on all inmates. For an occasional inmate, the SHU environment may actually prove beneficial. For others, the adverse psychological impact of the SHU will be relatively moderate. They may experience some symptoms but not others, and/or experience those symptoms to a minor or moderate degree. As Dr. Grassian testified, “[t]here clearly are people who are able to tolerate solitary confinement [or] small-group confinement and manifest only some of the symptoms. They don’t reach the point of psychotic disorganization that we see in some of the other prisoners.” Tr. 12-1869. For some, however, the conditions in the Pelican Bay SHU will likely lead to serious mental illness 175 or a massive exacerbation of existing mental illness. 176

The experts are essentially in agreement with respect to the types of persons that are most likely to suffer a serious mental injury from continued exposure to the conditions in the Pelican Bay SHU. Probably most vulnerable are inmates already suffering from mental illness. Dr. Haney testified that prisoners suffering from severe mental disorders should never be subjected to conditions that are as harsh as those imposed in the Pelican Bay SHU. ' Haney Decl. at 67. Defendants’ expert Dr. Dvoskin agreed that segregation may exacerbate pre-existing mental illness and that inmates who are in acute psychiatric distress or suicidal depressions should not be placed in the SHU, absent a few “very, very rare exceptions.” Tr. 27-4466, 4473-74.

Certain inmates who are not already mentally ill are also at high risk for incurring serious psychiatric problems, including becoming psychotic, if exposed to the SHU for any significant duration. As defendants’ expert conceded, there are certain people who simply “can [no]t handle” a place like the Pelican Bay SHU. Persons at a higher risk of mentally deteriorating in the SHU are those who suffer from prior psychiatric problems, borderline personality disorder, chronic depression, chronic schizophrenia, brain damage or mental retardation, or an impulse-ridden personality. Dvoskin Tr. 27-4374-75, 4473-74; Grassian Tr. 12-1869-71, 1882. Consistent with the above, most of the inmates identified by Dr. Grassian as experiencing serious adverse consequences from the SHU were either already suffering from mental illness or fall within one of the above categories. In contrast, persons with “mature, healthy personality functioning and of at least average intelligence” are best able to tolerate SHU-like conditions. Grassian Decl. at 21; see also Tr. 12-1870. Significantly, the CDC’s own Mental Health Services Branch recommended excluding from the Pelican Bay SHU “all inmates who have demonstrated evidence of serious mental illness or inmates who are assessed by mental health staff as likely to suffer a serious mental health problem if subjected to RES conditions.” Trial Exh. P-4495 at 3951.

Defendants were aware that the SHU had, “by design, [been] constructed so that the inmates’ environmental stimulation would be minimized,” Trial Exh. P-4495 at 3948, and that inmates would be subjected to virtually total social isolation. For example, defendants knew, among other things, that inmates in the SHU would have very little direct human contact with staff or inmates, other than possibly a cellmate, for potentially years on end, that visitors would be infrequent, and that there would be no window or view of the outside world from either the exercise pen or the cell.

Defendants were also aware that such conditions could pose a significant risk to the mental health of inmates, particularly for those who are mentally ill or otherwise at a high risk for suffering substantial mental deterioration in the SHU. Defendant Marshall, for example, knew before the SHU opened that RES was a potential risk for inmates, and had “some concerns that [mental decompensation in the SHU] was always a possibility.” Tr. 22-3821; see also Trial Exh. P-4596 at 81280. The CDC’s Mental Health Services Branch (“MHSB”) addressed the potential effects of RES on inmates confined in the Pelican Bay SHU in a September 1989 memorandum entitled “Possible Effects of Reduced Environmental Stimulation on Inmates Confined to the Pelican Bay State Prison.” Trial. Exh. P-4495. That memorandum recommended excluding from the SHU all inmates who were either seriously mentally ill or assessed as likely to suffer a serious mental health problem if subject to RES conditions. Trial Exh. P-4495 at 3951.

The recommendation was largely based on a “dramatic contrast” between two other SHUs which also impose substantial restrictions on human contact and environmental stimulation — one in Marion, Illinois and the other in Florence, Arizona (which served as the model for Pelican Bay). Id. at 3950; Trial Exh. P-3814 at 3954-55. The MHSB report found that the Marion SHU, which excludes mentally ill inmates and those whom the mental health staff feel are at risk for developing a serious psychiatric condition, does not experience a “significant level of psychological decompensation as a result of RES.” Id. However, the Florence SHU, which does not exclude such persons, has experienced “a significantly greater level of adverse behavioral and psychiatric consequences than the Marion facility. In particular, [Florence] cites experiencing problems with their Borderline Personality Disorder inmates who had an increased frequency of suicidal behavior.” Id.

As the record shows, however, the MHSB recommendation was essentially disregarded. As time progressed, defendants were aware that some inmates were developing serious psychiatric problems or suffering a serious exacerbation of an existing mental illness after transfer to the SHU. Notwithstanding this experience, and defendants’ knowledge that RES conditions have the potential for adversely affecting inmate mental health, defendants made no effort to investigate or address whether SHU conditions were adversely affecting inmates suffering from serious psychiatric problems. Nor is there any indication that the MHSB recommendation was given any serious consideration. Rather, inmates who were suffering severe effects from conditions in the SHU were, for the most part, simply medicated with psychotropic drugs or ignored.

E. CELL-HOUSING PRACTICES

Plaintiffs contend that defendants have violated their Eighth Amendment duty to protect inmates from assault by other inmates. They divide this contention into two separate parts. First, they allege that inmates in the maximum security and security housing units suffer a pervasive risk of assault by their cellmates because defendants do not routinely assign to single cells those inmates who have a history of assaulting their cellmate. Second, plaintiffs allege that minimum security inmates face a pervasive risk of harm because they are sometimes assigned to share a cell with a maximum security inmate. Plaintiffs further contend that defendants have been deliberately indifferent to the pervasive risk of harm to inmates from cellmate assaults. At trial, the Court heard testimony regarding this claim from plaintiffs’ expert, Steve Martin, 177 prison officials, and one inmate (Charles Campbell).

1. Double celling of Inmates

k. Overview of Double Celling

As previously described, Pelican Bay is composed of three separate units: (1) a general population unit, designated for persons classified as maximum security (Level IV) prisoners, (2) the SHU, designated the housing “of choice” for inmates “who are the greatest threat to prison security and safety,” Trial Exh. D-130, and (3) a minimum security facility, designated for persons classified as minimum security (Level I) prisoners.

In the general population unit, virtually all of the inmates (roughly 2,000) have cellmates. With respect to the SHU, California regulations provide that housing “shall be in single cells (when possible) in security hous-ing_” CaLCode of Regs., Tit. 15, § 3377.1. Due to ever-increasing population pressures, however, approximately two-thirds of the 1,500 inmates housed in the SHU have cellmates; thus about 1,000 inmates are double celled and about 500 inmates are single celled. Given that approximately one-third of the SHU inmates can be single celled at any given time, the prison has the opportunity to exercise a substantial amount of discretion in determining which SHU inmates should be single celled. The minimum security section of the prison houses approximately 200 inmates in dormitory-style facilities as opposed to traditional cells,

b. Cell-Assignment Decisions

Cell-assignment decisions are made by correctional sergeants. Neither Pelican Bay nor the CDC have promulgated a written policy that sets forth criteria to be used in making cell-assignment decisions for either the General Population section of Pelican Bay or the SHU, CDC regulations do provide that an inmate maybe given single cell status if the inmate “may not be safely housed” in a double cell, but it does not specify any criteria or factors that should be used in making such a decision. Calif.Code Reg., Tit. 15 § 3377.1(c). 178

The informal practice is that correctional sergeants making cell assignments attempt to determine whether two potential cellmates will be compatible by considering such factors as the inmates’ ethnicity, age, length of confinement, gang affiliation, if any, and whether the inmates have any documented “enemies.” Some sergeants may also consider the inmates’ prison job status and whether the inmates are smokers or non-smokers.

Consideration of these factors increases the likelihood of cellmate compatibility and thus reduces the likelihood of cellmate assaults. However, sergeants do not routinely, or as a matter of common practice, consider whether an inmate has a prior history of either assaulting his cellmate or being a victim of such assaults, although such information is usually the most important indicator of whether an inmate will assault a new cellmate or continue to be victimized by such assaults. Nor do prison officials automatically review the cell-assignment or cell-housing status of inmates after they are involved in cell fights. Plaintiffs’ expert found it “unbelievable” that prior assaultive history was not considered in making cell-housing decisions at Pelican Bay. Martin Tr. 8-1270.

An inmate can request single cell status, and this request will be reviewed by a classification committee; however, unless an inmate has actually killed a cellmate or the prison officials are convinced that there will be some mortal danger to an inmate, single cell status is typically denied. Thus,- very few inmates are given single cell status simply because of a history of assaulting cellmates or being the victim of such assaults. At the time of trial, 18 inmates in the SHU were officially designated as being on single cell status, and many of these designations were for reasons unrelated to prior assaultive behavior. 179 Thus, most inmates who have single cells in the SHU are single celled as a matter of chance.

All inmates are also periodically reviewed by a classification committee, which has the authority to single cell inmates even if the inmate has not formally requested single cell status. The central file for each inmate reviewed by the classification committee is made available to the committee. This file contains all documentation of an inmate’s behavior. While the classification committee has designated some inmates for single cell status because of a history of cellmate fights, this happens only rarely.

c. Number of Cell Fights and Resulting Injuries

According to prison records, there were 1,158 reported cell fights at Pelican Bay during a span- of slightly over three years, from the opening of the prison in December 1989 to January 1993. This figure includes 475 cell fights in the general population section and 683 cell fights in the SHU. 180

Most of the inmates involved in the above cell fights were involved in only one or two cell fights at Pelican Bay. However, a disturbing, albeit -relatively small, number of inmates were involved in repeated cell fights while at Pelican Bay: 52 inmates were involved in three cell fights; 15 inmates were involved in four cell fights; 14 inmates were involved in five cell fights; 4 inmates were involved in six to eight cell fights, and 2 inmates were involved in nine cell fights. 181 Thus, between December 1989 and January 1993 prison officials double celled 35 inmates even alter they had a well established history of assaulting three prior cellmates at Pelican Bay.

In many of these cases, the cell fights occurred over a short time frame. For example, prisoners with 6 or 7 reported cell fights often had 5 or 6 of the fights within a one-year period. While a few of these repeat offenders were ultimately single celled in light of their history of cellmate assaults, the vast majority were not. Plaintiffs’ expert testified that it was “unparalleled” in his experience to see this type of repeated fighting with cellmates within such a short period of time.

Many of these cell fights have resulted in serious injuries to the victimized inmate. Among the many injuries sustained are: fractured ribs (Julio Vasquez), major trauma to hand (David Funches), coma, paralysis and loss of eye (Miguel Barraza), severe bleeding (Alfredo Martinez), large facial wound (Jaime Pena), facial lacerations (Jose Lopez), and brain damage and disability (Allyn Hopkins),

d. Defendants’ State of Mind

While some evidence was presented bearing on defendants’ state of mind, we conclude that it falls short of proving that defendant Warden Marshall or defendant Chief Deputy Warden Peetz had actual knowledge of the extent to which inmates were repeatedly assaulting cellmates but continuing to be double celled. While information concerning cell fights is documented, the record is less than clear as to whether that information reached defendants Marshall or Peetz in any systematic way. Nor is there persuasive evidence that they actually knew, from other sources, the general number of inmates that were continuing to be double celled despite the fact that they had previously assaulted three or more cellmates.

2. Temporary Housing of Minimum Security Inmates with Maximum Security Inmates in the Same Cell

a. Classification System

All persons entering the CDC penal system are given a classification score which determines an inmate’s security level. Based on this score, an inmate will be given a designation ranging from Level I — reserved for the lowest security risk prisoner — to Level IV — reserved for the highest security risk prisoner. The score is arrived at by tabulating points that are based on an array of objective factors which include, among other things, length of sentence, nature of the crime committed, criminal history, employment history, military service, marital status, age, prior escape attempts, and prior incarceration behavior. See Cal.Code of Regs., tit. 15, § 3375.3. 182 An inmate’s classification score can be subsequently adjusted based on an inmate’s conduct and behavior in prison.

Level I “minimum security” prisoners are typically serving short terms for relatively minor felony offenses, including’drunk driving, certain drug offenses, and minor property-related offenses. Level IV “maximum security” prisoners are typically more sophisticated and experienced offenders who have committed serious offenses and are serving longer sentences.

Under California regulations, an inmate should normally be housed in a facility with a classification level that is commensurate with the inmate’s score. However, there are certain “administrative determinants” which allow the CDC to confine inmates in a prison that is not commensurate with their classification point score. Id. at § 3375.2 One such administrative determinant is that “[a]n inmate with a felony hold, warrant, detainer, or the equivalent thereof filed with the department who is likely to receive a significant period of consecutive incarceration or be deported, shall not be housed in a Level I facility without perimeter gun towers.” Id. at § 3375.2(a)(4).

b. Interaction of Level I and Level IV Prisoners

The general population section of Pelican Bay is reserved for Level IV maximum security prisoners. Level I minimum security, prisoners are housed in a separate dormitory style facility outside the perimeter towers. Prisoners must have security clearance before they can be released to the minimum security facility. This requires, among other things, a determination that the inmate has no outstanding warrants or detainers.

Minimum security inmates may be temporarily housed in the general population section of Pelican Bay for one of two reasons. The first is if the inmate is charged with a disciplinary violation. When this occurs, the inmate will likely be transferred to the general population section pending the outcome of the investigation of the charges. This process can take up to a month and a half, and in a few cases longer.

The second is if a Level I inmate is newly arrived, and the prison is awaiting arrest history and other background documents needed to determine whether the inmate can be cleared for placement in the minimum security facility. See CaLCode of Regs, Tit. 15 § 3375.2(a)(4). For reasons left unexplained at trial, this process can take anywhere from three weeks to two months or, in some cases, as long as several months.

Beginning in September 1992, a gymnasium in the General Population section was converted into a dormitory to use exclusively for incoming Level I inmates who were awaiting clearance to the minimum security facility. This gymnasium was closed temporarily in February 1993, but subsequently reopened. Thus, most incoming Level I inmates are segregated in the gymnasium. Some, however, are still temporarily housed with Level IV inmates depending upon available bed space. It is also possible that future budgetary constraints could force closure of the gymnasium.

Plaintiffs have identified two instances where Level I inmates were assaulted by Level IV cellmates. In one such case, a Level I inmate, Charles Campbell, had the tip of his nose bitten off during an attack by his Level IV cellmate. There was also some evidence that Level IV inmates may subject Level I cellmates to other types of “pressure tactics”; however, the exact nature of these tactics, and their frequency, was not developed in the record.

F. SEGREGATION OF PRISON GANG AFFILIATES

The CDC has determined, and plaintiffs do not dispute, that gangs present a serious threat to the safety and security of California prisons. 183 The CDC DOM statement of philosophy provides, in part, that “prison gangs and disruptive groups through their illegal activities are a threat to the security of all [prison] institutions ... [and] are also a definite danger to public order and safety”. DOM § 55070.5.

The term “prison gangs” refers to gangs that originate within the prison system. Such gangs first developed in California in the late 1950s and 1960s, and now include the Aryan Brotherhood, the Black Guerrilla Family (“BGF”), the Mexican Mafia (“EME”), and the Nuestra Familia (“NF”), as well as the Northern Structure (“NS”), the Texas Syndicate, the Vanguards, the Mexikanemi, and the New Mexico Syndicate. The term “disruptive groups” refers to gangs which originate outside of prison, such as street gangs, white supremacist groups, right- or left-wing revolutionary groups, and motorcycle gangs. DOM § 55070.17.4.

Although both prison gangs and disruptive groups pose threats to prison security, prison gangs are considered the greater threat. One gang investigator explained that this is because “prison gangs have within their own policy a mandatory ruling that [members] must participate in gang behavior, where the disruptive groups do not have that mandatory ruling ... Prison gangs also attempt to control the criminal enterprises of the prison system and attempt to ... exercise unlawful influence over the other inmates to participate in their behavior.” Hawkes Tr. 16-2613; see also Gomez Tr. 28-4610 (gang affiliated prisoners are the “most disruptive to the day-to-day management of a prison system”).

Activities typically associated with prison gangs include loan sharking and extortion, drug trafficking, and premeditated assaults ranging from unarmed attacks to fatal stabbings. Gang members also pressure nonmembers in the general prison population to assist gang activities by smuggling or obtaining weapons or providing information. Inmates may join a prison gang for a variety of reasons, including the desire to gain increased status in the inmate population or to obtain the “protection” that gang affiliation may offer. Some inmates may join under pressure from other inmates. Any new member, however, must pledge allegiance to the gang for life.

Under California regulations, the prison may place any inmate in administrative segregation whose presence in the general prison population “endangers institution security.” CaLCode Regs. tit. 15 § 3335(a). 184 Gang affiliation 185 is one such threat to prison security. Accordingly, once the CDC determines that an inmate is a member or associate of a prison gang, the inmate is routinely transferred to administrative segregation in the SHU. 186

The aim of this policy is to promote the overall security of the California prison system by taking a “pro-active stance in the arena of gang suppression.” DOM § 55070.1; Gomez Tr. at 28-4610 (gang affiliated prisoners are not placed in the SHU as a form of punishment for specific behavior but for “the safety and security of both inmates and staff in the Department of Corrections”). At the time of trial, approximately '625 inmates were confined in the SHU based on prison gang affiliation.

Inmates transferred to the SHU for prison gang affiliation are normally given an indeterminate term. This means that the inmate will remain in the SHU for the duration of his prison term, unless the inmate “drops out” of the prison gang by successfully completing what is referred to as a “debriefing” process. As one inmate succinctly testified, “the only way [a prison gang member] can get out of [the SHU] is to debrief, parole, or just die of old age.” Trujillo Tr. 9-1469.

“Debriefing” requires the inmate to admit that he was a gang member, identify other gang affiliates, and reveal everything he knows about the gang’s activities and organizational structure. Because prison gang members join “for life,” the CDC considers debriefings necessary to prove that renunciations of gang membership are genuine. As the DOM explains, the purpose of a “debriefing” is to “obtain sufficient verifiable information from the subject which adversely impacts the gang so the gang will no longer accept the subject as either a member or associate.” DOM § 55070.20.1. Although no evidence of actual reprisals was introduced at trial, a number of prison staff agree that inmates who debrief and gain release from the SHU are considered “snitches,” and thus face serious risks of being attacked or even killed by other inmates. Thus, a few inmates have elected to remain in the SHU for their own safety even after. debriefing. Defendants do not permit gang members to “drop out” of the prison gang simply by refraining, or promising to refrain, from participating in gang activities or associating with gang affiliates while in the SHU.

1. Procedure for Establishing Prison Gang Affiliation

The procedure for establishing gang membership or association is referred to as the “validation” process. Every institution within the CDC, including Pelican Bay, employs at least one Institutional Gang Investigator (“IGI”) 187 who is responsible for tracking gang activities and investigating those suspected of gang membership. Whenever a gang investigator obtains evidence that an inmate has associated with other gang affiliates, it is noted in the inmate’s central file (“C-file”). If the evidence is tangible, such as a membership list, gang constitution, letter, or photographs, the gang investigator will store the object itself in the C-file. If the evidence is intangible, such as the statement of a confidential informant, a staff observation of association, visits from persons with known gang connections, or an oral confession, the IGI prepares the appropriate paperwork memorializing the evidence and places it in the inmate’s C-file. There is no written requirement that contemporary information be relied on in making the initial validation; however, the practice at Pelican Bay is to rely on at least some current information.

The most common item of evidence is the statement of another inmate, generally referred to as a statement from a “confidential informant” or “CL” California regulations preclude reliance on such statements, unless “other documentation corroborates information from the source, or unless the circumstances surrounding the event and the documented reliability of the source satisfies the decision makers(s) that the information is true.” Cal.Code of Regs. tit. 15, § 3321(b)(1). These regulations also provide that reliability can be established if any one of the following criteria is satisfied: (1) the confidential source has previously provided information which proved to be true, (2) other confidential sources have independently provided the same information, (3) the information provided by the confidential source is self-incriminating, (4) part of the information provided is proven true, or (5) the confidential source is the victim. Id. at § 3121(c).

In order to “validate” an inmate as a full “member” of a gang, CDC regulations require that IGIs identify a minimum of three “original, independent source items of documentation indicative of actual membership.” DOM § 55070.19.2. In order to validate an inmate as an “associate,” CDC regulations require that IGIs identify a minimum of three “original, independent source items of documentation indicative of association with VALIDATED gang members and/or associates.” DOM § 55070.19.3. (emphasis in original). 188

Once an IGI believes that there is sufficient documentation to validate an inmate, the IGI prepares a “validation package” for submission to the Special Services Unit (“SSU”) in Sacramento, California. This package includes photocopies of each source document relied upon, a written itemization of the evidence, and a description of the inmate’s distinctive markings and tatoos, if any. Once the package is completed, the inmate is brought to the office of the IGI, where the inmate is told that he is suspected of gang affiliation, and provided a copy of a form summarizing the evidence relied upon.

When the evidence in the validation package includes information from a confidential informant, the inmate is provided with a Confidential Information Disclosure Form which briefly summarizes the substance of the accusation, insofar as that can be done without disclosing the informant’s identity. The form also identifies the basis for the IGI’s determination that the information is reliable. This typically consists of a eonclu-sory statement that the informant has provided reliable information in the past. The cursory nature of the information provided to the inmate makes it difficult to challenge evidence provided by confidential informants.

Nonetheless, the inmate is given an opportunity to present his views to the IGI and contest his alleged gang affiliation. He is not, however, given an opportunity to present evidence, examine witnesses or obtain assistance. If the IGI decides to pursue the validation after meeting with the inmate, the IGI submits the validation package to the SSU in Sacramento. 189

The SSU performs a quality review check to ensure consistency between institutions and to confirm that the IGI has submitted at least three items of evidence which satisfy CDC regulations. As a practical matter, this review is largely superficial since the SSU presumes that the documentation relied upon in the validation package is accurate unless there is an obvious or blatant flaw. The SSU does not interview the inmate as part of its review.

If a package contains more than the minimum three items of evidence, the SSU may reject certain items as “not acceptable” or “not usable,” and still validate the inmate so long as at least three items remain that are not rejected. Bruce Depo. at 182. 190 In such a case, the SSU does not inform the IGI or the prison that certain items have been rejected. As SSU Agent Addison explained, “if [the IGI] had only sent in three pieces of gang evidence, then they would know that all three were relied on, but if there were more than three sent in they wouldn’t know which three were relied on.” Addison Depo. at 144.

If the package appears to be in order, the SSU will officially “validate” the inmate as a member or associate of a prison gang.. This occurs in the overwhelming number of validation packages submitted for approval. Of over 300 packages submitted from Pelican Bay over a three-year period, only two were rejected. IGI Hawkes testified that only one of the packages he had submitted had ever been permanently rejected by the SSU, and that case involved a request to validate an inmate as a “drop out” of a prison gang. Hawkes Tr. 16-2676. 191

2. Assignment of Prison Gang Affiliates to the SHU

Once an inmate is validated as a gang member or associate by the SSU, an Institutional Classification Committee (“ICC”) is convened to decide whether the inmate should be retained in the SHU for an indeterminate term based on his gang affiliation. Cal Code Regs. tit. 15, § 3338(d). ICCs are composed of an Associate Warden, a Program Administrator, a Correctional Counsel- or II, and a Correctional Counselor I, and are charged with the responsibility for major classification decisions, including transfers.

Given that it is CDC policy to confine validated gang affiliates to the SHU for an indeterminate term, this is invariably the outcome of the ICC meeting. ICCs do not, as a general matter, closely reexamine the underlying basis of a new validation. The inmate, however, is brought to the ICC meeting and given an opportunity to address the ICC before a final decision is made. 192 ICC recommendations of transfer to a security housing unit for an indeterminate term must be “endorsed” by a Classification Staff Representative. Cal.Code Regs., tit. 15, §§ 3341.5, 3379. In the case of assignments to the SHU for gang affiliation, this approval is routine.

3. Subsequent Reviews of Status in the SHU

Inmates assigned to indeterminate terms in the SHU for gang affiliation are afforded two types of periodic reviews. A Unit Classification Committee (“UCC”) reviews an inmate’s indeterminate SHU assignment every 120 days, 193 and the ICC, described above, reviews an inmate’s SHU indeterminate status every 12 months. The inmate attends each UCC and ICC review and has an opportunity to address the committee, although sometimes the inmate will simply be asked if he is “ready to debrief.” Trujillo Tr. 9-1467. The inmate does not have an opportunity to present or examine witnesses; staff assistance may be provided occasionally.

The UCCs are composed of a Program Administrator, a Correctional Counselor II and a Correctional Counselor I. They have less authority than the ICCs and usually attend to day-to-day programming and less significant classification matters. Thus, unlike the ICC, the UCC is not empowered to reconsider an inmate’s validated status and order his release from segregation. However, the UCC can inquire into the propriety of the validation and recommend a change in status. One program administrator testified, regarding UCC reviews, that “[t]here have been occasions in the past where we have discovered information that was not corroborated or validated appropriately, in our view. When that has happened, we’ve returned the case to the IGI to re-evaluate their validation and to provide additional information to support the validation, if it exists.” Helsel Tr. 21-3539-40.

Prior to the ICC annual review, an IGI reviews the inmate’s file, compiles any new evidence pertaining to gang membership or activity, considers whether any previously relied upon evidence has been called into doubt, and determines whether there is still a sufficient evidentiary basis for satisfying current CDC validation requirements. The absence of any new evidence linking the inmate to gang activity or gang members is neither noted nor considered relevant. As long as the initial evidence used to validate the inmate still meets CDC requirements and has not been called into doubt, the inmate will be retained in the SHU as a validated gang affiliate. 194

III.

CONCLUSIONS OF LAW

A. EIGHTH AMENDMENT OVERVIEW

By virtue of their conviction, inmates forfeit many of their constitutional liberties and rights: they are isolated in prisons, and subject to stringent restrictions that govern every aspect of their daily lives. Nonetheless, those who have transgressed the law are still fellow human beings — most of whom will one day return to society. 195 Even those prisoners at the “bottom of the social heap_have, nonetheless, a human dignity.” Toussaint v. McCarthy (Toussaint VI), 926 F.2d 800, 801 (9th Cir.1990), cert. denied, 502 U.S. 874, 112 S.Ct. 213, 116 L.Ed.2d 171 (1991). In recognition of this fundamental principle, our jurisprudence is clear: while incarceration may extinguish or curtail many rights, the Eighth Amendment’s protection against cruel and unusual punishment still retains its “full force” behind prison doors. Michenfelder v. Sumner, 860 F.2d 328, 335 (9th Cir.1988).

It is a right animated by “broad and idealistic concepts of dignity, civilized standards, humanity, and decency.” Estelle v. Gamble, 429 U.S. 97, 102, 97 S.Ct. 285, 290, 50 L.Ed.2d 251 (1976); see also Hudson v. McMillian, 503 U.S. 1, 11, 112 S.Ct. 995, 1001, 117 L.Ed.2d 156 (1992); Patchette v. Nix, 952 F.2d 158, 163 (8th Cir.1991) (Eighth Amendment “ ‘draw[s] its meaning from the evolving standards of decency that mark the process of a maturing society”’); Michenfelder, 860 F.2d at 335; Spain v. Procunier, 600 F.2d 189, 200 (9th Cir.1979) (Eighth Amendment is based on the “fundamental premise that prisoners are not to be treated as less than human beings”). It is a right that recognizes that in a country such as ours, which aspires to the highest standards of civilization, there is simply no place for abuse and mistreatment, even in the darkest of jailhouse cells.

Consistent with these humanitarian concepts, our courts have long acknowledged that when the State, by imprisonment, prevents a person from caring for himself, the Constitution imposes “ ‘a corresponding duty to assume some responsibility for his safety and general well being.’ ” Helling v. McKinney, — U.S. —, —, 113 S.Ct. 2475, 2480, 125 L.Ed.2d 22 (1993). “[H]aving stripped [prisoners] of virtually every means of self-protection and foreclosed their access to outside aid,” society may not simply lock away offenders and let “the state of nature take its course.” Farmer v. Brennan, — U.S. —, —, 114 S.Ct. 1970, 1977, 128 L.Ed.2d 811 (1994). Rather, government officials must ensure that prisons, while perhaps “restrictive and even harsh,” Rhodes v. Chapman, 452 U.S. 337, 347, 101 S.Ct. 2392, 2399, 69 L.Ed.2d 59 (1981), do not degenerate into places that violate basic standards of decency and humanity. In short, while the Eighth Amendment “ ‘does not mandate comfortable prisons’ ... neither does it permit inhumane ones.” Farmer, — U.S. at —, 114 S.Ct. at 1976 (citation omitted).

Thus, it is well past dispute that the Eighth Amendment requires that prison officials provide inmates with such minimum essentials as adequate food,.shelter, clothing, medical care, and safety.

When the State by the affirmative exercise of its power so restrains an individual’s liberty that it renders him unable to care for himself, and at the same time fails to provide for his basic human needs — e.g. food, clothing, shelter, medical care, and reasonable safety — it transgresses the substantive limits on- state action set by the Eighth Amendment.

Helling, — U.S. at —, 113 S.Ct. at 2475 (quoting DeShaney v. Winnebago County Dep’t. of Social Services, 489 U.S. 189, 199-200, 109 S.Ct. 998, 1005-06, 103 L.Ed.2d 249 (1989)); see also Farmer, — U.S. at —, 114 S.Ct. at 1976; Toussaint v. McCarthy (Toussaint TV), 801 F.2d 1080, 1107 (9th Cir.1986) (“human needs that prison officials must satisfy include food, clothing, sanitation, medical care, and personal safety”), cert. denied, 481 U.S. 1069, 107 S.Ct. 2462, 95 L.Ed.2d 871 (1987).

The Eighth Amendment also prohibits those who operate .our prisons from using “excessive physical force against inmates.” Farmer, — U.S. at —, 114 S.Ct. at 1976; Hoptowit v. Ray, 682 F.2d 1237, 1246, 1250 (9th Cir.1982) (prison officials have “a duty to take reasonable steps to protect inmates from physical abuse”); see also Vaughan v. Ricketts, 859 F.2d 736, 741 (9th Cir.1988), cert. denied, 490 U.S. 1012, 109 S.Ct. 1655, 104 L.Ed.2d 169 (1989) (“prison administrators’ indifference to brutal behavior by guards toward inmates [is] sufficient to state an Eighth Amendment claim”). As courts have succinctly observed, “[p]ersons are sent to prison as punishment, not for punishment.” Gordon v. Faber, 800 F.Supp. 797, 800 (N.D.Iowa 1992) (citation omitted), aff'd, 973 F.2d 686 (8th Cir.1992). “Being violently assaulted in prison is simply not ‘part of the penalty that criminal offenders pay for their offenses against society.’” Farmer, — U.S. at —, 114 S.Ct. at 1977 (quoting Rhodes, 452 U.S. at 347, 101 S.Ct. at 2399).

In order to prevail on any Eighth Amendment claim alleging cruel and unusual punishment, a plaintiff must satisfy two requirements:

First, the deprivation alleged must be, objectively, sufficiently serious; a prison official’s act or omission must result in the denial of the minimal civilized measure of life’s necessities. For a claim ... based on a failure to prevent harm, the inmate must show that he is incarcerated under conditions posing a substantial risk of serious harm....

The second requirement follows from the principle that only the unnecessary and wanton infliction of pain implicates the Eighth Amendment. To violate the Cruel and Unusual Punishments Clause, a prison official must have a sufficiently culpable state of mind.

Farmer, — U.S. at —, 114 S.Ct. at 1977 (internal quotations and citations omitted); see also Wilson v. Seiter, 501 U.S. 294, 297, 111 S.Ct. 2321, 2323, 115 L.Ed.2d 271 (1991). Thus, every Eighth Amendment claim embodies an objective and subjective component. Wilson, 501 U.S. at 297-98, 111 S.Ct. at 2323-2326. The former focuses on whether there has been a deprivation or infliction of pain serious enough to implicate constitutional concerns, while the latter requires inquiry into the defendant’s state of mind to determine whether the infliction of pain was “unnecessary and wanton.” Jordan v. Gardner, 986 F.2d 1521, 1525-28 (9th Cir.1993) (en banc).

In considering whether the objective component has been met, the Court must focus on discrete and essential human needs such as health, safety, food, warmth or exercise. Wilson, 501 U.S. at 304, 111 S.Ct. at 2327. “Courts may not find Eighth Amendment violations based on the ‘totality of conditions’ at a prison.” Hoptowit, 682 F.2d at 1246 (quoting Wright v. Rushen, 642 F.2d 1129, 1132 (9th Cir.1981)). Thus, while courts may consider conditions in combination “when they have a mutually enforcing effect that produces the deprivation of a single, identifiable human need ...[,] [n]othing so amorphous as ‘overall conditions’ can rise to the level of cruel and unusual punishment when no specific deprivation of a single human need exists.” Wilson, 501 U.S. at 304-05, 111 S.Ct. at 2327. The question whether the objective component of an Eighth Amendment claim has been met presents an issue of law for the court to decide. Hickey v. Reeder, 12 F.3d 754, 756 (8th Cir.1993).

In contrast, the state of mind inquiry presents a question of fact, and is “subject to demonstration in the usual ways, including inference from circumstantial evidence.” Farmer, — U.S. at —, 114 S.Ct. at 1981. For most Eighth Amendment claims, the plaintiff satisfies the culpability requirement by proving that the defendants’ actions (or omissions) constitute “deliberate indifference.” This “baseline” standard, Jordan, 986 F.2d at 1527, applies in cases alleging inadequate protection from injury from other inmates or inhumane conditions of confinement that deprive an inmate of a basic necessity of life, such as shelter, food, health or exercise. See Farmer, — U.S. at —, 114 S.Ct. at 1977; Jordan, 986 F.2d at 1528.

As the Supreme Court recently clarified, the test for determining “deliberate indifference” is essentially equivalent to the standard for establishing subjective recklessness in criminal cases. Farmer, — U.S. at —, 114 S.Ct. at 1980. Thus, the plaintiff must show that:

the [prison] official knows of and disregards an excessive risk to inmate health or safety; the official must both be aware of facts from which the inference could be drawn that a substantial risk of serious harm exists, and he must also draw the inference.

Id. at —, 114 S.Ct. at 1979. In other words, the defendant must “consciously disregard a substantial risk of serious harm.” Id. at —, 114 S.Ct. at 1980 (internal quotation omitted). Such a standard presupposes that the defendant has not acted reasonably in the face of a known risk. Thus, a prison official can avoid liability if he “responded reasonably to the risk, even if the harm ultimately was not averted.” Id. at — - —, 114 S.Ct. at 1982-83. “Whether one puts it in terms of duty or deliberate indifference, prison officials who act reasonably cannot be found liable under the Cruel and Unusual Punishments Clause." Id. at —, 114 S.Ct. at 1983.

In sum, deliberate indifference occurs where the prison official “knows that inmates face a substantial risk of serious harm and disregards that risk by failing to take reasonable measures to abate it.” Id. at —, 114 S.Ct. at 1984. This standard does not require plaintiffs to “show that a prison official acted or failed to act believing that harm actually would befall an inmate; it is enough that the official acted or failed to act despite his knowledge of a substantial risk of serious harm.” Farmer, — U.S. at —, 114 S.Ct. at 1981. Nor does this standard mean that “prison officials will be free to ignore obvious dangers.” Id. While the obviousness of a risk is not conclusive, a factfin-der may “conclude that a prison official knew of a substantial risk from the very fact that the risk was obvious.” Id. at — - — and n. 8, 114 S.Ct. at 1981-82 and n. 8. Similarly, a defendant would “not escape liability if the evidence showed that he merely refused to verify underlying facts that he strongly suspected to be true, or declined to confirm inferences of risk that he strongly suspected to exist.” Id. at — n. 8, 114 S.Ct. at 1982 n. 8; see also McGill v. Duckworth, 944 F.2d 344, 351 (7th Cir.1991) (“Going out of your way to avoid acquiring unwelcome knowledge is a species of intent”), cert. denied, 503 U.S. 907, 112 S.Ct. 1265, 117 L.Ed.2d 493 (1992).

Although the deliberate indifference standard governs most claims, an even higher degree of culpability must be shown in one type of claim: when an inmate seeks to hold an individual prison officer liable for using excessive physical force against the inmate during a particular incident. Hudson, 503 U.S. at 1, 112 S.Ct. at 995; Whitley v. Albers, 475 U.S. 312, 320, 106 S.Ct. 1078, 1084, 89 L.Ed.2d 251 (1986).

In Whitley, an inmate was shot and seriously wounded during the course of a prison riot. The inmate sought damages under 42 U.S.C. § 1983 against the individual prison guards and officials directly involved in the incident, alleging use of excessive force. 475 U.S. at 316-317, 106 S.Ct. at 1083. The Supreme Court held that in order to prevail, the plaintiff must show more than deliberate indifference; he must show that the force used against him was applied, not in a “good faith effort to maintain or restore order, [but] maliciously and sadistically for the very purpose of causing harm.” Id. at 320-21, 106 S.Ct. at 1085 (internal quotations omitted).

In 1992, the Supreme Court revisited the state of mind, issue in Hudson, in which an inmate alleged that prison guards had beaten him for no reason during an escort. Extending Whitley beyond the context of a riot, the Supreme Court concluded that the “maliciousness” standard controlled, not just in major prison disturbances, but in smaller incidents as well. The Court held that “whenever prison officials stand accused of using excessive physical force ... the core judicial inquiry is ... whether force was applied in a good-faith effort to maintain or restore discipline, or maliciously and sadistically to cause harm.” Hudson, 503 U.S. at 7, 112 S.Ct. at 999. Put another way, plaintiffs must show that “officials used force with ‘a knowing willingness that [harm] occur.’ ” Farmer, — U.S. at —, 114 S.Ct. at 1978.

In determining whether the maliciousness standard has been met in any given ease, the factfinder may draw inferences from circumstances surrounding the challenged conduct. To assist this process, the Supreme Court identified five factors that should be taken into consideration: (1) the extent of the injury suffered, (2) the need for the application of force, (3) the relationship between that need and the amount of force used, (4) the threat reasonably perceived by the responsible officials, and (5) any efforts made to temper the severity of a forceful response. Hudson, 503 U.S. at 7, 112 S.Ct. at 999; Romano v. Howarth, 998 F.2d 101, 105 (2nd Cir.1993).

B. EXCESSIVE FORCE

Plaintiffs advance two related but distinct theories of Eighth Amendment liability with respect to their excessive force claim. The first is that there is a pattern of prison staff using excessive force against inmates, and that defendants, in permitting this pattern to develop and persist, have acted not only with deliberate indifference, but also with the malicious purpose of causing harm. The second is that defendants have failed to establish adequate systems for controlling use of force by prison staff against inmates — despite knowing the grave risks to inmate safety that this failure creates — and that inmates have suffered serious injuries as a result. Because we find that liability is established under the first theory, we do not address the second. 196

Although the “pattern of excessive force” theory is well established in class-action prison litigation, most, if not all, such cases predate the Supreme Court’s recent decisions in Wilson (which imposed a state of mind requirement in every Eighth Amendment claim) and Hudson (which extended Whitley’s “maliciousness” standard beyond the context of the prison riot). Thus, this action raises a number of issues concerning the proper application of the subjective and objective components of the Eighth Amendment in class-action excessive force claims. We first address the subjective component.

1. Subjective Component

As discussed above, Hudson and Whitley clearly delineate the state of mind standard applicable when an individual inmate seeks to recover for the use of excessive force in a particular incident: the inmate must show that the defendant used force maliciously and sadistically for the purpose of causing harm rather than in a good faith effort to restore or maintain order. Neither case, however, addresses the culpability requirement in the context of a class action, where the defendants are not individual officers accused of using excessive force in a particular instance, but Wardens and high ranking administrators who are charged with overall operation of the prison.

The positions of the parties in this uncharted area are predictable: plaintiffs argue that Hudson_ and Whitley do not apply, and that the appropriate benchmark for measuring the culpability of the defendants in this case is “deliberate indifference,” a result consistent with pre-Hudson authority. See, e.g., Fisher, 692 F.Supp. at 1564 (applying deliberate indifference test in class action excessive force case). Defendants, on the other hand, contend that the “maliciousness” test, articulated in Whitley and extended in Hudson, should control. We are not persuaded, however, for the reasons explained below, that Hudson requires application of the maliciousness standard to the case at bar.

First, on its face, Hudson applies to the situation where a prison official is accused of using excessive force or otherwise directly participating in an incident of excessive force. Thus, Hudson described the “maliciousness” test as applying “whenever prison officials stand accused of using excessive physical force.” 503 U.S. at 6-7, 112 S.Ct. at 999 (emphasis added). As the Supreme Court recently observed in Farmer, use of the deliberate indifference standard is “inappropriate in one class of prison cases: when ‘officials stand accused of using excessive physical force.’ ” — U.S. at —, 114 S.Ct. at 1978 (internal quotations omitted) (emphasis added); see also Buckner v. Hollins, 983 F.2d 119, 122 (8th Cir.1993) (“The Whitley/Hudson standard ... applies to eases in which a prison official is accused of using excessive force”) (emphasis added).

In this case, however, no defendant is charged with using excessive force or otherwise participating in any particular incident. See Defendants’ Trial Brief at 39 (“None [of the defendants] are alleged personally to have applied any force to any plaintiff’). Rather, as top-ranking administrators, they are charged with conduct of a completely different nature: abdicating their duty to supervise and monitor the use of force and deliberately permitting a pattern of excéssive force to develop and persist.

Second, the rationale underlying Hudson does not justify its extension to the case at bar. Hudson reasoned that in contrast to claims of inadequate medical care, where “the State’s responsibility to provide inmates with medical care ordinarily does not conflict with competing administrative concerns,” claims alleging the misuse of force in a particular instance implicate the prison’s competing obligation to restore and maintain order. Hudson, 503 U.S. at 6, 112 S.Ct. at 998; see also Whitley, 475 U.S. at 320, 106 S.Ct. at 1084. Moreover, a guard responding to a disruption must make immediate judgments regarding the amount of force needed under the circumstances. Thus, officers are placed in the dilemma of hastily balancing two competing interests: the interest in protecting the inmate from unnecessary pain and suffering and the interest in restoring and maintaining order.

Whether the prison disturbance is a riot ,[as was the case in Whitley] or a lesser disruption [as in Hudson ], corrections officers must balance the need to maintain or restore discipline through force against the risk of injury to inmates. Both situations may require prison officials to act quickly and decisively.

Hudson, 503 U.S. at 6, 112 S.Ct. at 998-99 (internal quotation omitted); see also Whitley, 475 U.S. at 320, 106 S.Ct. at 1084 (when confronted with a disturbance, prison officials must balance competing concerns of safety of others and risk of injury to the inmate “in haste, under pressure, and frequently without the luxury of a second chance”); Jordan, 986 F.2d at 1528 (“Because the critique of such [hastily made] decisions in hindsight could chill effective action by prison officials, the Supreme Court has held that the higher [maliciousness] standard is appropriate”); Farmer, — U.S. at —, 114 S.Ct. at 1978 (“[W]here the decisions of prison officials ... [to use force] are made in haste ... an Eighth Amendment claimant must show ... that officials applied force maliciously”) (internal quotations omitted).

In a class action such as this, however, where the plaintiffs allege a pattern of exees- sive force, 197 there is no competing security or administrative interest that must be balanced under tight time constraints. .Unlike an officer faced with an immediate decision of whether to use force, and, if so, how much force to use, none of defendants’ actions at issue here were made under time pressure. Indeed, the conduct alleged herein is the very antithesis of action under pressure — it is the practice of permitting and condoning a pattern of excessive force by subordinates, as reflected by defendants’ management of the prison over an extended period of time. As Judge Lasker recognized in Fisher, “ ‘[t]he exigencies and competing obligations facing prison authorities while attempting to regain control of a riotous cellblock’ [as in Whitley ] do not exist where the conduct challenged is ‘the municipality’s operation of the Jail generally,’ because ‘unlike in the prison riot setting, there can be no legitimate concern that liability will improperly be based on decisions necessarily made in haste, under pressure, and frequently without the luxury of a second chance.’ ” Fisher, 692 F.Supp. at 1562 n. 56 (quoting Morgan v. District of Columbia, 824 F.2d 1049, 1057-58 (D.C.Cir.1987)). Although this observation was made in the context of a class claim regarding a pattern of inmate-on-inmate violence, it is equally apropos where the class alleges that the defendants’ operation of the prison permitted and condoned a pattern of staff using excessive force on inmates. Thus, the time constraints that faced the defendants in Hudson do not apply here.

Given the above, we are persuaded that Hudson does not require us to measure defendants’ state of mind against a standard of maliciousness. 198 Rather, we conclude that, in a case such as this, where class representatives are seeking to obtain injunctive relief against high ranking prison administrators for an ongoing pattern of excessive force, the subjective prong of the Eighth Amendment is satisfied by a showing of deliberate indifference.

This standard — which requires a showing of criminal recklessness — properly balances two important interests: the deference due to prison administrators charged with operating prisons, and the constitutional right to be protected from a pattern of excessive force. Cf. Redman v. County of San Diego, 942 F.2d 1435, 1443 (9th Cir.1991) (en banc) (applying deliberate indifference standard to inmate’s personal safety claim struck “appropriate balance” between right not to be punished and the deference given to prison offi cials to manage prisons), cert. denied, 502 U.S. 1074, 112 S.Ct. 972, 117 L.Ed.2d 137 (1992); Roland v. Johnson, 856 F.2d 764, 769 (6th Cir.1988) (same).

The Jordan ease is particularly instructive. There, the Ninth Circuit, sitting en banc, held that a prison’s practice of subjecting women inmates to cross-gender clothed body searches violated the Eighth Amendment. 986 F.2d at 1521. The court found that the practice, albeit generally falling within the realm of security, was not, in fact, justified by any legitimate security interest. Id. at 1527. The court then ruled that the deliberate indifference test governed the liability of the prison Warden because the practice had not been adopted under time constraints and inflicted pain upon inmates on a routine basis:

When, as here, officials formulate a policy in circumstances where there are no particular constraints on the officials’ decision-making process ... and the implementation of the policy will inflict pain upon the inmates on a routine basis, we need not look for a showing of action taken ‘maliciously and sadistically’ before Eighth Amendment protections are implicated.

Id. at 1528. Thus, Jordan teaches that where the prison practice at issue (a) lacks a legitimate security justification, (b) will inflict pain on a routine basis, and (c) was not developed under time constraints, plaintiffs need not show that prison policy makers acted with the very purpose of causing harm. Rather, a showing of deliberate indifference will suffice.

This is the situation presented here. First, although a practice of permitting and condoning a pattern of excessive force may fall within the universe of matters relating to security, prison administrators have no legitimate security interest in maintaining such a practice. Certainly, defendants have not offered one. Second, a practice of permitting and condoning a pattern of excessive force will inflict pain on a routine basis. Third, as discussed above, the actions and omissions at issue did not occur in a time-pressured context but rather over an extended period that allowed for ample reflection, calculation, and forethought.

In sum, where defendants, through their deliberate indifference, permit and condone a pattern of excessive force against inmates, they have demonstrated sufficient culpability to incur liability under the Eighth Amendment. To hold otherwise would allow prison administrators, through their own criminal recklessness, to permit and condone the routine use of excessive force against inmates. We do not believe that such a result comports with either the basic duties of a prison administrator or the intended protections of the Eighth Amendment.

As discussed in section 11(A)(3), supra, plaintiffs have amply established defendants’ deliberate indifference. Defendants knew that unnecessary and grossly excessive force was being employed against inmates on a frequent basis, and that this practice posed a substantial risk of serious harm to the plaintiff class. Nonetheless, defendants consciously .disregarded the risk of harm, choosing instead to tolerate and even encourage abuses of force by deliberately ignoring them when they occurred, tacitly accepting a code of silence, and, most importantly, failing to implement adequate systems to control and regulate the use of force, despite their knowledge that such systems are important to ensuring that the use of force is effectively controlled. See Fisher, 692 F.Supp. at 1564 (defendants’ failure to deter misuse of force by adequately investigating and disciplining use of force predictably led to misuse of force); Ruiz v. Estelle, 503 F.Supp. 1265, 1302 (S.D.Tex.1980) (prison officials encouraged staff to indulge in excessive physical violence by rarely investigating reports of violence and failing to take corrective disciplinary action against officer whom they knew to have brutalized inmates), aff'd in part, 679 F.2d 1115 (5th Cir.1982), amended in part, vacated in part on other grounds, 688 F.2d 266 (5th. Cir.1982), cert. denied, 460 U.S. 1042, 103 S.Ct. 1438, 75 L.Ed.2d 795 (1983).

Even assuming, arguendo, that the more stringent maliciousness test applies, the evidence demonstrates that defendants are culpable under this standard as well. As discussed in section 11(A)(3), supra, “the ex tent to which force is misused at Pelican Bay, combined with the flagrant and pervasive failures in defendants’ systems for controlling the use of force reveal more than just deliberate indifference: they reveal an affirmative management strategy to permit the use of excessive force for the purpose of punishment and deterrence.” Section II(A)(3), supra, at 1199. As such, we are convinced that defendants’ actions and omissions evince a “‘knowing willingness that [harm] occur’” rather than a good faith effort to maintain order. Farmer, — U.S. at —, 114 S.Ct. at 1978; Hudson, 503 U.S. at 7, 112 S.Ct. at 999.

Defendants emphasize that a change in certain policies and practices has reduced the level of violence at Pelican Bay. 199 They contend that such actions constitute a “reasonable response” to the risk of harm to inmates, thus precluding any liability. See Farmer, — U.S. at —, 114 S.Ct. at 1983. (Prison officials “who act reasonably cannot be found liable under the Cruel and Unusual Punishments Clause”). The Court considers these changes a positive development, and they may be relevant to shaping an appropriate remedy. Fisher, 692 F.Supp. at 1521, 1564. They do not, however, shield defendants from liability. As explained in our factual findings:

[S]uch changes, which post-date the filing of this class action, were likely motivated by this litigation, and at least as of the time of trial, had not been cemented in any formal written policy. As such, they may well be transitory in nature, and the Court is not persuaded that such changes would not be undone in the absence of court intervention. Nor has any prison official ever suggested that such changes were made to address problems concerning the use of excessive force. On the contrary, defendants never acknowledged that there was a genuine problem to be addressed and always offered other reasons to explain these changes in practice. Moreover, defendants never offered the Court any firm or clear assurances that such changes would be permanent. Accordingly, the Court is not convinced that these recent changes represent a serious commitment by defendants to end the pattern of excessive force.

Section 11(A)(3), supra, at 1198-99. Nor do the changes address all the facets of the pattern of excessive force. Given the above, we are not satisfied that the current changes in practice constitute a reasonable response sufficient to avert a finding of liability.

2. Objective Component

As noted earlier, the objective component of an Eighth Amendment claim focuses on whether the plaintiff has suffered “an infliction of pain” that is “objectively ‘harmful enough’ ” to establish a constitutional violation. Hudson, 503 U.S. at 8, 112 S.Ct. at 999-1000; see also Farmer, — U.S. at —, 114 S.Ct. at 1988; Wilson, 501 U.S. at 298, 111 S.Ct. at 2324. Here, plaintiffs have demonstrated that (a) the use of excessive force is sufficiently prevalent to demonstrate a pattern, and (b) that this pattern has resulted in injuries that are objectively “harmful enough” to implicate the Constitution.

a. Pattern of Excessive Force

The Court concludes that prisoners at Pelican Bay have been subjected to excessive force — including assaults, beatings, and naked cagings in inclement weather — on a scale of sufficient proportions to demonstrate a pattern rather than a collection of isolated incidents. In reaching this conclusion, we note that plaintiffs need not show that the prison is a place “where sadistic guards regularly torture inmates without cause” or a “dark and evil world completely alien to the free world.” Fisher, 692 F.Supp. at 1532, 1563 (internal quotations omitted); Withers v. Levine, 615 F.2d 158, 161 (4th Cir.), cert. denied, 449 U.S. 849, 101 S.Ct. 136, 66 L.Ed.2d 59 (1980) (plaintiff need not demonstrate a reign of violence and terror). On the other hand, plaintiffs must show more than a “mere collection of isolated and aberrant acts which are not characteristic of the institution[ ].” Ruiz, 503 F.Supp. at 1302.

As discussed in our findings, plaintiffs have amply shown that the misuse of force at Pelican Bay is not merely aberrational, but an inevitable and all too common consequence of defendants’ actions and omissions which tolerate and encourage the use of grossly excessive and unnecessary force. Fisher, 692 F.Supp. at 1532. See also Hoptowit, 682 F.2d at 1249-50; Ruiz, 503 F.Supp. at 1302 (finding that violence by prison officers was routine and not restricted to dangerous situations).

Defendants, however, contend that plaintiffs can not rely on any specific incident of excessive force to illustrate a pattern unless they also show that the officer involved applied the force maliciously and sadistically for the purpose of causing harm. In other words, they contend that the objective question of whether there is a pattern of excessive force that inflicts sufficiently serious injury requires an inquiry into the state of mind of the individual officers involved in the acts giving rise to the pattern.

This approach erroneously collapses the mental state necessary to hold a particular defendant liable for use of excessive force with the separate question whether the force used was objectively “excessive.” Force can be “excessive” — that is, unnecessary or grossly disproportionate to the circumstances — even when it is not inflicted with the malicious purpose of causing harm. For example, an officer could use an excessive amount of force because of lack of training and supervision rather than out of malice.

If class members were seeking relief from individual officers, clearly no individual instance of excessive force would be actionable unless the officer involved acted with the malicious purpose of causing harm. Whitley, 475 U.S. at 320, 106 S.Ct. at 1084 (the infliction of excessive force alone does not violate the Constitution). But here, the liability of individual officers is not at stake. Rather, class members are seeking only injunctive relief against top-ranking prison administrators; as such, the Court is not required to make findings of liability with respect to individual prison staff. Fisher, 692 F.Supp. at 1532 (in class action excessive force case court made no attempt to “reach final conclusions as to liability in any particular incident, as would be necessary in the case of an individual claim under 42 U.S.C. § 1983”). Instead, it is the defendants’ mental state that is properly at issue.

This approach recognizes that-the subjective component of the Eighth Amendment is directed toward assessing whether the defendant in any given case can be held liable for the excessive force at issue because she or he acted with a sufficiently culpable state pf mind. Wilson, 501 U.S. at 305, 111 S.Ct. at 2328 (“the criterion of liability [is] whether the respondents acted ‘maliciously and sadistically’ ” (emphasis added)). Thus, in a case charging prison administrators with a pattern of excessive force, the pattern is actionable only if plaintiffs can satisfy the subjective component of the Eighth Amendment by proving that the pattern is attributable to the defendants’ wanton state of mind. However, plaintiffs need not also prove that non -defendant prison officers are sufficiently culpable to incur individual Eighth Amendment liability. Indeed, such a rule would let widespread patterns of excessive force go unaddressed where prison administrators were culpable but individual officers were not. For example, a prison population could be subjected to a “pattern of excessive force,” not because individual prison staff act with a purpose of inflicting harm but because prison administrators, either acting with deliberate indifference or maliciousness, fail to ensure that staff are properly trained and supervised. 200

Even assuming, arguendo, however, that plaintiffs are required to establish a pattern of excessive force that consists solely of incidents where force was applied maliciously to cause harm, they have met that burden here. In reaching this conclusion, we need not rely on direct evidence of a malicious purpose; rather, as discussed above, we may draw inferences from circumstances surrounding the challenged conduct, including (1) the extent of the injury suffered, (2) the need for the application of force, (3) the relationship between that need and the amount of force used, (4) the threat reasonably perceived by the responsible officials, and (5) any efforts made to temper the severity of a forceful response. Hudson, 503 U.S. at 7, 112 S.Ct. at 999; see also Hill v. Shelander, 992 F.2d 714, 717-18 (7th Cir.1993) (evidence that officer assaulted inmate without provocation and that inmate did not resist would allow finder of fact to “infer that [the officer] acted with malicious intent.”); Farmer, — U.S. at —, 114 S.Ct. at 1981 (state of mind can be inferred from circumstantial evidence).

As detailed in section 11(A)(1), supra, the record is replete with instances in a variety of settings where these closely related factors show that excessive force was used, not for the good faith purpose of restoring or maintaining order, but maliciously for the purpose of causing harm, i.e., with a knowing willingness that harm occur. We have described numerous examples where the inmate suffered significant pain or injury, there was little or no need for the use of force at all or the amount of force used was far out of proportion to the need, and little or no effort was made to temper the severity of the response. See, e.g., Hudson v. McMillian, 962 F.2d 522, 523 (5th Cir.1992) (on remand) (upholding magistrate’s finding that force was used maliciously where amount of force used was unnecessary and excessive and could only be seen as motivated by malice). The indefensible treatment of inmates Richard and Castillo are but two of the many examples in the record. See section 11(A)(1)(a), supra.

The Court recognizes that correctional officers must react, sometimes quite forcefully, to subdue an uncooperative or combative inmate. Nothing herein is intended to detract from this basic proposition. However, it is not a license for correctional staff to immediately resort to the maximum, rather than the minimum, amount of force needed to restore order, and to do so with a knowing willingness that harm occur. See Slakan, 737 F.2d at 372 (“Even when a prisoner’s conduct warrants some form of response, evolving norms of decency require prison officials to use techniques and procedures that are both humane and restrained.”). At Pelican Bay, officers immediately resorted — with alarming regularity — to unnecessary and excessive force with the purpose of causing harm. Indeed, the degree of force used by correctional staff was often so far beyond any penological justification that it was clearly a pretext for inflicting punishment and pain.

b. Degree of Harm Inflicted

Finally, we must determine whether plaintiffs have demonstrated that the pattern of excessive force caused injuries that were objectively “harmful enough” to constitute an “infliction of pain.” Having reviewed the record, we are satisfied that plaintiffs have demonstrated injuries of a sufficiently harmful nature to violate the Constitution. In many of the incidents described in section 11(A), supra, inmates suffered serious physical injuries that often required follow-up medical care and left behind lingering or long-term effects. In cases where the force used may not have resulted in serious physical injury, such as instances involving use of fetal restraints or the nude cagings, the Court has found that defendants acted with a punitive purpose. Thus, the injury inflicted by those uses of force is also sufficiently harmful to satisfy the objective component of the Eighth Amendment. Hudson, 503 U.S. at 9, 112 S.Ct. at 1000 (“When prison officials malicious and sadistically use force to cause harm, contemporary standards of decency always are violated ... whether or not significant injury is evident”); cf. Gordon v. Faber, 800 F.Supp. at 800 (forcing inmates to go outside in sub-freezing weather without hats or gloves for over one hour constitutes infliction of pain).

3. Conclusion

While the Eighth Amendment will countenance prisons that are restrictive, and even harsh, it does not permit the pattern of needless and officially sanctioned brutality that has invaded operations at Pelican Bay. Not only have plaintiffs established a pattern of excessive force at Pelican Bay that has caused sufficient harm to demonstrate the “infliction of pain” on a elasswide basis, but they have also shown that this pattern is attributable, not to inadvertence or mistake, but to defendants’ deliberate indifference and knowing willingness that harm occur. It is a conclusion we do not reach lightly. On the contrary, it is with considerable reluctance and regret that we find violations of this nature within an institution of our state. We are persuaded, however, that the testimonial and documentary evidence permit no other result.

C. MEDICAL AND MENTAL HEALTH CARE

Like other conditions of confinement, medical care provided to inmates is subject to scrutiny under the Eighth Amendment’s prohibition against cruel and unusual punishment. Helling, — U.S. at —, 113 S.Ct. at 2480; Wilson, 501 U.S. at 297, 111 S.Ct. at 2323. This does not mean, however, that every inattention to every medical need implicates the Constitution. Nor does the Eighth Amendment guarantee inmates the best medical care available. Rather, to establish Eighth Amendment liability, plaintiffs must demonstrate that prison officials are “deliberately indifferent” to “serious” medical needs of inmates. Estelle, 429 U.S. at 106, 97 S.Ct. at 292; Toussaint IV, 801 F.2dat 1111.

It is firmly established that “medical needs” include not only physical health needs, but mental health needs as well. Hoptowit, 682 F.2d at 1253; Balla v. Idaho State Board of Corrections, 595 F.Supp. 1558, 1576-77 (D.Idaho 1984). As far back as 1977, the Fourth Circuit observed that there is “no underlying distinction between the right to medical care for physical ills and its psychological or psychiatric counterpart. Modern science has rejected the notion that mental or emotional disturbances are the products of afflicted souls, hence beyond the purview of counseling, medication and therapy.” Bowring v. Godwin, 551 F.2d 44, 47 (4th Cir.1977). Nor can it be questioned that deliberate indifference to a serious mental illness can precipitate as much, if not sometimes more, suffering and distress than indifference to a serious complaint of a solely physical nature. Thus the Ninth Circuit has held that “requirements for mental health care are the same as those for physical health care needs.” Doty v. County of Lassen 37 F.3d 540, 546 (9th Cir.1994); Cody v. Hillard, 599 F.Supp. 1025, 1058 (D.S.D.1984) (adequacy of mental health care system “is governed by the same constitutional standard which applies when determining the adequacy of a prison’s medical ... system”), aff'd in part and rev’d in part en banc, 830 F.2d 912 (8th Cir.1987), cert. denied, 485 U.S. 906, 108 S.Ct. 1078, 99 L.Ed.2d 237 (1988).

It is clear, and defendants do not dispute, that members of the plaintiff class have “serious” medical and mental health needs. 201 Rather, the central issue is whether defendants have been deliberately indifferent to those needs. In class actions challenging the entire system of mental or medical health care, courts have traditionally held that deliberate indifference can be shown by proving either a pattern of negligent acts or serious systemic deficiencies in the prison’s health care program:

‘[D]eliberate indifference’ can be evidenced by ‘repeated examples of negligent acts which disclose a pattern of conduct by the prison medical staff or it can be demonstrated by ‘proving there are such systemic and gross deficiencies in staffing, facilities, equipment, or procedures that the inmate population is effectively denied access to adequate medical care.’

Wellman, 715 F.2d at 272 (emphasis added) (citing Ramos, 639 F.2d at 575); Casey v. Lewis, 834 F.Supp. 1477, 1543 (D.Ariz.1993); see also Hoptowit v. Ray, 682 F.2d 1237, 1253 (9th Cir.1982) (“[M]edical services provided at the penitentiary are so deficient that they reflect a deliberate indifference to the serious medical needs of the prisoners”); Todaro, 565 F.2d at 52 (“systemic deficiencies in staffing, facilities or procedures [that] make unnecessary suffering inevitable” are evidence of “deliberate indifference”).

As discussed above, however, the Supreme Court recently made it clear that the “deliberate indifference” standard requires a showing of culpability that can not be inferred solely from objective conditions. Farmer, — U.S. at — - —, 114 S.Ct. at 1979-80. Rather, it can only be found where the defendant actually knows of, and yet disregards, an excessive risk to inmate health or safety. Id. at —, 114 S.Ct. at 1979. Accordingly, to prove deliberate indifference, plaintiffs must demonstrate not only that the levels of medical and mental health care are constitutionally inadequate from an objective standpoint — based on either a “pattern of negligent conduct” or “systemic deficiencies” — but also that defendants (1) knew the risk to inmate health that this inadequacy posed, and (2) acted with disregard for this risk. In short, plaintiffs must show that defendants “ ‘consciously disregarded]’ a substantial risk of serious harm” to plaintiffs’ health or safety. Id. at —, 114 S.Ct. at 1980. Accidental or inadvertent failure to provide adequate care will not suffice. Ramos, 639 F.2d at 575.

Plaintiffs have met this burden here, with respect to both the treatment of physical health needs and of mental health needs. As discussed below, they have shown that appalling systemic deficiencies render the mental health care system and the medical care system incapable of satisfying minimum constitutional standards. They have also shown that defendants have consciously disregarded the substantial risk of harm posed by these deficiencies. We therefore conclude that defendants have been deliberately indifferent to the serious mental and medical health needs of the population at Pelican Bay.

1. Systemic Deficiencies in Medical and Mental Health Care Systems.

The Eighth Amendment does not require that prison officials provide the most desirable medical and mental health care; nor should judges simply “constitutionalize” the standards set forth by professional associations such as the American Medical Association or the American Public Health Association. Hoptowit, 682 F.2d at 1253; see also Bell v. Wolfish, 441 U.S. 520, 543-544 n. 27, 99 S.Ct. 1861, 1876 n. 27, 60 L.Ed.2d 447 (1979) (draft recommendations of the Department of Justice are not determinative of constitutional requirements). However, the Eighth Amendment does require that defendants “provide a system of ready access to adequate medical care.” Hoptowit, 682 F.2d at 1253; Casey, 834 F.Supp. at 1545.

Courts have considered a number of factors which bear upon whether or not a system meets constitutional minima. First, prisoners must be able “to make their medical problems known to the medical staff.” Hoptowit, 682 F.2d at 1253; Casey 834 F.Supp. at 1545. While a functioning sick call system can be effective for physical illnesses, there must be a “systematic program for screening and evaluating inmates in order to identify those who require mental health treatment.” Balla, 595 F.Supp. at 1577 (quoting Ruiz, 503 F.Supp. at 1339). This is particularly so since “[s]everely mentally ill inmates cannot make their needs known to mental health staff’ on their own. Casey, 834 F.Supp. at 1550.

Next, the facility must be sufficiently staffed. Ramos, 639 F.2d at 578 (staffing shortfalls effectively deny inmates access to diagnosis and treatment); Lightfoot v. Walker, 486 F.Supp. 504, 524-25 (S.D.Ill.1980) (finding that staff shortages render medical services below constitutional level); French v. Owens, 777 F.2d 1250, 1254 (7th Cir.1985), cert. denied, 479 U.S. 817, 107 S.Ct. 77, 93 L.Ed.2d 32 (1986) (gross deficiencies in staffing may constitute deliberate indifference). Mental health professionals must be employed in “sufficient numbers to identify and treat in an individualized manner those treatable inmates suffering from serious mental disorders.” Balla, 595 F.Supp. at 1577 (quoting Ruiz, 503 F.Supp. at 1339); see also Cabrales v. County of Los Angeles, 864 F.2d 1454, 1461 (9th Cir.1988) (understaffing contributed to inmate’s suicide because of a lack of diagnosis and treatment).

The prison may refer inmates to outside facilities for treatment; however, if defendants choose to refer inmates outside the prison, they must provide “reasonably speedy access” to these other facilities. Hoptowit, 682 F.2d at 1253. See also Lightfoot, 486 F.Supp. at 522 (finding unacceptable delays in transfer of residents in need of psychiatric care). For those inmates who are treated within the prison, access to medical treatment cannot be substantially delayed in a systematic manner. Although isolated instances of delay do not give rise to liability unless they have caused substantial harm to the inmate, Wood v. Housewright, 900 F.2d 1332, 1335 (9th Cir.1990), regular and significant delays in the delivery of medical care may be constitutionally unacceptable. Casey, 834 F.Supp. at 1545; see also Hoptowit, 682 F.2d at 1253 (must provide “ready” access); Todaro v. Ward, 431 F.Supp. 1129, 1146 (S.D.N.Y.1977) (substantially delayed access to treatment violates the constitution); Balla, 595 F.Supp. at 1567 (finding significant delays before seeing a doctor are part of violation).

Moreover, “the prison must provide an adequate system for responding to emergencies.” Hoptowit 682 F.2d at 1253. Security staff (or lack thereof) should not dangerously delay access to emergency treatment. See Casey, 834 F.Supp. at 1502, 1545 (delay in access to treatment outside prison caused by lack of security and transportation staff is part of violation); Ramos, 639 F.2d at 577. If outside facilities are too distant to handle emergencies promptly, then the prison must provide “adequate facilities and staff to handle emergencies within the prison.” Hoptowit, 682 F.2d at 1253. Staff must be adequately trained to cope with emergencies. Palmigiano v. Garrahy, 443 F.Supp. 956, 974 (D.R.I.1977) (nursing staff “not provided with guidance for use in commonly occurring emergencies” found inadequate); Balla, 595 F.Supp. at 1567 (lack of written procedures for emergencies an element of violation).

Health screenings are a necessary supplement to ordinary avenues of access to medical care. The facility should screen newly arrived inmates to identify potential medical problems and communicable diseases. Lightfoot, 486 F.Supp. at 524 (“Health care admission screening procedures, including a physical examination performed by a physician, are an essential element of a constitutionally adequate system”); Tillery v. Owens, 719 F.Supp. 1256, 1306 (W.D.Pa.1989), aff'd, 907 F.2d 418 (3rd Cir.1990) (three-minute intake physical performed by physician who does not touch inmates inadequate); Hoptowit, 682 F.2d at 1253 (failure to provide preventative health care or routine physical examinations part of violation). Screenings for communicable diseases should be sufficient to protect other inmates from infectious diseases. See Lareau v. Manson, 651 F.2d 96, 109 (2d Cir.1981) (failure to screen for communicable diseases poses a serious threat to well-being of other inmates and is sufficient to give rise to Eighth Amendment violation).

The requirement of ready access to adequate care precludes prison officials from preventing treatment which is medically nee- essary in the judgment of the treating doctor. Estelle, 429 U.S. at 104-105, 97 S.Ct. at 291 (intentional interference with prescribed treatment manifests deliberate indifference); Casey, 834 F.Supp. at 1545.

Of course, “[a]ceess to the medical staff has no meaning if the medical staff is not competent to deal with the prisoners’ problems.” Hoptowit, 682 F.2d at 1253; Cabrales, 864 F.2d at 1461. While medical technical assistants or their equivalent may permissibly be the first to examine inmates with physical ailments, they must be properly trained to perform this function and adequately supervised. Capps v. Atiyeh, 559 F.Supp. 894, 912 (D.Or.1982); see also Toussaint IV, 801 F.2d at 1112 (unqualified personnel may not regularly engage in medical practice). Medical technicians cannot be “left to operate in a vacuum” without physician supervision or guidance from written protocols. Capps, 559 F.Supp. at 912; Ramos, 639 F.2d at 576 (constitutional violation where inadequately supervised medical providers misdiagnose or mistreat inmates, causing life-threatening situations and needless pain and suffering); Palmigiano, 443 F.Supp. at 974; Newman v. Alabama, 349 F.Supp. 278, 284 (M.D.Ala.1972); Lightfoot, 486 F.Supp. at 517. Moreover, staff should receive “in-service” training or continuing education to ensure that they are adequately trained. See Capps, 559 F.Supp. at 912 (lack of continuing education for staff part of violation); Palmigiano, 443 F.Supp. at 974 (lack of in-service training programs part of violation).

Certain procedures are also all but indispensable to providing adequate care. First, “[a] primary component of a minimally acceptable correctional health care system is the implementation of procedures to review the quality of medical care being provided.” Lightfoot, 486 F.Supp. at 517-18. Reviews of records to evaluate the delivery of care are essential. Capps, 559 F.Supp. at 912 (lack of chart review is part of violation); Lightfoot, 486 F.Supp. at 517 (lack of chart review is element of violation); Todaro, 431 F.Supp. at 1160 (failure to audit system part of violation); see also Palmigiano, 443 F.Supp. at 975. In addition, peer review and death reviews should be instituted to improve the quality of care. Capps, 559 F.Supp. at 912 (lack of peer review part of violation); Lightfoot 486 F.Supp. at 517-18 (noting lack of peer review and expressing court’s “alarm[ ]” at the “lack of regular system of review of deaths”).

Medical records must be sufficiently organized and thorough to allow the provision of adequate care to inmates. Hoptowit, 682 F.2d at 1252-53 (deficient medical records part of violation); Capps, 559 F.Supp. at 912; Casey, 834 F.Supp. at 1503 (medical record-keeping system deficient where “medical records are not always available at sick call treatment ... [and] do not always have the appropriate or required documentation of treatment or assessment of medical problems”). Medical records that are “'inadequate, inaccurate and unprofessionally maintained’ ” constitute a “ ‘grave risk of unnecessary pain and suffering’ in violation of the eighth amendment.” Cody, 599 F.Supp. at 1057 (quoting Burks v. Teasdale, 492 F.Supp. 650, 676, 678 (W.D.Mo.1980)).

Finally, some constitutional minima are specific to mental health care. Psychotropic or behavior-altering medication should only be administered with appropriate supervision and periodic evaluation. Ruiz, 503 F.Supp. at 1339. In addition, there should be a basic program to identify, treat, and supervise inmates with suicidal tendencies, and mental health records should be adequately maintained. Id.; see also Balla, 595 F.Supp. at 1577.

The Court finds the delivery of both physical and mental health care at Pelican Bay to be constitutionally inadequate. The system of physical health care at Pelican Bay fails to provide “ready access to adequate medical care.” Hoptowit, 682 F.2d at 1253. As discussed at length in the findings of fact, staffing levels, although improved after an abysmal start, are still insufficient. Training and supervision of medical staff, particularly during the critical triage process when MTAs decide if inmates may see a physician, is almost nonexistent. Intake health screening is woefully inadequate and screening for communicable diseases has been poorly im plemented. Inmates often experience significant delays in receiving treatment. There are no protocols or training programs for dealing with emergencies or trauma, and the facility has no effective procedures for managing inmates’ chronic illnesses. The medical recordkeeping system is utterly deficient. Finally, there are no programs of substance in place to ensure that quality care is provided: Pelican Bay has no working quality control program, no genuine peer review, no death reviews. In combination, these systemic deficiencies in the provision of medical services make ready access to adequate medical care impossible at Pelican Bay.

Similarly, the mental health care system at Pelican Bay fails to provide “ready access to adequate [mental health] care.” Hoptowit, 682 F.2d at 1253. As detailed in the findings of fact, staffing levels, once outrageously low, are still seriously deficient. Screening and referral mechanisms are inadequate. Inmates with serious mental health needs are not receiving adequate monitoring and treatment on far more than just isolated occasions, particularly in the SHU. Some acutely psychotic inmates are left to suffer, in a hallucinatory and distraught state, without being referred to needed inpatient or intensive outpatient treatment. Inmates that are referred to other institutions for inpatient or intensive outpatient care often experience significant delays — delays that become particularly troublesome given Pelican Bay’s lack of procedures for involuntary administration of antipsychotic drugs. Certain mentally ill inmates may require temporary or permanent exclusion from the SHU in order to attain and sustain a psychiatric recovery; others may require temporary or permanent exclusion from the SHU to prevent a mental deterioration from progressing into a serious mental disorder. The professional mental health staff, however, is precluded from addressing such needs.

Defendants suggest that the mental health care provided at Pelican Bay should pass constitutional muster because it is not completely “systemless.” Defendants also argue that they have a “system in place” to provide medical care. Clearly, a prison with “sys-temless” health care would not withstand Eighth Amendment scrutiny; this is not, however, the dispositive inquiry. Indeed, whether or not a prison has some medical or mental health care “system,” unless it actually delivers ready access to adequate care it can not survive constitutional scrutiny.

In this case, the deficiencies discussed above and in the findings of fact show that the delivery of medical and mental health care at Pelican Bay is riddled with systemic and gross deficiencies — deficiencies which preclude ready access to adequate care. We therefore conclude, as we must, that defendants’ system for providing mental health care and medical care fails to comport with minimum constitutional standards.

2. Defendants’ State of Mind

Determination of the defendants’ state of mind presents a question of fact. Farmer, — U.S. at —, 114 S.Ct. at 1981. As set forth in the findings of fact above, we conclude that defendants knew they were subjecting the inmate population to a substantial risk of serious harm by virtue of their utter failure to provide for adequate medical and mental health care. This finding is based on information of which defendants were aware, coupled with the fact that the need for medical and substantial psychiatric services at Pelican Bay, and the risks of failing to address this need, were patently obvious to defendants. Id. at —, 114 S.Ct. at 1981; Hoptowit, 682 F.2d at 1253 (“medical services provided at the penitentiary are so deficient that they reflect a deliberate indifference to the serious medical needs of the prisoners”); Ramos, 639 F.2d at 578 (staff shortages make “unnecessary suffering inevitable” and evince deliberate indifference).

Defendants’ callous and deliberate indifference to inmates’ needs is particularly evinced by their failure to institute any substantive quality control. Quality control procedures represent the first critical steps of self-evaluation that could help defendants remedy widespread deficiencies; yet, at the time of trial, there were still no such procedures in operation.

Defendant Gomez’ actions, or lack thereof, after Pelican Bay became operational reveal a continuing, conscious disregard for the ongoing risk of harm to the mental health of inmates at Pelican Bay. Although some improvements have been made in recent years, they have not satisfactorily addressed the glaring deficiencies; further, we conclude that some of those improvements were prompted primarily by litigation. We find that the remaining deficiencies will not be cured absent some supervision by this Court.

In sum, plaintiffs have amply proven that the prison population at Pelican Bay has serious medical and mental health needs to which defendants have been deliberately indifferent. As was long ago established in Estelle, this unnecessary and wanton infliction of pain is inconsistent with contemporary standards of decency and violates the Eighth Amendment of the Constitution. Estelle, 429 U.S. at 103-104, 97 S.Ct. at 290-91.

There is no static test that determines whether conditions of confinement constitute cruel and unusual punishment. Davenport v. DeRobertis, 844 F.2d 1310, 1314-15 (7th Cir.), cert. denied, 488 U.S. 908, 109 S.Ct. 260, 102 L.Ed.2d 248 (1988). Rather, courts must assess whether the conditions are such that they are compatible with “civilized standards, humanity and decency.” Estelle, 429 U.S. at 102, 97 S.Ct. at 290 (internal quotation omitted); Young v. Quinlan, 960 F.2d 351, 359 (3rd Cir.1992). These civilized standards are measured, not by reference to any fixed historical point, but by “the evolving standards of decency that mark the progress of a maturing society.” Rhodes, 452 U.S. at 346, 101 S.Ct. at 2399. Of course, not every deficiency or inadequacy rises to the level of an Eighth Amendment violation. Young, 960 F.2d at 359. However, conditions that are “inhumane,” Farmer, — U.S. at —, 114 S.Ct. at 1976, deprive inmates of “basic human needs,” Helling, — U.S. at — - —, 113 S.Ct. at 2480-81, or fail to furnish a “minimal civilized measure of life’s necessities,” Wilson, 501 U.S. at 298, 111 S.Ct. at 2324; Chandler v. Baird, 926 F.2d 1057, 1064 (11th Cir.1991), are constitutionally wanting under contemporary Eighth Amendment standards. Young, 960 F.2d at 363-64.

At a minimum, these life necessities include adequate food, clothing, shelter, medical care and personal safety. Farmer, — U.S. at. —, 114 S.Ct. at 1976; Young, 960 F.2d at 364 (including sanitation). However, no simplistic litany of conditions should preclude the “fact-intensive inquiry” required by Eighth Amendment standards. Chandler, 926 F.2d at 1064; Toussaint v. McCarthy (Toussaint III), 597 F.Supp. 1388, 1393 (N.D.Cal.1984), aff'd in part and rev’d in part, 801 F.2d 1080 (9th Cir.1986), cert. denied, 481 U.S. 1069, 107 S.Ct. 2462, 95 L.Ed.2d 871 (1981). Thus, although courts have often focused on the minimum needed to physically sustain life, such as shelter, food, and medical care, courts have also recognized that conditions that inflict serious mental pain or injury also implicate the Eighth Amendment. As the Third Circuit recently observed, “[t]he touchstone is the health of the inmate. While the prison administration may punish, it may not do so in a manner that threatens the physical and mental health of prisoners.” Young, 960 F.2d at 364 (emphasis added). 202

We thus can not ignore, in judging challenged conditions of confinement, that all humans are composed of more than flesh and bone — even those who, because of unlawful and deviant behavior, must be locked away not only from their fellow citizens, but from other inmates as well. Mental health, just as much as physical health, is a mainstay of life. Indeed, it is beyond any serious dispute that mental health is a need as essential to a meaningful human existence as other basic physical demands our bodies may make for shelter, warmth or sanitation. As the Supreme Court has made quite clear, we can not, consistent with contemporary notions of humanity and decency, forcibly incarcerate prisoners under conditions that will, or very likely will, make them seriously physically ill. Helling, — U.S. —, 113 S.Ct. 2475. Surely, these same standards will not tolerate conditions that are likely to make inmates seriously mentally ill.

In this case, plaintiffs do not claim that SHU conditions deprive inmates of adequate food, heat, clothing, or sanitary conditions. Rather, plaintiffs allege that the conditions in the SHU, while sufficient to satisfy basic physical needs, pose a grave threat to the mental health of inmates. Specifically, plaintiffs contend that the conditions of extreme social isolation and reduced environmental stimulation in the SHU inflict psychological trauma, and in some cases deprive inmates of sanity itself. As such, they urge the Court to find that the SHU, as currently operated, deprives inmates of one of the “basic necessities of human existence.” Young, 960 F.2d at 364. They further contend that defendants have been deliberately indifferent to the mental health risks posed by conditions in the SHU. 203 Defendants, on the other hand, contend that plaintiffs have failed to establish any link between the conditions in the SHU and mental illness and that, in any event, the conditions in the SHU comport with contemporary Eighth Amendment standards. This claim has generated considerable attention, not only because it raises issues of a very serious dimension but because the Pelican Bay SHU is considered a state-of-the-art, “modern day” SHU, and thus a potential forerunner for other similar units around the country.

Having given the matter careful deliberation, we conclude that the record and the law do not fully sustain the position advocated by either plaintiffs or defendants. As explained below, we are not persuaded that the SHU, as currently operated, violates Eighth Amendment standards vis-a-vis all inmates. We do find, however, that conditions in the SHU violate such standards when imposed on certain subgroups of the inmate population, and that defendants have been deliberately indifferent to the serious risks posed by subjecting such inmates to the SHU over extended periods of time.

1. Whether conditions in the SHU are sufficiently injurious to mental health so as to deprive inmates of a basic necessity of life

We begin our analysis by underscoring that the general concept of segregating inmates for disciplinary or security reasons is a well established and penologically justified practice. Indeed, segregation “may be a necessary tool of prison discipline, both to punish infractions and to control and perhaps protect inmates whose presence within the general population would create unmanageable risks.” Young, 960 F.2d at 364. Thus, there is nothing per se improper about segregating inmates, even for lengthy or indefinite terms. Toussaint v. Yockey, 722 F.2d 1490, 1494 n. 6 (9th Cir.1984). There is also little question that prison gang activity and violence within California prisons are serious problems that require strong measures from prison administrators. The decision to segregate inmates who threaten the security of the general population falls well within defendants’ far ranging discretion to manage California’s prison population.

Defendants’ discretion to determine the specific conditions of segregation is similarly broad. Given the “limitations of federalism and the narrowness of the Eighth Amendment” it is not the Court’s function to pass judgment on the policy choices of prison officials. Hoptowit, 682 F.2d at 1246. Rather, prison administration is a matter “peculiarly within the province of the legislative and executive branches of government.” Turner v. Safley, 482 U.S. 78, 84-85, 107 S.Ct. 2254, 2259, 96 L.Ed.2d 64 (1987). Defendants are thus entitled to design and operate the SHU consistent with the penal philosophy of their choosing, absent constitutional violations. Peterkin v. Jeffes, 855 F.2d 1021, 1033 (3rd Cir.1988). They may impose conditions that are “‘restrictive and even harsh’ ” Farmer, — U.S. at —, 114 S.Ct. at 1977 (quoting Rhodes, 452 U.S. at 347, 101 S.Ct. at 2399); they may emphasize idleness, deterrence, and deprivation over rehabilitation. This is not a matter for judicial review or concern unless the evidence demonstrates that conditions are so extreme as to violate basic concepts of humanity and deprive inmates of a minimal level of life’s basic necessities. Young, 960 F.2d at 364 (“Segregated detention is not cruel and unusual punishment per se, as long as the conditions of confinement are not foul, inhuman or totally without penological justification”); Toussaint III, 597 F.Supp. at 1413-14 (although doubting wisdom of certain policies, the court observed that its function is “not to sit in judgment of the policy choices of state officials”). In short, absent a showing of constitutional infringement, courts may not substitute their judgment or otherwise interfere with decisions made by prison officials. Peterkin, 855 F.2d at 1033; Hoptowit, 682 F.2d at 1246.

It is equally clear that the very nature of prison confinement may have a deleterious impact on the mental state of prisoners, for reasons that are self-evident. Especially for those facing long sentences, “depression, hopelessness, frustration, and other such psychological states may well prove to be inevitable byproducts.” Jackson v. Meachum, 699 F.2d 578, 584 (1st Cir.1983); Davenport, 844 F.2d at 1313 (it is “highly probable that the experience of being imprisoned inflicts psychological damage”). This is particularly true for inmates placed in segregation, given that they are, by definition, subjected to additional isolation beyond that experienced by other general population inmates. Such inmates, for example, are often excluded from participating in prison work and vocational programs, leaving them to endure a regimen of prolonged and forced idleness. The resulting extreme boredom may cause prisoners to suffer loneliness and “psychological pain.” Toussaint III, 597 F.Supp. at 1414. Nonetheless, there is no right to recreational, vocational or rehabilitative programs. “The lack of these programs simply does not amount to the infliction of pain.” Toussaint IV, 801 F.2d at 1106 (internal quotation omitted).

Thus, as the Toussaint ease highlights, the “psychological pain” that results from idleness in segregation is not sufficient to implicate the Eighth Amendment, particularly where the exclusion from prison programs is not without some penological justification. Toussaint TV, 801 F.2d at 1108. As the district court observed in Toussaint III, “[a]t least in theory, each [inmate in segregated housing] has been selected for segregation on the basis of criteria indicating that he is in some way unfit or unsuited for intermingling with other inmates, whether because he has misbehaved, because he presents a threat to the safety of other inmates, or because he has requested isolation from other inmates for his own protection.” Toussaint III, 597 F.Supp. at 1414.

Indeed, the import of Toussaint is that the mental impact of a challenged condition should be considered in conjunction with penological considerations. Toussaint IV, 801 F.2d at 1108. On the one hand, a condition that is sufficiently harmful to inmates or otherwise reprehensible to civilized society will at some point yield to constitutional constraints, even if the condition has some penological justification. Thus, defendants’ insistence that the SHU is “working” as a secure environment for disruptive pris oners 204 does not and cannot determine whether the SHU passes constitutional muster. No prison, for example, can deprive inmates of a basic human need, even though the underlying conditions might otherwise arguably promote some penological objective. “A punishment may be so below civilized norms as to be cruel and unusual no matter what its provocation....” O’Brien v. Moriarty, 489 F.2d 941, 944 (1st Cir.1974). Sedating all inmates with a powerful medication that leaves them in a continual stupor would arguably reduce security risks; however, such a condition of confinement would clearly fail constitutional muster.

On the other hand, a condition or other prison measure that has little or no penological value may offend constitutional values upon a lower showing of injury or harm. See Gordon, 800 F.Supp. at 800 (“The lack of legitimate penological interest is relevant to the determination of whether the objective [Eighth Amendment] standard has been violated”); 205 see also Sheley v. Dugger, 833 F.2d 1420, 1428 (11th Cir.1987) (Eighth Amendment “forbids inflictions of pain which are totally without penological justification”) (internal quotation omitted).

In this case, the conditions at issue primarily affect three inmate populations: (1) those who are being disciplined for committing serious rules violations, (2) those who the CDC has determined are affiliated with a prison gang, and (3) those who are otherwise considered security risks because of disruptive or assaultive behavior. The severe restrictions on social interaction further defendants’ legitimate interest in precluding opportunities for disruptive or gang related activity and assaults on other inmates or staff. 206 For those serving short-term disciplinary terms, they also serve a punitive function. Other aspects of the conditions in the SHU, however, appear tenuously related to legitimate penological interests, at least with respect to those inmates that are segregated in the SHU not as a disciplinary measure, but for other reasons. For example, it is not clear how the lack of an outside view, the extreme sterility of the environment, and the refusal to provide any recreational equipment in the exercise pen (even a handball) furthers any interest other than punishment, and defendants have not advanced one. Thus, in the Court’s view, the totality of the SHU conditions may be harsher than necessary to accommodate the needs of the institution with respect to these populations. However, giving defendants the wide-ranging deference they are owed in these matters, we can not say that the conditions overall lack any penological justification.

Accordingly, as was the case in Toussaint, plaintiffs can not prevail on the instant claim simply by pointing to the generalized “psychological pain” — i.e. the loneliness, frustration, depression or extreme boredom — that inmates may experience by virtue of their confinement in the SHU. Toussaint IV, 801 F.2d at 1107-08; see also Jackson, 699 F.2d at 581 (where social isolation of segregated inmate “caused him to become depressed,” district court could not order that daily inter action with other inmates be provided). Cf. Doty, 37 F.3d at 546 (“mild stress-related ailments are the type of ‘routine discomfort’ that may result merely from incarceration” and do not constitute serious mental health need). The Eighth Amendment simply does not guarantee that inmates will not suffer some psychological effects from incarceration or segregation. Jackson, 699 F.2d at 583. However, if the particular conditions of segregation being challenged are such that they inflict a serious mental illness, greatly exacerbate mental illness, or deprive inmates of their sanity, then defendants have deprived inmates of a basic necessity of human existence — indeed, they have crossed into the realm of psychological torture.

Courts have recognized that conditions in segregation could cross this line, particularly, where the length of segregation is indefinite or long term. For example, in Jackson, the Court observed that “although depression, hopelessness, frustration, and other psychological states may well prove to be inevitable byproducts of lifelong incarceration, the threat of substantial, serious and possibly irreversible if not critical psychological illness together with prolonged or indefinite segregated confinement would increase the burden on prison authorities to explore feasible alternative custodial arrangements.” 699 F.2d at 584-5; see also O’Brien, 489 F.2d at 944 (segregated confinement involving “neither intolerable isolation nor inadequate food, heat, sanitation, lighting or bedding” is not cruel and unusual) (emphasis added); Grubbs v. Bradley, 552 F.Supp. 1052, 1124 (M.D.Tenn.1982) (“[T]he mere fact that inmates may tend to degenerate as a result of incarceration is not actionable. On the other hand, if conditions are so bad that serious physical or psychological deterioration is'inevitable, then the result is cruel and unusual punishment”) (emphasis added). 207

In short, while courts will reject Eighth Amendment claims where there is no persuasive evidence that the challenged conditions lead to serious mental injury, 208 where such injury can in fact be shown, Eighth Amendment protections clearly come into play. Thus, we must ask the following question; does the evidence before the Court demonstrate that the conditions in the Pelican Bay SHU inflict mental harm so serious or severe that they cross the constitutional line?

Clearly, the constellation of conditions at issue here go well beyond the simple absence of prison programs which formed the basis of the challenge in Toussaint. What plaintiffs object to is not merely an absence of programs, but a more universal deprivation of human contact and stimulation. As described more fully in the factual findings, this deprivation is achieved through various factors, including the stark physical environment, the lack of any window to the outside world, the geographically remote location of Pelican Bay, and the extreme degree of social isolation stemming from the tightly restricted contact with prison staff, inmates and others. As we have already noted, some of these conditions appear, at best, tenuously related to legitimate concerns. We must also consider that many in the SHU face indefinite and potentially lengthy terms. 209

As plaintiffs also point out, they need not show that every inmate will suffer a serious mental illness or injury that is attributable to conditions in the SHU. In Helling, — U.S. —, 113 S.Ct. 2475, the plaintiff challenged a condition of confinement — his compelled exposure to second hand smoke — on the ground that it posed an unreasonable risk to his health. As the Supreme Court made plain, the plaintiff need not prove that every inmate would become ill from the second hand smoke. Rather, it indicated that the critical inquiry was whether (1) the risk involved was “unreasonable” in that the challenged conditions were “sure,” “very likely” or “imminently]” likely to cause “serious” damage to the inmate’s future health, and (2) whether society considers the risk to b.e “so grave that it violates contemporary standards of decency to expose anyone unwillingly to such a risk.” Id. at —, 113 S.Ct. at 2482. “In other words,” the Court continued, “the prisoner must show that the risk of which he complains is not one that today’s society chooses to tolerate.” Id. 210

Here, the record demonstrates that the conditions of extreme social isolation and reduced environmental stimulation found in the Pelican Bay SHU will likely inflict some degree of psychological trauma upon most inmates confined there for more than brief periods. Clearly, this impact is not to be trivialized; however, for many inmates, it does not appear that the degree of mental injury suffered significantly exceeds the kind of generalized psychological pain that courts have found compatible with Eighth Amendment standards. While a risk of a more serious injury is not non-existent, we are not persuaded, on the present record and given all the circumstances, that the risk of developing an injury to mental health of sufficiently serious magnitude due to current conditions in the SHU is high enough for the SHU population as a whole, to find that current conditions in the SHU are per se violative of the Eighth Amendment with respect to all potential inmates.

We can not, however, say the same for certain categories of inmates: those who the record demonstrates are at a particularly high risk for suffering very serious or severe injury to their mental health, including overt paranoia, psychotic breaks with reality, or massive exacerbations of existing mental illness as a result of the conditions in the SHU. Such inmates consist of the already mentally ill, as well as persons with borderline personality disorders, brain damage or mental retardation, impulse-ridden personalities, or a history of prior psychiatric problems or chronic depression. For these inmates, plac- ' ing them in the SHU is the mental equivalent of putting an asthmatic in a place with little air to breathe. The risk is high enough, and the consequences serious enough, that we have no hesitancy in finding that the risk is plainly “unreasonable.” Helling, — U.S. at —, 113 S.Ct. at 2481. Such inmates are not required to endure the horrific suffering of a serious mental illness or major exacerbation of an existing mental illness before obtaining relief. Id. at — - —, 113 S.Ct. at 2480-81. 211

We are acutely aware that defendants are entitled to substantial deference with respect to their management of the SHU. However, subjecting individuals to conditions that are “very likely” to render them psychotic or otherwise inflict a serious mental illness or seriously exacerbate an existing mental illness can not be squared with evolving standards of humanity or decency, especially when certain aspects of those conditions appear to bear little relation to security concerns. A risk this grave — this shocking and indecent — simply has no place in civilized society. It is surely not one “today’s society [would] choose[] to tolerate.” Id. at —, 113 S.Ct. at 2482. Indeed, it is inconceivable that any representative portion of our society would put its imprimatur on a plan to subject the mentally ill and other inmates described above to the SHU, knowing that severe psychological consequences will most probably befall those inmates. Thus, with respect to this limited population of the inmate class, plaintiffs have established that continued confinement in the SHU, as it is currently constituted, deprives inmates of a minimal civilized level of one of life’s necessities.

The above conclusions do not end our inquiry. In addition to demonstrating an injury that is “sufficiently serious” to violate objective Eighth Amendment standards, plaintiffs must also satisfy the subjective component of the Eighth Amendment. Specifically, they must show that the alleged injury is attributable to defendants’ “wanton” state of mind, and can therefore be fairly characterized as a form of cruel and unusual punishment. Wilson, 501 U.S. at 299, 111 S.Ct. at 2324.

Where, as here, the plaintiff contends that inhumane conditions are depriving inmates of their mental health, wantonness is established by proving that defendants have been deliberately indifferent to the risk of harm. Farmer, — U.S. at —, 114 S.Ct. at 1977 (“In prison-conditions cases [the relevant] state of mind is one of ‘deliberate indifference’ to inmate health or safety....”); Helling, — U.S. at —, 113 S.Ct. at 2480 (“Whether one characterizes the treatment received by [the prisoner] as inhuman conditions of confinement, failure to attend to his medical needs, or a combination of both, it is appropriate to apply the deliberate indifference standard articulated in Estelle ”) (internal quotations omitted). 212

Based on the Court’s findings of fact, and the evidence presented, we conclude that defendants had actual subjective knowledge that the conditions in the SHU present ed a substantial or excessive risk of harm with respect to inmates who were mentally ill or otherwise particularly vulnerable to conditions of extreme isolation and reduced environmental stimulation. 213 Yet defendants, in continued disregard for this risk, took no action to either exclude such inmates from the SHU, ameliorate the offending conditions with respect to these inmates, or otherwise seriously address the issue. This constitutes deliberate indifference. Farmer, — U.S. at —, 114 S.Ct. at 1979. As found above, defendants have also been deliberately indifferent to the lack of adequate mental health care provided to inmates at Pelican Bay, particularly in the SHU. See section II(A)(C)(3), supra. This merely underscores defendants’ callous lack of concern for the mental health of those inmates that are particularly at risk in the SHU.

In sum, while the conditions in the SHU may press the outer bounds of what most humans can psychologically tolerate, the record does not satisfactorily demonstrate that there is a sufficiently high risk to all inmates of incurring a serious mental illness from exposure to conditions in the SHU to find that the conditions constitute a per se deprivation of a basic necessity of life. We emphasize, of course, that this determination is based on the current record and data before us. We can not begin to speculate on the impact that Pelican Bay SHU conditions may have on inmates confined in the SHU for periods of 10 or 20 years or more; the inmates studied in connection with this action had generally been confined to the SHU for three years or less. We do, however, find, for the reasons stated above, that continued confinement in the SHU, under present conditions, constitutes cruel and unusual punishment in violation of the Eighth Amendment for two categories of inmates: those who are already mentally ill and those who, as identified above, are at an unreasonably high risk of suffering serious mental illness as a result of present conditions in the SHU. Defendants, of course, are not precluded from segregating either category of inmates from the remainder of the prison population where such segregation is otherwise justified; they simply can not segregate them under conditions as they currently exist in the Pelican Bay SHU.

1. Overview

In Farmer, supra, the Supreme Court reiterated the well-settled rule that prison officials must ‘“take reasonable measures to guarantee the safety of the inmates.’” In particular, prison officials have a duty, under the Eighth Amendment, “to protect prisoners from violence at the hands of other prisoners.” — U.S. at —, 114 S.Ct. at 1976 (quoting Hudson v. Palmer, 468 U.S. 517, 526-27, 104 S.Ct. 3194, 3200, 82 L.Ed.2d 393 (1984); Wilson, 501 U.S. at 298, 111 S.Ct. at 2324; Redman, 942 F.2d at 1442-1443 214 ; Hoptowit, 682 F.2d at 1249-1250.

Here, plaintiffs contend that defendants have failed to take reasonable steps to protect inmates’ safety in two respects: (1) they do not routinely assign to single cells those inmates who have a history of assaulting their cellmates, and (2) minimum security inmates are at times forced to share a cell with a Level IV maximum security inmate. Plaintiffs further allege that defendants have been deliberately indifferent to the risk of harm to inmates engendered by these practices.

a. Subjective Component of the Eighth Amendment

As all parties agree, the subjective component of this claim is governed by the deliberate indifference standard. Farmer, — U.S. at —, 114 S.Ct. at 1977.

b. Objective Component of the Eighth Amendment

Given the violent propensities of some prisoners, no prison can prevent all attacks on inmates by other inmates. Marsh v. Arn, 937 F.2d 1056, 1070 (6th Cir.1991); Shrader v. White, 761 F.2d 975, 980 (4th Cir.1985); Morgan, 824 F.2d at 1057 (the State is “not obliged to insure an assault-free environment”). Thus, the risk of inmate assaults must reach sufficiently serious proportions in order to implicate constitutional concerns.

In Farmer, — U.S. at —, 114 S.Ct. at 1975, an individual inmate brought suit against prison officials after being beaten and sexually assaulted by another prisoner. The Supreme Court stated that the objective component of the Eighth Amendment would be satisfied if the victimized inmate was “incarcerated under conditions posing a substantial risk of serious harm.” Farmer, — U.S. at —, 114 S.Ct. at 1977 (emphasis added). The court did not address the appropriate standard to be used when the claim of inmate violence is raised in the context of a class action, rather than an individual suit. However, a number of lower courts that have addressed class claims have utilized a standard that is similar, albeit articulated somewhat differently.

Specifically, such courts have typically required proof of a “pervasive risk of harm to inmates from other prisoners.” See, e.g., Fisher, 692 F.Supp. at 1560 (emphasis added) (internal quotations omitted). Cf. Gilland v. Owens, 718 F.Supp. 665 (W.D.Tenn.1989) (plaintiffs must establish that risk of personal harm is “a serious problem of substantial dimensions”).

To show a pervasive risk of harm, plaintiffs need not demonstrate a “reign of violence and terror.” Gilland, 718 F.Supp. at 665. On the other hand, pervasive risk of harm “may not ordinarily be shown by pointing to a single incident or isolated incidents.” Fisher, 692 F.Supp. at 1560 (internal quotations omitted. Rather, courts look for whether violence among inmates occurs with sufficient frequency that prisoners are put in reasonable fear for their safety, or whether there are constant threats of violence and assaults from other inmates. See, e.g., Ramos, 639 F.2d at 574; Fisher, 692 F.Supp. at 1560; Grubbs, 552 F.Supp. at 1128. Cf. Jensen v. Gunter, 807 F.Supp. 1463, 1482 (D.Neb.1992), appeal dismissed on other grounds sub. nom. El-Tabech v. Gunter, 992 F.2d 183 (8th Cir.1993) (risk was pervasive where level of violence was more than “unavoidable” and tensions were increased by limited cell size, noise, lack of reporting, and restricted surveillance).

In determining whether the above standards are satisfied, courts have typically examined one or more of the following: the number of inmate altercations, as reflected by lay testimony and/or statistics, whether there is evidence that inmates live in fear of assaults from other inmates, and whether there are particular conditions in the prison that contribute to a pervasive risk of inmate assault. See, e.g., LaMarca, 995 F.2d at 1535; Alberti v. Klevenhagen, 790 F.2d 1220, 1224-1226, clarified on other grounds, and reh’g en banc denied, 799 F.2d 992 (5th Cir.1986); Hoptowit, 682 F.2d at 1249); Ramos, 639 F.2d at 572-574; Fisher, 692 F.Supp. at 1527-30, 1559-1561; Jensen, 807 F.Supp. at 1482; Gilland, 718 F.Supp. at 673-688.

For example, in Gilland, supra, the district court’s finding that there was a “pervasive and constant threat of personal harm to inmates from attacks by other inmates” was based, inter alia, on inmate and staff testimony showing “constant contact” with violence, statistics showing that there were 685 inmate altercations in six months in a prison housing approximately 2,300 inmates, and “proof that inmates live in fear of personal harm.” 718 F.Supp. at 686-87. In Fisher, the court relied, inter alia, on statistics (1,300 inmate altercations in one year in a prison with a daily average population of 2,500-2,600), inmate testimony regarding actual assaults and fear of assaults, and the fact that inmates who had attacked other inmates were permitted to continue their assaultive behavior. Fisher, 692 F.Supp. at 1523-29.

2. Double celling and Cellmate Assaults

As the Tenth Circuit observed in Ramos, 639 F.2d at 572, “[a] prison setting is, at best, tense. It is sometimes explosive, and always potentially dangerous.” Given that prisons can not realistically monitor every cell at every moment, cell fights are an inevitable fact of prison life, particularly in maximum security prisons and security housing units where inmates are more likely to have violent histories or tendencies. The question then becomes whether inmate assaults in the prison are occurring at a level sufficiently beyond that which can reasonably be protected against so as to create a pervasive risk of harm.

Here, plaintiffs provided little evidence that the overall total number of cell fights over the three-year-plus period (1,158) for both the General Population facility and the SHU is significantly more than would be expected for facilities of their size and security designation. While one prison official stated that cell fights occur “frequently,” Helsel Depo. at 60-61, it is difficult to assess this statement without knowing whether the official believed that the frequency was more than would be expected. Notably, another prison official also testified that cell fights were “frequent” in the SHU, but “not as frequent as one might expect.” Lopez Tr. 14-2231.

The primary thrust of plaintiffs’ claim, however, is not that the overall number of cell fights is substantially out-of-line, but that inmates are subject to being double celled with other inmates that have a history of assaulting their cellmates. We agree that the failure to routinely consider an inmate’s prior assaultive history in making cell assignments reflects poorly on the administration of the prison. However, based on the evidence presented to the Court, only a relatively small group of inmates have engaged in repeated assaults on cellmates while at Pelican Bay. 215 For the particular inmates affected, the results have often been substantial injuries; in a few cases they have been tragic and profound. Whether the scope of the problem is of sufficient dimensions to constitute a “pervasive” or “constant” risk of harm to the class, however, is another question. In this regard, we note that whether inmates live in fear of assaults by other inmates bears on the pervasiveness of the risk. See Gilland, 718 F.Supp. at 687 (relying in part on “proof that inmates live in fear of personal harm”). Plaintiffs did not, however, identify any inmate testimony addressing this issue; in fact, the Court heard from only one inmate who discussed an inmate assault, and that testimony was directed toward plaintiffs’ claim regarding the temporary housing of Level I inmates with Level IV inmates. As such, the Court would be required to speculate as to the level and frequency of fear regarding inmate assaults that is experienced by inmates at Pelican Bay. 216

Given the above, it is less than clear that the problem of repeat assaults by inmates with a history of assaulting their cellmates has created a risk of injury to the class that is sufficiently pervasive to implicate the Eighth Amendment. Moreover, as found above, plaintiffs have not satisfactorily established that defendants knew the parameters of this problem. As such, plaintiffs can not demonstrate that defendants acted with deliberate indifference, even assuming the risk of harm was pervasive. Farmer, — U.S. at —, 114 S.Ct. at 1979. Accordingly, plaintiffs’ evidence falls short of establishing defendants’ liability under the Eighth Amendment.

In so concluding we in no way condone the current failure of correctional sergeants to routinely consider prior assaultive behavior in making cell assignments and considering cellmate “compatibility.” Indeed, this unfortunate and unexplained practice leaves those prison officials responsible for celling decisions vulnerable to incurring liability to any individual plaintiff who suffers injuries at the hands of a cellmate with an established history of assaulting cellmates. Further, with the issuance of this ruling, prison administrators at Pelican Bay will no longer lack actual knowledge of the problem. This, of course, would clearly bear on any future inquiries into the question of deliberate indifference.

3. Temporary Housing of Minimum Security Inmates with Maximum Security Inmates

With respect to this aspect of plaintiffs’ claim, plaintiffs need not demonstrate a pervasive risk of harm to the entire class; rather, it is sufficient to demonstrate that the identifiable group at issue — Level I, minimum security inmates — suffers from a pervasive risk of harm from being temporarily housed with Level IV inmates. In this regard the evidence is particularly scant. For example, there is no showing that assaults against Level I cellmates increased during the time the gymnasium was closed. Indeed, plaintiffs have identified only two instances where Level I inmates were assaulted by Level IV cellmates. In one such instance, the inmate, Charles Campbell, suffered a gruesome attack causing him to lose the tip of his nose. While Campbell’s testimony provided a moving and chilling account of this incident, neither a single incident nor isolated incidents is sufficient to demonstrate a pervasive risk of harm to Level I inmates. Fisher, 692 F.Supp. at 1560.

Given that plaintiffs have not established a pervasive risk of harm to Level I inmates, we do not reach the question whether defendants possessed the requisite mental state of deliberate indifference.

The Due Process clause of the Fourteenth Amendment provides that no State shall “deprive any person of life, liberty or property, without due process of law.” Under defendants’ current policy, inmates who are found to be affiliated with a prison gang are removed from the general prison population and confined in the SHU for an indeterminate term. Whether this practice is implemented in a manner consistent with constitutional guarantees of procedural due process is the issue before the Court. Defendants assert that current procedures satisfy or exceed due process requirements, while plaintiffs argue that they are constitutionally flawed in a number of respects. These flaws fall into two categories: (1) flaws in the procedural safeguards afforded to inmates suspected of gang affiliation, and (2) flaws in procedures governing the periodic review of inmates assigned to indeterminate terms in the SHU for prison gang affiliation. 217

To resolve this dispute we must first determine whether plaintiffs have a constitutionally protected liberty interest in remaining in the general prison population. Toussaint IV, 801 F.2d at 1089. If so, we must determine the amount of process due before they can be deprived of this liberty interest because of affiliation with a prison gang. Id. at 1098. 218 Finally, we must assess whether the plaintiff class has, in fact, been denied the quantum of process required by the Constitution.

1. Existence of a Liberty Interest

A liberty interest may arise from either of two sources: the due process clause itself or state law. Hewitt v. Helms, 459 U.S. 460, 466, 103 S.Ct. 864, 869, 74 L.Ed.2d 675 (1983); Smith v. Noonan, 992 F.2d 987, 989 (9th Cir.1993); Toussaint IV, 801 F.2d at 1089. State law 219 creates a liberty interest if it substantively limits official discretion by establishing particularized standards or criteria that govern state decision-makers. Kentucky Dep’t of Corrections v. Thompson, 490 U.S. 454, 462, 109 S.Ct. 1904, 1909, 104 L.Ed.2d 506 (1989); Conner v. Sakai, 994 F.2d 1408, 1411, amended 15 F.3d 1463 (9th Cir.1993), cert. granted in part, — U.S. —, 115 S.Ct. 305, 130 L.Ed.2d 217 (1994). The state must also require, “in explicitly mandatory language,” that if these standards or criteria are met, a particular outcome must follow. Kentucky Dep’t of Corrections, 490 U.S. at 463, 109 S.Ct. at 1910 (internal quotations omitted); Conner, 994 F.2d at 1411; Toussaint IV, 801 F.2d at 1095 (regulations that follow a “shall/unless” formula create liberty, interest); Nelson v. Bryan, 607 F.Supp. 959, 961 (D.Nev.1985) (state creates liberty interest by requiring decision-makers to base decisions on objective and defined criteria).

The liberty interest at issue here is the interest of prisoners in remaining in the general prison population and not being confined in a security housing unit. As the higher courts have held, the due process clause itself does not protect such an interest. See, e.g., Hewitt, 459 U.S. at 467-468, 103 S.Ct. at 869-70; Smith v. Noonan, 992 F.2d at 989.

In Toussaint IV, however, the Ninth Circuit held that sections 3335(a) 3336, and 3339(a) of Title 15 of the California Code of Regulations, taken together, do create a constitutionally protected liberty interest to be free from placement in administrative segregation. 801 F.2d at 1097-98. As the Court explained, these regulations combine to prohibit state officials from retaining an inmate in administrative segregation unless one of three substantive predicates is met: (1) the inmate presents an immediate threat to the safety of the inmate or others, (2) the inmate endangers institution security, or (3) the inmate jeopardizes the integrity of an investigation of an alleged serious misconduct or criminal activity. As such, they sufficiently fetter official decision-making to create a protected liberty interest.

Although not addressed by Toussaint IV, we conclude that another California regulation, Cal.Code Regs. tit. 15, § 3341.5(c)(3), also creates a liberty interest in freedom from administrative segregation. That' section provides that an inmate shall not be retained in the SHU beyond 11 months absent a classification committee determination that retention in the SHU is required because of one of three specific reasons: “(A) The inmate has an unexpired [Minimum Eligible Release Date] from the SHU, (B) Release of the inmate would severely endanger the lives of inmates or staff, the security of the institution, or the integrity of a investigation into suspected criminal activity or serious misconduct, [or] (C) The inmate has voluntarily requested continued retention in segregation.” Cal.Code Regs. tit. 15, § 3341.5(c)(3).

Like the regulations examined in Toussaint IV, section 3341.5(c) explicitly and substantively limits the exercise of official discretion by imposing a mandatory duty on state officials to release an inmate from the. SHU unless one of the above three predicates is met. Accordingly, section 3341.5(c) provides a separate basis for plaintiffs’ liberty interest in being housed in the general prison population with respect to those inmates that have been confined in the SHU for over 11 months. 220

Given the above, defendants may not confine prison gang members in the SHU, nor hold them there on indeterminate terms, without providing them the quantum of procedural due process required by the Constitution.

2. Amount of Process Required by the Due Process Clause

The Supreme- Court has twice addressed the amount of due process that the Constitu tion affords inmates with protected liberty interests. In Wolff v. McDonnell, 418 U.S. 539, 94 S.Ct. 2963, 41 L.Ed.2d 935 (1974), the Court considered the process required before a prison official can punish an inmate for serious misconduct after incarceration. Although the specific sanction at issue was denial of “good-time” credits, Wolff applies equally where the sanction is disciplinary segregation in a security housing unit such as the Pelican Bay SHU. Id. at 571, n. 19, 94 S.Ct. at 2982, n. 19; Conner, 994 F.2d at 1410-13; Toussaint IV, 801 F.2d at 1099.

After balancing the competing interests at stake, the Court held that the inmate in Wolff was entitled to the following due process protections: (1) advance written notice of the disciplinary charges, (2) an opportunity to call witnesses and present evidence if doing so would not unduly jeopardize institutional safety or correctional goals, (3) assistance from another inmate or prison staff if the inmate is illiterate or the complexity of the issues makes it difficult to collect and present the evidence necessary for an adequate comprehension of the.case, and (4) a written decision and summary of the evidence relied on. Wolff, 418 U.S. at 563-70, 94 S.Ct. at 2978-82. The prison was not, however, required to permit the cross-examination of witnesses or the participation of counsel. Id. at 567-69, 94 S.Ct. at 2979-81.

In Hewitt, 459 U.S. 460, 103 S.Ct. 864, the Supreme Court considered the amount of process required before the state can transfer an inmate to a security housing unit for “administrative” reasons. This type of “administrative segregation” is not utilized to punish the inmate for specific misconduct, as was the case in Wolff, but to further some legitimate need of the prison. Taylor v. Koon, 682 F.Supp. 475, 477 (D.Nev.1988). Thus, administrative segregation may properly be used to protect the prisoner’s safety, to protect other inmates from a particular prisoner, to break up potentially disruptive groups of inmates or to await completion of an investigation into misconduct charges. Hewitt, 459 U.S. at 468, 476-77, 103 S.Ct. at 869-70, 874; Toussaint IV, 801 F.2d at 1098.

The Court held that the amount of process required in cases of administrative segregation is substantially less than that required in Wolff-type proceedings. Hewitt, 459 U.S. at 473-476, 103 S.Ct. at 872-874. As the Ninth Circuit recently summarized:

Due process, in the administrative context, merely requires that the prison officials provide the inmate with some notice of the charges against him and an opportunity to present [the inmate’s] views to the prison official charged with deciding whether to transfer [the inmate] to administrative segregation.

Barnett v. Centoni, 31 F.3d 813, 815 (9th Cir.1994) (internal quotations omitted). These requirements must be satisfied within “a reasonable time after the prisoner is segregated.” Toussaint IV, 801 F.2d at 1100. Due process does not require “detailed written notice of charges, representation by counsel or counsel-substitute, an opportunity to present witnesses, or a written decision describing the reasons for placing the prisoner in administrative segregation.” Toussaint IV, 801 F.2d at 1100-01.

Thus, under Wolff and Hewitt, the amount of process due depends, in significant part, on whether the prisoner’s transfer to the SHU is characterized as disciplinary or administrative. Neither case, however, examines whether transferring an inmate to an indeterminate term in the SHU for gang affiliation is administrative or disciplinary. In Toussaint v. Rowland (Toussaint V), 711 F.Supp. 536 (N.D.Cal.1989), the district court assumed, without deciding, that segregation of prison gang members in a security housing unit falls under the ambit of administrative segregation. See id. at 539, n. 9. See also Toussaint VI, 926 F.2d at 804 (Wiggins, J., concurring). The court concluded, however, in affirming a report submitted by the court-appointed Monitor, that “difficulties engendered by determining prison gang membership may create a need for special due process procedures to ensure compliance with constitutional requirements.” Toussaint V, 711 F.Supp. at 541. 221 The court also affirmed the Monitor’s finding that inmates subject to indeterminate segregation have a more significant liberty interest than was present in Hewitt, which only involved a temporary transfer to administrative segregation pending investigation into misconduct charges. Id. at 541-42. See Mims v. Shapp, 744 F.2d 946, 953-64 (3rd Cir.1984).

Accordingly, the district court held that due process required the following procedures with respect to inmates suspected of gang affiliation: (1) the prisoner should be afforded the opportunity to present his views to the CAC [Criminal Activities Coordinator, now referred to as the IGI] prior to any decision to retain the prisoner in segregation for an indeterminate period, 222 (2) the CAC must designate the prisoner as being a current active member of a prison gang prior to any ICC decision to retain the prisoner in segregation for an indeterminate period, and (3) the CAC must reevaluate his determination regarding gang membership of the prisoner every 90 days. Toussaint V, 711 F.Supp. at 540, 541-42, n. 15, 543; Monitor’s Report at ¶ 47. On appeal, the Ninth Circuit expressly affirmed that inmates must be allowed to present their views to the CAC, but held that it was within defendants’ discretion to hold reviews of administrative segregation decisions every 120 days, rather than every 90 days. Toussaint VI, 926 F.2d at 803.

In addition to purely procedural protections, due process also requires prison officials to have an evidentiary basis for their decisions to confine an inmate to a security housing unit, whether the purpose of that segregation is disciplinary or administrative. Superintendent, Massachusetts Correctional Institution v. Hill, 472 U.S. 445, 455, 105 S.Ct. 2768, 2774, 86 L.Ed.2d 356 (1985). In particular, due process requires that such decisions be supported by “some evidence.” Id.; Toussaint IV, 801 F.2d at 1103-06. This standard is only “minimally stringent.” Cato v. Rushen, 824 F.2d 703, 705 (9th Cir.1987). The relevant inquiry is whether there is any evidence in the record that could support the conclusion reached by the prison decision-makers. Id.

The Ninth Circuit has also held, in the context of prison disciplinary proceedings, that the information relied upon must have at least “some indicia of reliability.” Cato, 824 F.2d at 705 (citing Mendoza v. Miller, 779 F.2d 1287, 1295 (7th Cir.1985), cert. denied, 476 U.S. 1142, 106 S.Ct. 2251, 90 L.Ed.2d 697 (1986)). When this information includes statements from confidential informants, as is often the case, the record must contain “some factual information from which the committee can reasonably conclude that the information was reliable.” Zimmerlee v. Keeney, 831 F.2d 183, 186 (9th Cir.1987), cert. denied, 487 U.S. 1207, 108 S.Ct. 2851, 101 L.Ed.2d 888 (1988). The record must also contain “a prison official’s affirmative statement that safety considerations prevent the disclosure of the informant’s name.” Id.

In Toussaint IV, the Ninth Circuit confirmed that in administrative segregation proceedings due process does not require disclosure of the identity of confidential informants, 801 F.2d at 1101, but it has yet to squarely address whether the “indicia of reliability” standards applied in the context of disciplinary hearings also, apply when inmates are placed in administrative segregation.

The district court in Toussaint did, however, conclude that “defendants have the burden [of showing] some evidence in the record to support a [administrative] segregation decision, and that evidence must have some indicia of reliability.” Toussaint V, 711 F.Supp. at 542 (internal quotations omitted). We agree that the evidence relied upon to confine an inmate to the SHU for gang affili ation must have “some indicia of reliability” to satisfy due process requirements.

The “touchstone of due process is protection of the individual against arbitrary action of government.” Wolff, 418 U.S. at 558, 94 S.Ct. at 2976. Allowing prisons to consign an inmate to the SHU for an indeterminate term, without ascertaining whether the information relied upon has “some indicia of reliability,” fails to protect against such arbitrary action. This is particularly so given the realities of prison life. As one court observed:

In a prison environment, where authorities must depend heavily upon informers to report violations of regulations, an inmate can seek to harm a disliked fellow inmate by accusing that inmate of wrongdoing. Since the accuser is usually protected by a veil of confidentiality that will not be pierced through confrontation and cross-examination, an accuser may easily concoct the allegations of wrongdoing. Without a bona fide evaluation of the credibility and reliability of the evidence presented, a prison committee’s hearing would thus be reduced to a sham which would improperly subject an inmate accused of wrongdoing to an arbitrary determination.

Kyle v. Hanberry, 677 F.2d 1386, 1390 (11th Cir.1982). Nor can we ignore that the information relied upon may often be obtained, at least in part, from “debriefings.” While some effort is made to assess the accuracy of such information, the clear incentive to fabricate or exaggerate information in order to gain release from the SHU significantly heightens the risk that false information will be relied upon. 223

Not only is the risk of false information high, but the consequences of an improper validation — confinement in the SHU for an indeterminate term, with all its attendant restrictions and adverse impacts on parole— are severe. Moreover, inmates improperly validated as gang members have little chance of rectifying such an error or otherwise obtaining release from the SHU. Such inmates will be unable to “debrief,” since, never having been gang members, they will not have acquired gang information that can be divulged; nor are they likely to possess any means of affirmatively proving that the information relied upon is false. Finally, we note that defendants have not asserted that applying a “reliability” standard in administrative segregation proceedings would unduly hinder those proceedings or otherwise jeopardize institutional security.

Accordingly, it is our conclusion that, in order to satisfy due process, an inmate may not be confined to the SHU for gang affiliation unless the record contains “some factual information” from which the IGI and classification committee “can reasonably conclude that the information was reliable.” Zimmerlee, 831 F.2d at 186. Such a requirement will “help prevent arbitrary deprivations without threatening institutional interests or imposing undue administrative burdens.” Hill, 472 U.S. at 455, 105 S.Ct. at 2774.

3. Plaintiffs’ Specific Objections

Having in the mind the above guidance, we now turn to plaintiffs’ specific objections regarding the amount of process afforded to Pelican Bay inmates transferred to the SHU based on their membership or association with a prison gang.

a. Failure to Provide Wolff protections

Under current policy, prison gang members and associates are assigned to administrative segregation for indeterminate terms on the ground that they pose a risk to the security of the prison. Plaintiffs contend that although defendants invoke the rubric of administrative segregation, their policy of confining gang affiliates in the SHU is in fact designed to punish and deter rather than to advance legitimate administrative purposes. Thus, they urge the Court to require the more stringent due process safeguards required by Wolff. There is no dispute that such safeguards are not currently provided.

We begin with the premise that prison gang membership and association is a threat to institutional security, and that therefore such members and associates are properly subject to administrative segregation. Cf . Toussaint VI, 926 F.2d at 804 (Wiggins, J., concurring); Toussaint V, 711 F.Supp. at 540-43. Plaintiffs ask us to find, however, that the harsh conditions in the Pelican Bay SHU render such segregation disciplinary rather than administrative.

As plaintiffs point out, the conditions imposed on gang members in the SHU are the same conditions imposed on inmates who are transferred to the SHU for set terms as punishment for specific misconduct pursuant to Wolff proceedings. Nor is there any disputing that the conditions in the SHU, described more fully in section 11(D)(1), supra, are severe.

Indeed, there is little doubt that the SHU’s decidedly harsh regimen contains an element of punishment and creates a deterrent effect. As one gang investigator agreed, the policy of sending gang members to the Pelican Bay SHU “sendfs] a message to other prisoners, that if you join a gang you’re going to get sent to Pelican Bay and life’s going to be tough.” Hawkes Depo. at 530-531; see also Gomez Tr. at 28-4461 (acknowledging deterrent effect). However, this is largely an inevitable byproduct of the fact that inmates in segregation are typically subject to restrictions that are substantially more onerous than those imposed on the general population. This fact alone, however, does not justify greater procedural protections than would otherwise be required in cases of administrative segregation. See Toussaint IV, 801 F.2d at 1099-1100 (fact that conditions in administrative segregation may involve “severe hardships” including denial of access to vocational, educational, recreational, and rehabilitative programs, restrictions on exercise, and confinement to one’s cell for lengthy periods, does hot justify a heightened level of due process).

Nor are conditions in the SHU, when taken as a whole, so extreme in relation to defendants’ stated administrative purposes that we must infer that their actual primary purpose is to “punish” or discipline gang members. Prison gang members rely on communication networks and opportunities for interaction to maintain the organization and carry out gang activities. For the most part, the conditions in the SHU serve to undermine these networks and opportunities by separating gang members from one another, and others, through a regimen of social isolation. 224 Thus, the nature of the conditions, by itself, does not persuade us that prison gang members are transferred to the SHU primarily as discipline for specific misconduct. Accordingly, we decline to find that Wolff governs the segregation of prison gang members and associates at Pelican Bay. 225

b. Failure to Permit Hearing Before SSU Agents

As set forth in the findings of fact, inmates are given an opportunity to present their views to the IGI before the IGI submits a validation package to the Special Services Unit (“SSU”) in Sacramento. They are also given an opportunity to present their views to the Institutional Classification Committee (“ICC”) prior to being assigned to the SHU for an indeterminate term for gang affiliation. They are not, however, permitted to present their views to the SSU agent who actually' validates the inmate as a prison gang member after reviewing the validation package submitted by the IGI.

As Hewitt makes clear, inmates assigned to administrative segregation are entitled to an informal hearing where they can present their views to the official “charged with deciding whether to transfer [the inmate] to administrative segregation.” Hewitt, 459 U.S. at 476, 103 S.Ct. at 874. As plaintiffs emphasize, only SSU agents can formally validate an inmate as a gang member or associate. Moreover, such a validation is usually the functional equivalent of deciding that the inmate will be transferred to the SHU for gang affiliation. Therefore, plaintiffs argue, defendants must afford inmates an opportunity to present their views to the SSU agent charged with reviewing the validation package submitted by the IGI.

While plaintiffs’ argument has superficial appeal, it promotes form over substance. Although the SSU agent formally validates the inmate, it is clear that the critical “decision-maker” in the process is still the IGI. As detailed in the factual findings, the SSU plays a technically important but substantively nominal role in the process. Nor are we persuaded that IGIs are unaware of the significance of their role. Given that inmates have an opportunity to present their views to the IGI and the ICC, the failure to provide a hearing before the SSU officer does not violate due process. 226

c. Failure to Provide for Meaningful Hearings

Plaintiffs contend that the hearings that are provided before the IGI and ICC are perfunctory formalities because prison officials have already made up their minds before meeting with the inmate. As such, the hearings violate the fundamental tenet of due process that opportunities to be heard must be granted “in a meaningful manner.” Parratt v. Taylor, 451 U.S. 527, 540, 101 S.Ct. 1908, 1915, 68 L.Ed.2d 420 (1981), overruled on other grounds in Daniels v. Williams, 474 U.S. 327, 106 S.Ct. 662, 88 L.Ed.2d 662 (1986); see also Toussaint IV, 801 F.2d at 1102 (affirming district court’s imposition of substantive criteria to “assure that plaintiffs’ due process rights are not meaningless gestures”).

While the potential for hollow gestures can not be denied, plaintiffs have not presented evidence that hearings before the IGI are meaningless as matter of course so as to establish liability on a classwide basis. Plaintiffs rely on a deposition excerpt from IGI Briddle in which he admits that his mind had already been “made up” with respect to a particular inmate. Briddle Depo. at 335-36. 227 However, we can not infer from this particular example a systematic failure to provide meaningful hearings. Indeed, Brid-dle also testified that the inmate’s views can impact the process. See Briddle Depo. at 347 (“If after discussing the case with the inmate I have concerns or problems with pursuing the case as far as the validation goes, the opportunity would be that the validation package would not be sent. And we’ve actually gone back and we bring the inmates back and some interaction does occur.”).

With respect to the ICC, it is clear that the Committee is predisposed to transfer any validated inmate to the SHU. However, this simply reflects prison policy regarding prison gangs and underscores the importance of the hearing with the IGI, particularly in light of the Court’s finding that the IGI, and not the SSU, is the critical decision-maker. When we consider the opportunities for hearing before the IGI and the ICC together, and the record as a whole, we decline to find that the process provided to the class is no more than a meaningless gesture.

d.Reliability Determinations for Confidential Informants

Under California regulations, prison officials may conclude that the information provided by a confidential informant is “reliable” if one of five criteria is met: (1) the confidential informant (“Cl”) has previously given information which has proved to be true, (2) other confidential sources have independently provided the same information, (3) the information provided by the Cl is self-incriminating, (4) part of the information provided is proven true, or (5) the confidential source is the victim. Cal.Code Regs. tit. 15 § 3321(c).

The Monitor in Toussaint considered the first four criteria (apparently the fifth criterion is a subsequent development) and found that they “may be appropriate safeguards of the reliability of confidential information, provided that they are not applied in a rote fashion without regard to the realities of the particular informant report under consideration. However, the ultimate judgement as to whether an ICC has sufficient indicia of the reliability of confidential information to use that information as the basis for segregating a prisoner must always be made by each ICC on a ease-by-case basis.” Monitor’s Report at ¶ 58 (emphasis added). This aspect of the Monitor’s report was adopted by the district court. Toussaint V, 711 F.Supp. at 543.

Plaintiffs now challenge the fifth criterion on the ground that a status of “victim” does not imbue the informant with any reliability, and that, in fact, such a person may well harbor ulterior motives to retaliate against the aggressor. Such potential undoubtedly exists. However, we conclude that, like the other criteria, it may be an appropriate safeguard of reliability so long as it is “not applied in a rote fashion without regard to the realities of the particular informant report under consideration.” Monitor’s Report at ¶ 58.

Plaintiffs also assert that defendants ignore this important caveat and routinely apply all five criteria in a “rote fashion.” The evidence in the record, however, is too sparse to draw allow us to draw such a conclusion. This does not, of course, preclude any individual inmate who believes he was wrongfully validated from challenging his validation on the ground that the record lacks “some indi-cia of reliability” with respect to the confidential information relied upon. In such a case, the prison must do more than simply invoke “in a rote fashion” one of the five criteria. It must also show that the “realities of the particular informant report” were taken into consideration.

e.Reliance on Hearsay

Plaintiffs also assert, citing Cato, 824 F.2d at 704-06, that defendants violate the evidentiary standards discussed above because IGIs rely on hearsay statements. Cato, however, did not preclude the use of hearsay per se. See also Helms v. Hewitt, 655 F.2d 487, 502 (3rd Cir.1981), rev’d on other grounds, 459 U.S. 460, 103 S.Ct. 864, 74 L.Ed.2d 675 (1983) (“We realize that in prison disciplinary proceedings hearsay may serve a useful purpose and we do not preclude it”). Rather, Cato held that where the only evidence in the record is a single, uncorroborated, hearsay statement, this does not constitute “some evidence” that has “some indicia of reliability.” Cato, 824 F.2d at 705. Plaintiffs have not demonstrated that defendants are failing to comply with the holding in Cato.

f.Inadequacy of Periodic Reviews

(i) Timing of Reviews

Administrative segregation can not be used as a “pretext for indefinite commit ment of an inmate.” Hewitt, 459 U.S. at 477 n. 9, 103 S.Ct. at 874 n. 9. Thus, prison officials must conduct some sort of periodic review of the confinement of prisoners in administrative segregation. Id.; Toussaint IV, 801 F.2d at 1101.

California regulations provide that inmates confined to the SHU for indeterminate terms must be reviewed by a classification committee every 180 days for consideration of release to the general inmate population. Cal. Code Regs. tit. 15, § 3341.5(c). Plaintiffs contend that a 180-day timetable violates constitutional requirements. See Toussaint VI, 926 F.2d at 803 (finding 120-day schedule permissible without deciding whether 120 days represents outer constitutional limit). We conclude, however, that this issue is not ripe for review. Despite the 180-day allowance in the regulations, the record reflects that defendants actually conduct reviews at Pelican Bay every 120 days; plaintiffs have identified no class member that has been required to wait 180 days for a classification review. Nor have plaintiffs submitted any evidence indicating that the current 120-day schedule is likely to be extended to 180 days in the future.

(ii) Failure to Consider Lack of Gang Activity or “Exculpatory Evidence”

As discussed above, due process requires that there be “some evidence” with “some indicia of reliability” to support an inmate’s placement in administrative segregation. Toussaint IV, 801 F.2d at 1103-06. This same standard applies to subsequent periodic reviews conducted every 120 days by the UCC, and annually by the ICC. Under current policy, an inmate is considered to be a security threat so long as the inmate is validated as a gang affiliate and has not yet debriefed. Thus, a validated inmate will continue to be retained in the SHU, absent a successful debriefing, even if the inmate has, for some period of time “remained clean”— i.e., there is no evidence of continued commitment to the gang, as reflected by participation in gang activity or association with other gang members. For example, an inmate who was validated in 1979, but has not engaged in any gang activity or otherwise associated with gang members since then, will still be retained in the SHU in 1994, fifteen years later, absent a successful debriefing. The lack of continuing evidence of gang membership or activity is simply considered irrelevant since the justification for administration segregation is the fact of gang membership itself, not any particular behavior or activity.

Plaintiffs contend that this policy improperly fails to consider “exculpatory” evidence. What plaintiffs are essentially arguing, however, is that, at some point, there is no longer “some evidence” to retain an inmate in the SHU, despite the absence of a debriefing, where the inmate has not engaged in any prison gang activity and there is no new evidence confirming the inmate’s continued association with the prison gang.

Plaintiffs’ objection must fail in light of our factual findings. As set forth in section 11(F), supra, the record supports defendants’ position that gang members and associates are threats to prison security, and that inmates who join such gangs join “for life.” As such, the fact that the inmate may not have affirmatively engaged in gang activity after confinement in the SHU does not, in and of itself, vitiate the inmate’s gang membership. Therefore, the premise for finding that the inmate is a security risk — gang-membership or association — is not affected by the lack of subsequent gang activity. We also note that opportunities for such activity-in the SHU are extremely limited, see section 11(D)(1), supra, although, as defendants acknowledge, a few inmates have nonetheless managed to continue some limited gang activity from the SHU. Accordingly, defendants do not violate due process by failing to give persuasive value to the fact that an inmate’s record reflects an absence of gang-related activity or association over some period of time. 228

(iii) Reliance on Evidence Previously Rejected

As set forth in thé factual findings, the SSU may reject any number of items submitted in a validation package and still validate an inmate, so long as there are at least three independent source items that the SSU does not reject. The rejection of such items, however, is not recorded in the inmate’s record. This practice creates a disturbing likelihood that the UCC or ICC, as assisted by the IGI, will improperly determine that there is still “some evidence” that an inmate is gang affiliated, in those cases where some of the items initially relied upon to validate the inmate are later called into doubt.

The following example illustrates this point. A validation package, containing five items of evidence, is submitted to the SSU. Rejecting two of the items, the SSU validates the inmate based on the remaining three. Two of these three items are later discredited based on subsequent information showing that the informants who provided the information are no longer considered reliable. This leaves only one circumstantial item of evidence supporting the validation which, standing alone, fails to satisfy the “some evidence” standard. However, the prison officials involved in the inmate’s periodic review will erroneously conclude that there is “some evidence” to support the continued validation and segregation of the inmate because they will be unaware that the SSU has already determined that two of the remaining three items can not be relied upon.

This flaw can not simply be dismissed as a matter of internal management. The very purpose of the “some evidence” requirement is to protect inmates from confinement in the SHU on arbitrary or baseless grounds. Yet, under defendants’ procedures, this minimum safeguard against arbitrary deprivations may be undermined in those very instances where the need for careful periodic review is needed most. If the “some evidence” requirement is to have meaning, then the decision-makers applying that standard can not be allowed to unknowingly rely on evidence previously rejected by the SSU. Accordingly, we conclude that due process requires that, when the SSU rejects an item of evidence, this fact must be included in the inmate’s central file so that it will be made available to those participating in the inmate’s periodic review.

IV.

SUMMARY

Throughout these proceedings, we have been acutely sensitive to the fact that our role in Eighth Amendment litigation is a limited one. Federal courts are not instruments for prison reform, and federal judges are not prison administrators. We must be careful not to stray into matters that our system of federalism reserves for the discretion of state officials. At the same time, we have no duty more important than that of enforcing constitutional rights, no matter how unpopular the cause or powerless the plaintiff. The challenge, then, in prison condition cases, is to uphold the Constitution in such a manner that respects the state’s unique interest in managing its prison population. It is a challenge that requires us to draw constitutional lines when necessary, yet minimize any intrusion into state affairs.

It was with these principles in mind that we studied the voluminous record in this ease and rendered our findings of fact and conclusions of law set forth above. And it is these principles that have compelled us to conclude that defendants have unmistakably crossed the constitutional line with respect to some of the claims raised by this action. In particular, defendants have failed to provide inmates at Pelican Bay with constitutionally adequate medical and mental health care, and have permitted and condoned a pattern of using excessive force, all in conscious disregard of the serious harm that these practices inflict. With respect to the SHU, defendants cross the constitutional line when they force certain subgroups of the prison population, including the mentally ill, to endure the conditions in the SHU, despite knowing that the likely consequence for such inmates is serious injury to their mental health, and despite the fact that certain conditions in the SHU have a relationship to legitimate security interests that is tangential at best.

As to the above matters, defendants have subjected plaintiffs to “unnecessary and wanton infliction of pain” in violation of the Eighth Amendment of the United States Constitution. We observe that while this simple phrase articulates the legal standard, dry words on paper can not adequately capture the senseless suffering and sometimes wretched misery that defendants’ unconstitutional practices leave in their wake. The anguish of descending into serious mental illness, the pain of physical abuse, or the torment of having serious medical needs that simply go unmet is profoundly difficult, if not impossible, to fully fathom, no matter how long or detailed the trial record may be.

The record does not, however, sustain other allegations advanced by plaintiffs. Conditions in the SHU may well hover on the edge of what is humanly tolerable for those with normal resilience, particularly when endured for extended periods of time. They do not, however, violate exacting Eighth Amendment standards, except for the specific population subgroups identified in this opinion. We have also found for defendants with respect to plaintiffs’ allegations regarding the use of force between inmates. Finally, with the exception of one issue, we have rejected plaintiffs’ challenges to the procedures governing the assignment of prison gang members to the SHU for indeterminate terms. 229

V.

APPROPRIATE RELIEF AND FURTHER PROCEEDINGS

Once constitutional violations have been found, federal courts have broad equitable powers to formulate appropriate relief. Stone v. City and County of San Francisco, 968 F.2d 850, 861 (9th Cir.1992), cert. denied, — U.S. -, 113 S.Ct. 1050, 122s L.Ed.2d 358 (1993); Hoptowit, 682 F.2d at 1245. We should only exercise, however, the least power necessary to accomplish this goal. Courts must “ ‘fashion a remedy that does no more and no less than correct [the] particular constitutional violation.’” Doty, 37 F.3d at 543; Toussaint IV, 801 F.2d at 1086 (“Injunctive relief against a state agency or official must be no broader than necessary to remedy the constitutional violation”). Such a remedy may include relief that the Constitution would not of its own force initially require, but only “if such relief is necessary to remedy a constitutional violation.” Toussaint IV, 801 F.2d at 1087; Gluth v. Kangas, 951 F.2d 1504, 1510 n. 4 (9th Cir.1991).

To facilitate a remedy that both cures the constitutional deficiencies and minimizes intrusion into prison management, most district courts require the development and implementation of a remedial plan that is narrowly tailored to correct the specific constitutional violations at issue. See, e.g., Casey, 834 F.Supp. at 1552-53; Lightfoot, 486 F.Supp. at 527-528. We see no reason to deviate from this approach in the case at bar. Injunctive or equitable relief is appropriate, and indeed necessary, where there is a “contemporary violation of a nature likely to continue.” See Farmer, — U.S. at -, 114 S.Ct. at 1983 (internal quotation omitted); Williams v. Lane, 851 F.2d 867, 885 (7th Cir.1988), cert. denied, 488 U.S. 1047, 109 S.Ct. 879, 102 L.Ed.2d 1001 (1989) (“District courts may ... order appropriate injunctive relief to prevent any continuing deprivation of an inmate’s constitutional rights”). We are firmly convinced that the constitutional violations identified above will not be fully redressed absent intervention by this Court.

In reaching this conclusion we have heeded the United States Supreme Court’s recent admonition that, where injunctive relief is sought, the plaintiff must show not only that defendants possess the subjective state of mind necessary to establish Eighth Amend ment liability, but that this state of mind will persist beyond the instant litigation. Farmer, — U.S. at —, 114 S.Ct. at 1988. We must thus evaluate defendants “attitudes and conduct” not only as of the time the suit was filed, but also during the litigation and “into the future.” Farmer, — U.S. at —, 114 S.Ct. at 1983.

Our assessment of defendants’ current attitudes and conduct only reinforces our view that injunctive relief is not only appropriate in this case, but perhaps “indispensable, if constitutional dictates — not to mention considerations of basic humanity— are to be observed in the prison[ ].” Stone, 968 F.2d at 861. Throughout this litigation, defendants have shown no indication that they are committed to finding permanent solutions to problems of serious constitutional dimension. On the contrary, defendants have expended most of their energies attempting to deny or explain away the evidence of such problems. Even when defendants modify certain policies (as they have done in the use-of-foree area), they do not argue that such changes evidence an intent to address the problems raised by this complaint; rather, defendants typically assert that they were precipitated by unrelated matters. 230

In short, we glean no serious or genuine commitment to significantly improving the delivery of health care services, correcting the pattern of excessive force, or otherwise remedying the constitutional violations found herein which have caused, and continue to cause, significant harm to the plaintiffs. Indeed, the Court is left with the opinion that, even given the evidence presented at trial, defendants would still deny that any condition or practice at Pelican Bay raises any cause for concern, much less concern of a constitutional dimension.

Nor are we confident, given the history of other prison litigation, that defendants will promptly rectify constitutional deficiencies absent intervention by this Court. See, e.g., Thompson v. Enomoto, 915 F.2d 1383, 1387 (9th Cir.1990), cert. denied, 502 U.S. 1071, 112 S.Ct. 965, 117 L.Ed.2d 131 (1992) (Court Monitor reports showed that state prison officials had not complied with decree governing conditions for death row inmates in 1988 and 1989); Gates v. Deukmejian, No. Civ. S-87-1636 LKK (October 27, 1994 Order at 8) (finding state prison officials in contempt of decree after “four-year pattern of delay and obstruction regarding the planning and implementation of a satisfactory OPP [Outpatient Psychiatric Program]”). The Office of Legislative Analyst has also observed that defendants have failed over the years to undertake adequate planning to address the medical needs of inmates in California prisons. See Trial Exh. P-3958 at 32 (“Our review indicates that, although the CDC has made some improvements in administration of its medical programs, the programs are too often characterized by a lack of adequate long-term planning, and ‘crisis management,’ often brought about by litigation.... The lack of long-term planning has been apparent over the years.”).

Considering all of the above, we conclude that injunctive relief is both necessary and appropriate to ensure an effective remedy of the constitutional violations at issue here. We also believe, given the above, that the participation of counsel for both parties, as well as a Special Master experienced in prison administration, will be essential to the formulation of a remedy that is both effective and narrowly tailored.

The appointment of a Special Master, with appropriately defined powers, is within both the inherent equitable powers of the court and the provisions of Rule 53 of the Federal Rules of Civil Procedure. Ruiz, 679 F.2d at 1159-62.

In this case, the assistance of a Special Master is clearly appropriate. Developing a comprehensive remedy in this case will be a complex undertaking involving issues of a technical and highly charged nature. The Court strongly believes that the participation of a well-qualified and impartial Special Master will greatly assist the Court in developing an appropriate remedy. The assistance of a Special Master will also be necessary to properly monitor the implementation of any remedy that this Court may order. Such a task will require a substantial expenditure of time and the expertise of someone experienced in prison administration. See Stone, 968 F.2d at 859 n. 18 (noting that federal courts “repeatedly have approved the use of Special Masters to monitor compliance with court orders and consent decrees”) (citations omitted); Williams, 851 F.2d at 885 (appointment of a “knowledgeable and impartial special master to implement a just remedy consistent with the needs of prison security and legitimate penological goals should assure compliance with the court’s ultimate decision”); Ruiz, 679 F.2d at 1159-62, 1165; Mercer v. Mitchell, 908 F.2d 763, 785 (11th Cir.1990) (district court appointed temporary Special Master to recommend remedial measures); Armstrong v. O’Connell, 416 F.Supp. 1325, 1340 (E.D.Wis.1976) (Special Master to assist in developing remedy in school desegregation ease). We note that the court is “not required to await the failure or refusal of [prison officials] to comply with [a] decree before appointing an agent [under Rule 53] to implement it.” Ruiz, 679 F.2d at 1161.

In addressing the scope and substance of the remedial plan, the parties and Special Master are reminded that “federal courts do not sit to supervise state prisons, the administration of which is of acute interest to the states.” Toussaint IV, 801 F.2d. at 1087. However, it is also the duty and responsibility of this Court to ensure that constitutional rights are fully vindicated. Thus, the parties and Special Master should keep in mind that any equitable remedy must “strike a balance ... that will both redress the constitutional violations found and yet accord appropriate deference to the defendants’ interests in running their own institution.” Fisher, 692 F.Supp. at 1567. This requires that any remedial plan be minimally intrusive and accord substantial deference to defendants’ legitimate interest in managing a correctional facility. Toussaint IV, 801 F.2d at 1087 (court has duty to fashion least intrusive remedy that is still effective). Accordingly, defendants’ policy preferences must be given deference unless doing so would preclude an effective remedy. See Hoptowit, 682 F.2d at 1254 (remedy “should permit, if possible within constitutional restraints, the prison officials to use the general approach that they find most effective and efficient”).

Accordingly, and good cause appearing, it is HEREBY ORDERED that:

1. Defendants’ February 11, 1994 Renewed Motion for Partial Judgment under Fed.R.Civ.P. 52(c), and to Strike Declarations of Grassian and Start, is denied.

2. The Court appoints Mr. Thomas F. Lonergan 231 to serve as a neutral Special Master, pursuant to Fed.R.Civ.P. 53 and the inherent powers of the Court, for the purpose of assisting the Court to fulfill its obligation to fashion an appropriate remedy and to monitor the implementation of that remedy. The specific duties and powers of the Monitor, along with other terms of his appointment, shall be governed by a separate order of reference to be issued forthwith.

3. As soon as practicable, counsel for plaintiffs and defendants shall begin working together jointly and in good faith, with the Special Master, to develop a satisfactory remedial plan that addresses the constitutional violations set forth in the accompanying conclusions of law. 232 The parties shall submit their proposed remedial plan to the Court within 120 days from the date of this order. The Special Master shall provide a progress report to the Court every 30 days and may recommend, for good cause, an extension of time beyond the 120 day deadline.

4. The Court fully anticipates that an appropriate remedial plan can be fashioned through the above process. In the event, however, that the parties are unable to develop a mutually acceptable remedial plan within the 120 day deadline (or such later deadline as the Court may allow by way of extension), the parties shall, no later than 7 days after such deadline, jointly submit to the Court any part(s) of such a plan that have been agreed to, or a statement that the parties were unable to agree on any aspect of a remedial plan. The Special Master shall then make recommendations to the Court with respect to any remaining areas of disagreement, after giving consideration to the input and concerns of both parties. Any such recommendations shall be consistent with the principles set forth above, and shall be filed and served no later than 30 days after the parties have jointly submitted any part(s) of the plan that have been agreed to (or a statement that no such agreement was possible). The parties shall have an opportunity to file any objections to the Special Master’s recommendations within 10 days after such recommendations have been served and filed with the Court.

5.This Court shall retain jurisdiction over this action until such time as the Court is satisfied that all constitutional violations found herein have been fully and effectively remedied.

IT IS SO ORDERED.

APPENDIX A

Glossary of Terms

CAC Criminal Activities Coordinator

CDC California Department of Corrections

Cl Confidential Informant

CMC . California Men’s Colony

CMF California Medical Facility

DOM CDC Department Operations Manual

IAD Internal Affairs Division

ICC Institutional Classification Committee

IGI Institutional Gang Investigator

ISU Investigative Services Unit

MHSB Mental Health Services Branch

MTA Medical Technical Assistant

PBSP Pelican Bay State Prison

RES Reduced Environmental Stimulation

SHU Security Housing Unit

SRB Shooting Review Board

SRT Shooting Review Team

SSU Special Services Unit

UCC Unit Classification Committee

1 . 42 U.S.C. § 1983 provides that "Every person who, under color of any statute, ordinance, regulation, custom, or usage, of any State ... subjects or causes to be subjected, any citizen of the United States or other person within the jurisdiction thereof to the deprivation of any rights, privileges, or immunities secured by the Constitution and laws, shall be liable to the party injured in an action at law, suit in equity, or other proper proceeding for redress.”

2 . On January 5, 1995, the Court was notified by defendants' counsel that three of the defendants have retired or resigned from their positions, and that new officials have been named in their place: Warden Marshall has been succeeded by interim Warden J.S. Stainer; Chief Deputy Warden Peetz has been succeeded by Chief Deputy Warden Robert L. Ayers, and Chief Medical Officer Astorga has been succeeded by interim Chief Medical Officer Dr. David Cooper.

3 . Aside from a limited group of documents, discovery in this action closed on February 26, 1993. See April 21, 1993 Order at 2-4. Accordingly, the evidence presented in this case primarily concerns incidents and events that occurred between December 1989 and February 1993.

4 . Several prison staff admitted to a code of silence problem. For example, one Program Administrator agreed, in his deposition, that a code of silence "frequently'' operates among officers of Pelican Bay, and that this fact can make it difficult, as a supervisor, to determine what really happened during an incident. Helsel Tr. 21-3577 (quoting deposition). Notably, in open court this same administrator was reluctant to testify about the code of silence; after agreeing that he knew "what the code of silence is at Pelican Bay,” he was asked if it operates frequently, to which he would respond only that "I don't know how to answer that." Helsel Tr. 21-3576. The Chief Deputy Warden also acknowledged the code of silence, see Peetz Tr. 19-3243 (Q: “Do you have concerns that there is a code of silence among staff at Pelican Bay?” A: "Yes. I have at times, yes”), as did former Program Administrator Rippetoe. See Rippetoe Depo. at 197-98 (“I believe that there’s always been a code amongst peers to protect each other, okay, whether it be law enforcement official [sic], whether it be, you know — it’s no greater in the prison setting. Yes, I think that there is that.”). Captain Jenkins also acknowledged that the code of silence has hampered his investigations of excessive force at Pelican Bay. Tr. 2-288. Defendant Gomez similarly agreed that lack of candor can impede some investigations at Pelican Bay: "There are people that ... are not forthcoming ... that are not as honest as they should be, and that makes an investigation more difficult to prove.” Tr. 28-4653.

We also observed at trial that prison staff frequently could not recall the identity of other staff whom they testified did or said certain things, although other details were easily recalled. Prison staff also report to internal investigators, with notable frequency, that they had just looked the other way, been distracted by something else, or had their visibility impaired at the moment the alleged misuse of force was said to have occurred. See, e.g., Trial Exh. P-3083 at 79046 ("Officer Bare claimed that he was unable to observe what was happening ... because the helmet he was wearing ... blocked his view.”).

Those who violate the code of silence risk hostility from other prison staff. After Sergeant Cox testified that he witnessed an inmate being hit on the head with the butt of a 38 millimeter gas gun, he was recalled as a witness. He testified that, after his appearance at trial, he had been told by various senior staff (whom he would not name unless ordered by the Court) that he had been a snitch and that he should "watch his back” and that "the administration wasn't very happy with me.” Cox Tr. 18-3015-17. Similarly, Officer Powers made the following comments to an interned investigator: “what do I say; you know the position I’m in ... about ratting off that, how am I going to work here any more.... [1]f an officer 'rats’ on a fellow officer, that officer becomes an outcast.” Trial Exh. P-3084 at 79860. Captain Jenkins also agreed that he has "seen evidence of the situation where an officer reports another officer'and is not too long thereafter reported upon himself,” and that this “has been an issue" at Pelican Bay. Tr. 2-290.

5 . The direct testimony of all experts was submitted by way of declaration, supplemented by two hours of live testimony. The parties were also permitted unlimited live cross examination and redirect examination.

Defendants filed a motion to strike portions of the declaration submitted by one of plaintiffs’ experts, Dr. Craig Haney, asserting that (1) Ms 56-61 should be stricken as inadmissible legal opinion outside Haney’s area of expertise, and (2) that all inmate statements contained in the declaration should be stricken as inadmissible hearsay. With respect to the latter, defendants stipulate that the statements may be admitted to show the basis of Dr. Haney's opinions, but not for the truth of the matters asserted.

We conclude that this motion should be denied. With respect to the inmate statements specified as numbers 1-32 in plaintiffs’ supplemental opposition, such statements are admissible under Fed.R.Evid. 803(3) to show the declar-ant’s then existing mental, emotional or physical condition. With respect to the inmate statements specified as numbers 32-49, plaintiffs have clarified that they are not offering such statements to show the truth of the matter asserted therein, and thus they do not constitute hearsay. Finally, the objection to ¶¶ 56-61 appears moot given that plaintiffs do not rely on the opinions set forth in these paragraphs in their proposed findings of fact and conclusions of law; nor has the Court otherwise considered them in making its findings of fact or conclusions of law.

6 . Charles Fenton worked for the Federal Bureau of Prisons for 27 years prior to his retirement in 1980, and served as Warden of the federal prison at Marion, Illinois (which during Fenton’s tenure housed those inmates posing the greatest security risk in the federal prison system, although ethnic prison gangs had not yet emerged as a significant presence). Fenton also served as warden at the federal maximum security prison at Lewisburg, Pennsylvania, and as the first warden at a new federal prison in Wisconsin. Since 1980, Fenton has served as a consultant for the Massachusetts Department of Corrections. He has also testified as an expert for the defense in prison lawsuits filed in several states. Notably, this is the first , time that Fenton has ever testified on behalf of ah inmate class.

7 . Steve Martin has more than 20 years experience in prison management and policy and holds two degrees in correctional science. He has worked as a correctional officer at a maximum security prison, as chief legal representative for the Texas Department of Corrections, and as Chief of Staff to the Director of the Texas Department of Corrections (the third ranking operational officer in the department). Since leaving the Texas department in 1985, Martin has served as a consultant for prison systems in California, Texas,' Nebraska, and Ohio; he currently serves on the Texas Punishment Standards Commission.

8 . Vince Nathan, a lawyer by training and former law professor, has served as a court-appointed expert and monitor in a number of cases challenging prison conditions. He served as a court monitor in the seminal Ruiz v. Estelle litigation, 503 F.Supp. 1265 (S.D.Tex.1980), aff'd in part and rev'd in part, 679 F.2d 1115 (5th Cir.1982) (involving, inter alia, staff misuse of force in the Texas correctional system), and Guthrie v. Evans, 93 F.R.D. 390 (S.D.Ga.1981) (involving Georgia State Prison). Nathan has also worked as a court-appointed monitor in the Puerto Rico and Michigan prison systems, and as a court-designated expert consultant in litigation involving Rhode Island prisons. He has also done consulting work for the National Institute of Corrections and the New Mexico and Arkansas Departments of Corrections. In Fisher v. Koehler, 692 F.Supp. 1519, 1522 n. 6 (S.D.N.Y.1988), aff'd, 902 F.2d 2 (2d Cir.1990), Judge Lasker observed that Nathan "has the reputation as one of the most knowledgeable experts in his field.”

9 . Daniel McCarthy was employed for almost forty years by the California Department of Corrections before his retirement in 1987. During that time, he worked as the Director of the California Department of Corrections for four years, and as the Warden of California Men's Colony for 12 years. His earlier positions included Deputy Superintendent at the California Men's Colony, Correctional Administrator at the California Correctional Institution in Tehachapi, California, Program Administrator at the California Medical Facility in Vacaville, California, Assistant Departmental Training Officer at Departmental Headquarters, Assistant Departmental Transportation Officer at Departmental Headquarters, In-Service Training Officer and Correctional Sergeant at California Medical Facility, and Correctional Officer at San Quentin.

10 . Larry DuBois has been in the field of corrections for over 20 years. He has previously served as the Warden at the Federal Correctional Institute in Englewood, Colorado, and at a facility in Lexington, Kentucky. Other prior positions include Associate Warden at federal correctional facilities in Talladega, Alabama and Marion, Illinois, and Regional Director for the Federal Bureau of Prisons. After his retirement, and before accepting his present position, he worked for three months assisting a Fulton County, Georgia facility on issues pertaining to overcrowding and staffing.

11 . Dr. Armond Start is an Associate Professor at the University of Wisconsin Medical School, where, in 1991, he founded the National Center for Correctional Health Care Studies. He obtained his M.D. in 1957 from the University of Michigan, and has been "Board Certified" as a pediatrician since 1964. Dr. Start also received an M.P.H. from the University of Oklahoma College of Health in 1977. After twelve years in private practice, Dr. Start served as the Director of the Division of Communicable Disease Control for the Oldahoma State Department of Health from 1975 to 1977. From 1977 to 1983 he served as the Medical Director of the Oklahoma Department of Corrections. At that time, he became the Director of Health Care for the Texas Department of Corrections, then the largest correctional system in the United States.

Dr. Start has reviewed over one hundred jails and prisons over the course of his career. He has acted as a correctional medical consultant or expert witness on numerous occasions. Notably, he was retained by the California Department of Corrections in Gates v. Deukmejian, a case concerning medical care at the California Medical Facility in Vacaville.

12 . Dr. Stuart Grassian has been a faculty member at Harvard Medical School since 1974 and a "Board Certified” psychiatrist since 1979. He obtained his M.D. in 1973 from New York University Medical School. In addition to his teaching, Dr. Grassian maintains a private practice, is the Psychiatric Director at the Melrose Wakefield Hospital Day Treatment Program for Addictions, and serves as a supervising psychiatrist in the Outpatient Department at the New England Memorial Hospital in Stoneham, Massachusetts. From 1977 to 1980 he was also Director of Inpatient Services at a community mental health center, where he was responsible for implementing policies regarding staffing, quality assurance, and supervision of psychiatric residents and other mental health professionals. He had similar responsibilities when he served as Chief of Staff (1991-92) and Director of Adult Inpatient Services (1980-84) at the New England Memorial Hospital. He has also testified as an expert witness in other prison litigation. Dr. Grassian has published two articles on the psychological effects of solitary confinement. Grassian & Friedman, Effects of Sensory Deprivation in Psychiatric Seclusion and Solitary Confinement, 8 Int'l J. of Law & Psychiatry 49 (1986); Grassian, Psychopathological Effects of Solitaiy Confinement, 140 Amer.J. of Psychiatry 1450 (1983).

13 . Dr. Craig William Haney is a Professor of Psychology and Director of the Program in Legal Studies at the University of Santa Cruz, where he has been teaching for the past 16 years. Dr. Haney earned both his Ph.D. in Social Psychology and his J.D. from Stanford University in 1978. He has published over 40 articles and book chapters on topics in law and psychology, including works on the conditions of confinement and the psychological effects of incarceration. Dr. Haney has testified as an expert witness in prison litigation in state and federal courts in California, Washington, and Illinois.

14 . Dr. Jay Harness received his medical degree in 1969 from the University of Michigan Medical School, and has been a “Board Certified" surgeon since 1970. Since 1992, Dr. Harness has been a Professor of Surgery at the University of California, Davis Medical School, and is also Chief of Surgical Oncology at Highland General Hospital in Oakland, California. From 1975 to 1985 Dr. Harness was the Director of the Office of Health Care for the Michigan Department of Corrections. He has served, inter alia, as a member of the American Medical Association's ("AMA's") Advisory Committee to improve medical care at correctional institutions, as the chairman of the AMA's National Consultant Advisory Group on accreditation of jails, and AS Chairman of the Board of Directors for the National Commission on Correctional Health Care. Dr. Harness has also served as a consultant on correctional health care for both the United States Department of Justice and for attorney generals in several states.

15 . Dr. Joel Dvoskin, a clinical psychologist, obtained his Ph.D. from the University of Arizona in 1981. Since 1984, he has directed the Bureau of Forensic Services for the New York State Office of Mental Health. In that position, he has line authority for, among other things, inpatient services at three large forensic hospitals and two regional forensic units, and all mental health services in New York state prisons. Since 1988, he has held the title of Associate Commissioner for Forensic Services in the New York State Office of Mental Health.

Dr. Dvoskin was previously employed, inter alia, as the Acting Executive Director of the Kirby Forensic Psychiatric Center, a new maximum security forensic psychiatric hospital in New York, and by the Arizona State Prison system, where he served as a psychologist, supervising psychologist, and Inmate Management Administrator. He has acted as a consultant to approximately 18 jurisdictions regarding the provision of mental health care to incarcerated persons. He has also worked in prisons and prison hospitals in Massachusetts.

16 . Vince Nathan also undertook extensive preparations. He spent over 300 hours reviewing all available written policies concerning the use of force at Pelican Bay, all Pelican Bay training materials, all control restraint memoranda, all shooting review reports, half of the files of internal investigations on excessive force, more than one hundred incident reports, and hundreds of pages of deposition testimony. Nathan visited Pelican Bay in July, 1993 for two days. On the first day he interviewed 7 inmates. On the second day he toured the facilities and conducted informal interviews with staff, including a sergeant, line officers, and housing unit control booth officers.

17 . The Court does not, of course, mean to imply that all correctional staff at Pelican Bay engage in excessive or unnecessary force for the purpose of causing harm. Indeed, the Court has no doubt that many Pelican Bay personnel are committed to performing their jobs in a conscientious and professional manner, notwithstanding the difficulties and inevitable frustrations of working in a prison setting. We note, for example, that Sergeant Cox testified that a "lot of staff" objected to the practice of chaining inmates in fetal restraints to a stationary object in their cells. “It's not that we were inmate lovers," he explained, but staff "didn't like it ... didn’t think it was right." Tr. 15-2345.

18 . We also note that plaintiffs’ expert Vince Nathan, who also served as an expert witness in Fisher v. Koehler, 692 F.Supp. 1519 (S.D.N.Y.1988), aff 'd, 902 F.2d 2 (2nd Cir.1990), found that the "use of unnecessary and grossly excessive force at [Pelican Bay] far exceeds that which I found at Rikers Island and to which I testified in Fisher v. Koehler." Nathan Deck at 10. Judge Lasker in that case found a pattern of excessive force that he described as "significant and widespread.” Id. at 1532.

19 . Because this case focuses on supervisorial responsibility for prison-wide misuse of force, we have not attempted to resolve or reach final conclusions as to the liability of any particular individual for any particular incident, as would be necessary in the case of an individual claim under 42 U.S.C. § 1983. See Fisher v. Koehler, 692 F.Supp 1519, 1532 (S.D.N.Y.1988), aff'd, 902 F.2d 2 (2nd Cir.1990).

20 . As further discussed in section II(A)(lXa)(4), infra, a cell extraction is the process by which an inmate is forcibly removed from his or her cell.

21 . A 38 millimeter gas gun, similar to a large shotgun in appearance, shoots rubber blocks at high velocity; a taser is an electrical gun that shoots darts up to a range of 15 feet. The darts, connected to the gun by wires, deliver up to 50,000 volts and temporarily incapacitate the victim. See Start Decl., Exh. K at 73.

22 .According to the medical report, the wound on Castillo's head was a seven centimeter by eight centimeter avulsion with a “deep groove under flap [of scalp] that appeared one half centimeter-plus deep, running from [the] frontal to [the] parietal area of the skull." Trial Ex. P-1991 at 23036. Castillo testified that, as a result of the incident, he suffered hearing loss, recurrent headaches, ringing in his ears, and nightmares.

23 . On June 7, 1991, Castillo filed a pro se action under 42 U.S.C. § 1983 alleging that excessive force was used in his January 30, 1991 cell extraction (C91-1745 SBA). The presiding judge dismissed the action on May 17, 1993, after concluding that Castillo's allegations failed to state a claim for excessive force. See May 17, 1993 order at 4-5 and n. 2. As that court only had before it the pleadings of a pro se litigant, and had no opportunity to consider the evidence presented in the instant action, its dismissal of Castillo’s action does not lend any credence to defendants' version of events.

24 . Even if Richard had, in fact, assaulted another officer, the Eighth Amendment does not permit an officer to impose corporal punishment in the form of a broken jaw. Ort v. White, 813 F.2d 318, 324 (11th Cir.1987) (prison officials step over the line of constitutionally permissible conduct if they summarily and maliciously inflict harm in retaliation for past conduct); see also Cal.Penal Code § 673 (making corporal punishment, e.g., the use of physical punishment as a penalty for violating a prison rule, a misdemean- or); Cal.Code of Regs., Tit. 15, § 3281 ("No cruel or corporal punishment of an inmate ... will be permitted for any reason”).

25 . Although Officer Parson testified that he did not participate in the cell extraction, the Court does not find this testimony credible in light of the fact that (1) the incident report lists Parson as a participant, (2) Lieutenant Carl, who supervised the extraction, told investigators during his taped interview that he believed that Parson participated in the extraction, (3) Sergeant Miller, also interviewed on tape, said that he was "sure” that Parson participated in the extraction, (4) the control booth officer on duty stated that he believed that one of the officers who pulled another officer away from Martinez was Parson, and (5) the medical report reflects that Parson informed the Medical Technical Assistant that he was involved in the cell extraction and that some of the mace used hit him in the face. Parson testified that he was only exposed to the mace when he went into the cell after the extraction was over to clean up, at which time he had a reaction to the mace. The Court does not find this explanation persuasive, given that it is unsupported by any other evidence and is contradicted by several other sources who had no motive to be untruthful.

26 . Indeed, there is no indication that Dortch requested a bath over a shower. Rather, because of Dortch’s uncooperativeness, custody staff was unable to successfully dress him, and he arrived at the SHU infirmary essentially nude, except for his restraints and part of a blue isolation gown that was wrapped around his upper torso.

27 . Shortly after the incident, a nurse returned to the tub. The water had been drained but she turned it on to test the temperature at the hottest setting, which was 140 degrees. A device has since been installed that prevents the water from reaching such a high temperature.

28 .According to the evidence in the record, Dortch maintained his silence because he knew that some retaliation would be coming for his having bitten an officer the week before, and he was determined to “take it like a man” and not let them enjoy it. However, there is also some evidence that Dortch passed out in the tub at some point, which would have alerted the officers.

29 . The Program Administrator denied to investigators that Brown had been kicked, although she admitted that Brown had yelled "quit kicking me." Trial Exh. P-3085 at 77601.

30 . One inmate testified that "each time [in fetal restraints], it's very painful, because you in a cramped-up position and you unable to ... relax one way or another because you cramped up, because you — you pulling on your ... arms, your legs. You know, you pulling one way or another, you pulling.” Jones Tr. 3-521-2.

31 .Plaintiffs' expert Vince Nathan similarly described the use of these forms of restraint as "unnecessary, ineffective, excessive, and altogether lacking in any legitimate penological value.” Nathan Decl. at 91.

32 . We note that the governing policy for the SHU instructs officers to leave an inmate in fetal restraints on his mattress on the floor rather than on his bunk “to ensure the inmate will not fall from the bunk and risk injury.” Trial Exh. D-49 at 18308.

33 . Under governing SHU policy, staff were directed to fill out a form in all cases where they used a restraint in a manner that matches the description of the fetal restraint described above. Trial Exh. D-49 at 18182, subsection (K). The record shows that between January 1990 and August 1992, over 170 "Attachment A" forms were filled out. In many of these forms, the staff did not specifically use the phrase “fetal restraints." However, this does not mean that fetal restraints were not in fact used. First, it does not appear that the policy required staff to fill out "Attachment A” unless they were using a restraint that was equivalent to a fetal restraint. Second, neither the policy nor Attachment A ever refer specifically to the term “fetal restraint” or require that such term be used. Even assuming, however, that some of the Attachment A forms did not involve use of fetal restraints, the Court finds that usage of the restraint was substantial between January 1990 and August 1992.

34 .It also permits use of fetal restraints for destruction of properly or any conduct that “may incite other inmates.” Trial Exh. D-49 at 18307.

35 . We also note that the late 1991 date, which corresponds to a tapering off of the use of the fetal restraint, also corresponds to the time period when plaintiffs filed their amended complaint challenging the fetal restraint practice. See Plaintiffs' Amended Complaint, filed November 21, 1991 (referring to fetal restraint as "hog tying" practice).

36 . At the time of trial, Baker had been on medical leave as of March of 1993; it is unclear how long she had been employed at Pelican Bay prior to that time.

37 . For example, in Georgia State Prison, only hands-on force and restraints are used in cell extractions; in Texas, staff are limited to handcuffs, belly chains, leg irons and straight jackets, unless the inmate is armed, in which case the baton, shield and/or mace may also be used. Texas staff may only use tasers if separate requirements for use of lethal force are also met.

38 . All references to the “Martin Declaration” in this section refer to the Amended and Corrected Declaration of Steve J. Martin Re: Issues Regarding Staff Use of Force.”

39 . The routine use of full-scale cell extractions for failure to return meal trays, which drew particular criticism from plaintiffs' experts, stopped some time after February 1992. Defendants testified that the policy was changed after they realized that inmates were not making weapons out of meal trays. However, out of the 70 inmates who were extracted for meal trays in 1990 and 1991, not one of the prison’s Incident Reports indicated that the inmate had broken or tried to break the tray in order to make a weapon. It is also notable that the Incident Reports prepared after cell extractions for meal trays rarely, if ever, noted the location of the meal tray or whether it was ultimately retrieved. Indeed, the record indicates that full scale cell extractions were conducted even where the tray was accessible at the front of the cell and the inmate was at the back. Accordingly, we are not persuaded that the meal tray extractions were primarily motivated by a genuine concern that inmates were converting meal trays into weapons. Similarly, the program administrator instrumental in securing the policy change did not believe that security concerns justified routine cell extractions to retrieve meal trays. See Helsel Tr. 21-3528.

Under the current policy, an inmate is allowed to keep his tray unless it appears that he is attempting to turn it into a weapon. If the tray is still not returned by the time of the night watch, staff will conduct a cell extraction to retrieve the tray. Notably, this new policy has resulted, in far fewer extractions. As one program administrator explained, trays were generally withheld because the inmate perceived some sort of problem, and by having staff talk to them, the situation can usually be resolved short of conducting an extraction.

40 . According to Pelican Bay written policy, a cell extraction should only take place after "attempts to talk to the inmate have failed.” The Court heard considerable testimony regarding how this policy was carried out, and to what extent staff attempted to talk the inmate into compliance or allowed a “cool down” period so that an extraction could be avoided. It is clear that practices varied, from basically "not talking” to inmates to "some talking.” Plainly, however, many cell extractions were performed without fully exploring whether the situation could be resolved through talking or a cool down period. One staff member, David Rickman, testified that he was categorized as a "weak lieutenant” because he was a "communicating” lieutenant. It was his understanding from "some of the cell extraction teams if the lieutenant has to talk to the inmates it's time to extract regardless of the situation.” Rickman Depo. at 55-56.

41 . For example, in the extraction of inmate Taylor on December 22, 1990, staff used the taser three times, discharged a gas gun three times, and used mace once (extraction for refusal to surrender cup, spoon, and excess clothing). In the extraction of inmate White on January 16, 1991, staff used the taser twice, the gas gun twice, and mace once. In the extraction of inmate Fierro on February 3, 1991, staff used the taser twice, the gas gun twice, and mace once. Defendants' expert DuBois stated that he had never been involved in cell extraction that utilized this level of force where the inmate, like Fierro, was unarmed except for a food tray. Tr. 29-4812-13. DuBois also testified that while he did not consider it an objectional practice to use tasers in cell extractions, he had never worked in a prison where the taser was used on a regular basis. Tr. 29-4709-10.

42 . In the 231 cell extractions that occurred between 1991 and May 31, 1993, six inmates had a weapon. One was a razor, while the remainder were primarily soap wrapped in a sock or shirt, or some similar variation. Indeed, in more than a few cases, the inmate was simply sitting in the back of the cell, and not acting in an aggressive manner.

43 . See “Electric Shock Devices and their Effects on the Human Body,” Medical Science Law, Vol. 30, No. 4, at 299 (1990); Hickey v. Reeder, 12 F.3d 754, 757 (8th Cir.1993) (taser inflicts a painful and frightening blow which is "exactly the sort of torment without marks with which the Supreme Court was concerned in [Hudson v. McMillian, 503 U.S. 1, 112 S.Ct. 995, 117 L.Ed.2d 156 (1992)], and which, if inflicted without legitimate reasons, supports the Eighth Amendment’s objective component”). Questions about significant health risks have also led some prisons to preclude its use. Martin Decl at 10; Start Decl. at 122-26. Notably, defendants’ expert McCarthy testified that he is "not a big supporter of tasers.” Tr. 15-2470.

44 . In the extractions in which the taser was used during this period, the gas gun or mace also was used approximately 79% of the time (82 out of 104 extractions). These numbers strongly support the conclusion that if the purpose of the taser was to immobilize the inmate, it manifestly was not accomplishing that purpose. Yet there was testimony from a senior staff member at the prison that in his experience the taser was a "very effective” weapon that disabled almost every inmate on which it was used. Scribner Tr. 7-1169-70. If that is the case, then the repeated use of multiple weapons, in addition to the taser, supports the Court’s conclusion that unnecessary and excessive force was applied in cell extractions.

45 . Defendants suggested that the change in policy was prompted by a September 1991 departmental bulletin which required staff to contact the medical department before using the taser to ensure that inmate was not on psychotropic medication, was not being treated for cardiac arrhythmia or did not have a pacemaker. However, the weight of the testimony shows that it had always been the policy that medical clearance should be obtained prior to using the taser. Thus, the Court is not persuaded that this bulletin fully explains the change in practice. Program Administrator Lopez, for example, offered that the change was a response to officers playing with the tasers and wearing the batteries down. Tr. 14-2247-48.

46 .The Pelican Bay cell extraction policy for the general population prohibits direct shots in ac-knowledgement of this risk: “Great care and conservative judgement [sic] must be used when utilizing these [gas gun] rounds at close distances. Staff must not aim the round directly at the main torso or head, of inmates, as the impact at short distance can severely injure or kill.” Trial Exh. D-43 at 4132.

47 . In one such instance, the medical report stated that inmate Calhoun had sustained eight baton welt marks on his back. Trial Exh. P-3738 at 3217-18. At trial, the officer involved denied that he ever hit Calhoun on his back with the baton, and offered another explanation for the marks: his kneeling on Calhoun's back in order to gain control over Calhoun during an extraction. The Court did not find this testimony or explanation to be credible.

48 . Indeed, the routine use of highly violent cell extractions, for all types of infractions, is difficult to square with the directive in Cal.Code of Regs., Title 15, § 3279 that "[i]n the event the inmate ... refuses to cooperate in a change of location, only the minimum force required to complete the move will be used."

49 .DuBois testified that a mock demonstration he was given at Pelican Bay by a four person cell extraction team was similar to cell extractions in other systems at which he had worked, and that he did not find use of the gas gun "as a distraction," or use of a taser to be objectionable in a cell extraction. Tr. 29-4704-10 (emphasis added).

50 . Notably, the log says "Quiet day up to 1345 [hours], when inmate Dunn got extracted in C-SHU. He liked it so much, they probably did him again by the time, this is written." Trial Exh. P-3044 at 27715.

51 . This practice appears at least in part designed to compensate for the fact that California is ranked 47th in the nation in the number of correctional officers per inmate. Gomez. Tr. 28-4612-13. Defendant's expert McCarthy also hypothesized that "I guess we [in California] were [in] the wild west days there and never got out.” Tr. 15-2539-40.

52 . The firearms used at Pelican Bay are: (1) the Ruger Mini-14 .223 caliber rifle, (2) the Heckler & Koch Model 94 ("H & K 94”) 9 millimeter carbine, using the Glaser Safety Slug, (3) the Smith & Wesson .38 caliber revolver, and (4) the Remington 12-gauge pump shotgun. Firearms were discharged 177 times in 129 incidents between the time the prison opened and September 9, 1993. Of the 177 shots fired, 23 were intended to be for effect (i.e. were fired with the intent to hit a person), 152 were intended to be warning shots, and 2 were accidental. 109 shots were fired outdoors and 68 indoors. Of the 152 warning shots, 13 caused or were alleged to have caused inmate injuries from ricochets or bullet fragments.

53 . The general population yards are also supervised by unarmed staff who patrol the yard on the ground.

54 . For example, control booth officers in the SHU have resorted to semi-automatic rifles to break up an unarmed fight between inmates on the pod because no gas gun or taser or mace is kept in the control booth. Instead, at least through June of 1992, the gas gun was stored in the corridor control, notwithstanding the fact that the SHU operating procedure contemplates that the control booth officer would have access to a gas gun. Similarly, at least as of June of 1992, it was not expected that staff would attempt to break up unarmed inmate fights in SHU pods through use of cell extraction teams before resorting to lethal force.

In one instance, two inmates were fighting in the SHU pod while approximately ten officers, some with batons, stood on the other side of the pod door. However, they were not allowed to intervene absent specific orders from a lieutenant or sergeant, neither of whom had arrived by the time the control booth officer felt action was necessary to stop the fight. Given this, and the absence of a gas gun, the control booth officer testified that his only alternatives for breaking up the unarmed fistfight were to shout warnings or use his firearm. Brodeur Tr. 24-4014-15. As a result, inmate Ashker was shot in the arm. He later developed an aneurysm in the arm, which, because of lack of treatment, ultimately necessitated an airlift for emergency surgery. Start Decl. at 174-175.

Another inmate fistfight a month before trial resulted in an inmate named Jose Padilla being shot and wounded. On this occasion, as previously, the inmates were unarmed and were fighting in the pod, in clear view of the control booth officer. As before, backup officers were waiting at the pod door for the supervisor with a gas gun to. arrive. In this instance, there were three backup officers at the time of the shot and there were three inmates involved in the fight. Once again, the control booth officer did not believe that she could allow the fight to continue, and had no alternative available other than to use her firearm. She therefore shot Mr. Padilla in the hip, causing him to sustain extensive trauma to his hip and necessitating an air evacuation.

55 . The records produced in this action show that three prisoners have been killed, two of whom were not the person at whom the shooting officer was aiming. Another 20 inmates have been hit by bullets or bullet fragments; six of these inmates appear from the prison records to have been uninvolved bystanders, and another three appear not to have been the aggressor. When defendants’ expert reviewed reports from three separate shooting incidents between March and August 1993 (involving six inmates and 10 separate shots), he concluded that none of the reports set forth sufficient facts to justify any of the shots and that, upon receipt, he would have referred the reports to an investigative unit. Trial Ex. P-1267, P-4900, and P-5599; DuBois Tr. 29-4770-71, 74, 75, 79-80.

56 . See Martin Decl. at 17 (“I believe that there is a substantial level of abuse of force that never makes its way into the descriptions of extractions in the incident reports”); id. at 89 ("Apart from cell extractions, inmates and staff come into contact at Pelican Bay on escort, in the dining hall (for general population), and in a variety of other contexts. The documentation of the use of force in these circumstances is not as standardized as is the case in cell extractions, and therefore in my experience it is more likely that the misuse of force will go unreported"). We also observe that while staff are required to fully document "each incident of use of force,” see Title 15, § 3279, there is no clear definition of what constitutes "force” for purposes of this requirement. See note 58, infra.

57 . If anything, that pattern is, in reality, stronger than reflected in the findings here, given the operative code of silence and the fact that the testimony by prison staff often seemed calculated to reveal no more than necessary. Inadequacies in the supervision of the use of force and the investigatory process also obscure the full parameters of the pattern. Certainly, much has transpired at Pelican Bay of which the Court will never know.

58 .Cal.Code Regs. tit. 15, § 3279 provides that: "No employee will use physical force on an inmate or parolee unless it be in the employee’s defense or the defense of others, or unless it is necessary to prevent escape or serious injury to persons or property. In the event the inmate or parolee refuses to cooperate in a change of location, only the minimum force required to complete the move will be used”; see also, Cal.Code Regs., tit. 15, § 3281 ("No cruel or corporal punishment of an inmate or parolee will be permitted for any reason”). In Fisher, 692 F.Supp. at 1551, the court cited, with approval, the opinion of plaintiffs’ expert Nathan that a similar, generalized statement regarding use of force “ ’provide[d] virtually no guidance to a facility manager, let alone a line correctional officer as to the conduct that's expected.’"

Nor do we find a clear definition of force itself. While the DOM attempts to define force, the definition is circular (force refers to force which is force): "Force as used in this section refers to necessary or reasonable force which is defined as that amount of force used to overcome resistance or to accomplish the lawful performance of peace officer duties." DOM, § 55050.3, Trial Exh. D-37. Thus, while staff are required to fully document "each incident of use of force,” see Cal.Code Regs., Tit. 15, § 3279, this vague standard provides little, if any, guidance on when such a report is, in fact, required.

59 . Defendants did not ask their expert, Larry DuBois, to give an opinion regarding the adequacy of the written policies at Pelican Bay.

60 . A comment made by Sergeant Sandor illustrates this point. When asked if it was "standard procedure” to obtain medical clearance before utilizing a taser, he answers "well, that’s standard procedure for me. I can only speak for myself.” Sandor Depo. at 38 (emphasis added). Another officer was asked whether it was proper to use force against an inmate if there was a threat of immediate destruction to state property, but no danger that the inmate would escape. He responded that "It’s always possible. Just depends on who is — what supervisor's calling the shots." Van de Hey Depo. at 110. Indeed, the Court heard much live testimony that made it clear that practices could change as quickly as the "changing of the guard,” notwithstanding whatever the written policy might be. For example, under one Program Administrator, staff did not make any serious attempts to talk to inmates before undertaking a cell extraction, but under another, "both the sergeant and the lieutenant [became] involved in talking with the inmate, attempting to identify what the problem was, and determining if there was a way of resolving the problem before we resorted to force.” Helsel Tr. 21-3528.

61 .For example, when Sergeant Sandor was asked in February 1993 if the taser was the first weapon of choice in a cell extraction, he responded yes, provided that the inmate is not barricaded and is close enough to reach. However, when Program Administrator Lawrence was asked in February 1993 under what circumstances he would permit the use of taser, he responded that he would probably only do so when an inmate had a weapon, like a bat or baton, that he would not relinquish. Lawrence Depo at 119. Similarly, former Associate Warden Bark stated in February 1993 that the taser would be used as the “last resort" piece of equipment, after mace and the 38 millimeter gas gun, although he could not recall any document that set forth that hierarchy. Bark Depo at 74. But then Captain Scribner testified at trial that he viewed the taser as the "weapon of choice.” Scribner Tr. 7'-1228-30. Charles, Ham, the lieutenant in charge of training, had yet a different understanding. He stated that when tasers were used at Pelican Bay, it was only in instances where the inmate had undertaken some kind of overt act of hostility as opposed to resisting passively. Yet he also acknowledged that the training material for officers used language indicating that use of the taser in cell extractions was mandatory. Ham Tr. 20-3423, 3429. •

62 . According to In-Service Training Supervisor Ham, a properly-conducted cell extraction can typically be completed within 15 seconds. Tr. 20-3434.

63 . Title 15, § 3276(14)(b), for example, requires that employees use "the greatest caution and conservative judgment” when carrying and using firearms, and provides that firearms shall only be used when it is absolutely necessary to prevent or stop- escapes, physically assaultive behavior, the taking of hostages, or other disturbances and disorders which present an immediate danger of escape, loss of life, great bodily injury, or damage to a substantial amount of valuable properly. Further, firearms shall be used only as a last resort' unless other options are clearly inappropriate in view of the immediate need to use armed force. Id. DOM § 55050.8 uses similar language, but contains different limits on the use of firearms, providing that firearms may only be used in the defense of oneself or another from death or serious injury, to prevent escape or to kill a dangerous or badly injured animal. The regulations also require that a warning (either a whistle, shout, or warning shot) be given before aiming a shot at a person, unless the -life of a person is in immediate danger. Cal.Code of Reg., tit. 15, § 3276(14)(b)(5).

64 . As a result of the CDC task force study, Pelican Bay adopted a post order that provides for the following alternative before resort to lethal force: "A cell extraction team shall be utilized for those inmates continuing to fight after being ordered to stop. The cell extraction team shall be deployed in accordance with established procedures. In an extreme emergency, in order to prevent the death or great bodily injury of an innocent cell partner, floor staff will vacate the affected section. Floor staff will notify the Control Officer of the immediate need for armed force. The Control Officer will then attempt to control the assault with the use of armed force in compliance with DOM 55050." Trial Exh. D-58 at 52886. Warden Marshall also testified that the current practice is that, in the event of a cell fight, the control booth officer waits until staff arrives, and then when the door is opened, staff goes in and breaks up the fight. Marshall Tr. at 22-3808. Marshall also testified that, in the event lethal force is used, the control booth officer may fire a warning shot or a shot for effect. We note, however, that the post order does not give any specific guidance as to where such warning shots should be fired.

65 . For example, policy changes regarding fetal restraints, staff response to meal tray retentions, and use of tasers have never been memorialized in writing. Of course, practices that are not formalized in written policies are also more susceptible to change. For example, although Chief Deputy Warden Peetz testified that the rule was now that fetal restraints were "completely discontinued," Tr. 20-3256-57, Program Administrator Helsel testified that he was led to believe that he still had "some latitude” or "discretion" in this area. Tr. 21-3566-67.

66 . For example, Program Administrator Lopez was unaware at the time of his deposition in 1993 whether Pelican Bay even had written policies describing when cell extractions are appropriate. Tr. 14-2185-2186. Similarly, Officer Brodeur did not recall seeing the written operational policy for the SHU until the time of his deposition. Tr. 24-4017. Captain Scribner provided another example, stating that when he was assigned as a full-time lieutenant in the SHU, he was not required to read the post orders for his subordinate staff, and could not recall what they said about warning shots. Scribner Depo. at 107-08. He also testified that extractions generally were appropriate only if there was an imminent threat to safely or security, but was unsure what written policy, if any, set forth that standard. Tr. 7-1210-14. Nor could he recall whether there were any written procedures governing the use of fetal restraints. Tr. 7-1237-38.

67 . For example, although the SHU operating procedure contemplates that the control booth officer will have access to a gas gun, control booths do not contain such guns. Specifically, the SHU operating procedure provides that in the event of a violent cell fight in the SHU, the control booth officer is to open the cell door in order to allow the victim to exit the cell into the pod, which the procedure notes, will also "give the gun officer a clear field of fire, if required.” If the fight continues into the pod, and assuming no weapon is observed, "the first weapon of choice should be the 37-mm gun utilizing the single rubber baton round 264-R. Obviously, the firing of this round will be from the control booth, not from the floor. In directing this shot, staff are advised to avoid any direct hits, but rather should ricochet it around.” Trial Ex. D-49 at 18289 (emphasis added). Notwithstanding the above, the only weapon available to officers in the SHU control booths is the H & K 9-millimeter rifle. As a consequence, significant injuries have been inflicted when control booth officers have resorted to using rifles. See note 54, supra. No explanation was offered at trial as to why gas guns have not been stationed in the control booths. Other examples of staff actions which contravened written policy include leaving an inmate in fetal restraints on the bunk instead of on the floor, and using fetal restraints with the hands and ankles tied behind the back instead of in the front.

68 . At the Academy, officers must take several classes (which appear to last from four to eight hours) that address, in whole or part, the use of force: effects of force, weapons defense, side-handle baton, application of restraint gear, constitutional rights, and ethics. Once at Pelican Bay, officers assigned to the SHU are required to take additional classes, which include body searching and escort procedure, cell extraction training, and restraint gear application. There are also additional lesson plans relating to use of force that supervisors can check out and use to provide additional training.

69 . Nathan reviewed only three use of force lesson plans and did not attend any training sessions or review video instruction-tapes.

70 . There is also some anecdotal testimony in the record that points to inadequacies in the training. For example, while officers are taught to use the minimum amount of force necessary, some indicated by their testimony that they could benefit from additional guidance as to how this standard is to be applied. For example, one internal affairs lieutenant explained that the use of force on an inmate in a particular instance was an appropriate (i.e. minimum) amount because it was not “the most force you could apply.” This is clearly not an appropriate interpretation of the minimum force standard. See also Van De Hey Depo. at 112 (officers are supposed to apply "minimum necessary force” principle without aid of more specific guidelines); Bare Depo. at 127 (Q: Under what circumstances have you been trained that it is appropriate to use force against an inmate? A: "I think that is a gray area with [California Department of Corrections”]. Q: "Okay. What is your understanding?” A: "To use any amount of force.” Q: "Any amount of force?” A. "Well, basically, you need to use the amount of force necessary to complete the task at hand”).

71 .The issue of report writing provides one such example. Experts for both sides criticized the incident reports written by officers as unsatisfactorily vague and general. These inadequacies could be attributed to either poor training or a system of supervision that is willing to accept, if not prefer, reports that reveal as little as possible. As discussed infra, the weight of the evidence in this case supports the latter explanation much more strongly than it does the former.

72 . Of course, videotaping not only provides a record of what occurs, but is also likely to have the added effect of discouraging misuses of force. The Warden’s reasons for declining to videotape cell extractions or taser use were spelled out in a memorandum from the Wardens’ Advisory Group, chaired by Warden Marshall. Specifically, the memorandum noted that (1) videotaping could result in the granting of discovery motions in court litigation, (2) the extra time required would be prohibitive, (3) the cell extraction scene is crowded, (4) the public would “simply see five to six staff members applying force to a single individual in confined quarters, giving the appearance of overkill to simply accomplish control and removal,” and (5) buying the equipment and having an additional staff member operate it would involve additional cost. ' Trial Exh. P-4144 at 52974-48.

73 . For example, the reports of two baton men assigned to the cell extractions of two cellmates (inmates Molano and Moreno) utilize identical language except for changing the name of the inmate. When questioned about this at trial, one of the baton men admitted that he had "no explanation" of how that occurred. Owens Tr. 28-4581.

74 . As one officer noted in his deposition testimony, striking an inmate on his back does not assist in any way in restraining the inmate. Consequently, “all you're doing is inflicting punishment." Coleman Depo. at 101.

75 . The baton officer on the extraction team testified that he did not recall ever being asked about the incident by any investigator or prison administrator. Owens Tr. 28-4594. Providing a remarkable insight into the perspective of mid-level supervisors at the prison, the lieutenant who supervised the extraction testified that although he had "no idea how those [markings] were caused,” he saw nothing in the medical report that would have given him any concern. Trujillo Tr. 21-3681-3682.

76 . , See DuBois Tr. 29-4814; McCarthy Tr. 15-2485 ("Q: And if you'd been director and received this [report regarding the cell extraction of Moreno] — A: I probably would have fired somebody. Q: Well, even if you’d been able to review what somebody might have explained for the baton welts, you probably would have sent it to internal affairs [to investigate]. A: Yes.”).

77 . The report of the baton man, Craig Franklin, simply states as follows: "Upon entering the cell, I observed Inmate CALHOUN at the back wall in an aggressive stance. Officer Conway made contact using a shield trying to gain control of the inmate. At this time, CALHOUN kicked me in the left knee using his right foot. The team then took Inmate CALHOUN down onto the lower bunk, gained control and applied mechanical restraints.” Trial Exh. P-4925 at 3208.

78 . Without belaboring the point, we briefly summarize some additional incidents that clearly called for further inquiry by the Warden. A review of the incident report of the extraction of inmate Martinez on October 17, 1994, for example, reveals that: (1) the incident was precipitated when an inmate threw a food tray and spat on an officer; (2) the officer (Parson) is listed as a member of the extraction team; (3) the inmate lost four teeth as a result of the extraction, and (4) there is no indication in the incident report of the inmate receiving a blow to the mouth. Trial Ex. P-1178 at 5581-83; Martin Decl. at 79-80. Yet the supervising lieutenant stated that he was never queried about the incident by his supervisors until five months later when a complaint from Martinez prompted an internal investigation. Martin Decl. at 80; Trial Ex. P-3 111 (tape interview with Carl).

In another incident involving the extraction of four inmates, one of the inmates had a tooth knocked out and another sustained abrasions to the side of his face and displayed sluggish pupil reaction — suggestive of a blow to the head— without any record of how these injuries to the head occurred. Martin Decl. at 52-55. In another extraction of two inmates from their cell, the inmates again ended up with substantial inju-ríes to the head area that are completely unexplained by the incident report: one inmate had a "grossly swollen, contused nose which is out of alignment ... has numerous small lacerations to lower lip and front teeth appeared to be loose"; the other had an “approximately 2 inch laceration to mid parietal area of skull.” Trial Ex. P-1100 at 4121-4122; Martin Decl. at 73-74. In a third incident, a 3-second charge from the taser is supposed to have “temporarily incapacitated” the inmate, yet the medical record reveals nine separate injuries, all to the face and head area and all unexplained: lacerations on the top of his head, the side of his nose, and his upper lip; a number of abrasions on the sides of his neck and the back of his head; and a 2-inch reddened area on the top of his head. Trial Ex. P-1008 at 2201; Martin Decl. at 28-29.

While the above examples focus on incidents that occurred during cell extractions, the record shows the same lack of supervision in other situations. One example about which the Court heard testimony, and which is discussed in greater detail above, resulted in an inmate having his upper arm broken during a cuffing up through port in his cell. Neither defendants nor their experts testified regarding the incident. Plaintiffs’ experts, however, testified that where there is a major injury of this type and there is "the added factor of an injury occurring where under proper correctional practice it should not,” further investigation of the investigation should have been "automatic.” Martin Decl. at 127-128. Similarly, Fenton stated: "I would anticipate that an event like that would be the subject of the most minute examination by all supervisory levels clear to the central office.... I would certainly never want it to happen again. I would certainly take every step in the world to prevent it from happening again and, depending on what the examination and the investigation revealed, might very well take action against the person who did it.” Tr. 5-791. It does not appear, however, that either of the officers involved ever discussed the incident with prison administrators, even after the inmate filed a grievance claiming guard brutality. Rowland Depo. at 32-33; Trial Ex. P-343 at 34721; Martin Decl. at 127-128.

79 . For the period during which he did conduct personal reviews, he does not recall ever seeing an incident report package where the inmate suffered injuries that were unexplained.

80 . A warden may also refer a non-injury shooting incident for review by a SRB. Warden Marshall has utilized this option once in a case where an officer used a .38 revolver to shoot a warning shot inside a housing unit.

81 . Between January and June of 1992, the number of shooting incidents not reviewed at all amounted to half (6 of 12) of all shooting incidents. Nathan Decl. at 65 n. 128.

82 . For example, some reports simply state that the officer was responding to a physical altercation without indicating that the, standard for use of lethal force had been met. See DuBois Tr. 29-4770.

83 . According to the evidence presented, every decision to shoot has been found to be within department policy (although in some instances there has been criticism regarding some aspect of staff conduct, such as the decision to utilize a particular type of firearm.

84 .In one incident reviewed by DuBois, a control booth officer shot an inmate in the hip while he was involved in a fist fight in the pod area. Trial Ex. P-931 at 42716. DuBois found that both the officer's initial warning shot, and the shot that hit the inmate were "questionable,” and that the Shooting Review Board report approving the shootings did not include facts explaining its decision. DuBois Tr. 29-4774-4475. DuBois further observed that when the officer asserted for the first time to the Shooting Review Board that she thought she had seen a weapon, it was, in his opinion, “obvious that this is brought up to cover her behind.” DuBois Tr. 29-4776. He testified that if he had received that SRB report, he would have turned the matter over to his investigative unit. Tr. 29-4776-77.

85 . Investigations into allegations of misuse of force can also be conducted by the Special Services Unit ("SSU") of the California Department of Corrections. This unit generally only handles such investigations when the allegations are "severe ... or newsworthy.” Beckwith Tr. 17-2708.

86 . The majority (68-80 percent) of investigations concern staff misconduct not involving inmates.

87 . There are only three other instances that have been brought to the Court's attention in which Internal Affairs concluded that a correctional officer engaged in improper physical force: one officer was found, based on his own admission, to have taped an inmate's mouth shut, and another was found to have slapped an inmate on the back of the head. In the third incident. Internal Affairs sustained the allegation that the officer had used excessive force (by slamming an inmate's head into a wall) but did not conclude that any specific rules or codes against utilizing excessive force had been violated. The officer was subsequently disciplined only for verbal misconduct.

88 . The evidence indicates that only one investigation at Pelican Bay was ever instigated as a result of a supervisor’s review of an incident report, and that supervisor, Program Administrator Helsel, was not a member of the ISU. That incident, involving an inmate named Marco Perez, was referred to Internal Affairs by Helsel. Trial Exh. P-3094 at 77690. All other investigations about which the Court heard testimony appear to have been initiated by an inmate complaint or by an officer or other eyewitness reporting what they believed to be the possible misuse of force.

89 . This incident is unrelated to the cell extraction of Jesse Calhoun discussed supra.

90 . In a telling exchange at trial, the Captain in charge of the IAD suggested that the Calhoun incident was nothing unusual at Pelican Bay. He implied that Officer Hlebo may have overreacted to the incident because he did not usually work in that area and thus was not "used to" seeing those types of things. The regular staff, he stated, would not have found this incident to be a striking or unusual occurrence. Jenkins Tr. 3-390.

91 . Officer Rader was found to have used excessive force in this incident but was given the lowest possible level of discipline, a letter of reprimand, and only for improper verbal behavior. See section 11(A)(2)(e), infra.

92 . The summary of Cox's statement in the investigative report does not address whether or not Cox could corroborate Nietschke’s report; thus, it is not clear from the report whether Cox was even asked about this point. The summary does state that Cox observed an officer “rushing” Martinez inside the cell, and pulled him off. However, it is clear from Cox’s description that this occurred before Martinez was fully outside the cell and restrained, and thus is unrelated to what Nietsehke reported. Nonetheless, the investigative report appears to use this unrelated incident to conclude that Cox contradicted Nietsehke because Cox “describes the action of the Officer as only an attempt.” Trial Exh. 3083 at 79054-5, 79058.

93 . At closing argument, defendants argued that the MTA’s report supports Parson’s claim that he did not participate in the cell extraction, but instead received his mace-related injuries while cleaning up some time after the extraction. Defendants noted that the report reflects a forty-five minute gap between the time the extraction occurred and the time Parson sought medical treatment, the implication apparently being that a person struck in the face with mace would not wait so long to seek treatment.

However, the Court believes that such a delay would not necessarily be inconsistent with Parson’s having participated in the extraction. The report does not foreclose the possibility, for example, that Parson came into contact with fumes during the extraction but was not struck directly in the face with the chemical spray, making immediate medical assistance unnecessary. Further, the fact that the MTA’s report lists "Lt. Carl, Sgt. Cox, Sgt. Miller, and cell extraction team” as witnesses to the incident in which Parson was injured provides additional strong support for the conclusion that Parson participated in the Martinez extraction. Tr. Exh. P-1178 at 5600. In light of the large amount of evidence which indicates that Parson participated in the extraction, defendants' argument regarding the MTA’s report does not persuade us that we are unjustified in so finding.

94 . At trial, Warden Marshall displayed a similar attitude in discussing the investigation into an incident which left an inmate with a fractured jaw. When asked whether he was satisfied that the "whole story” had been disclosed, he stated, "I believe it was. I think we — I think we know what happened, yes.” Tr. 22-3839. When asked in the next question whether he then knew "why there was blood on the floor in that counselor’s office," he responded, "we haven't figured that out.” Id. In an incident in which significant evidence indicated that an officer had engaged in excessive force, and the inmate ended up in the infirmary for six weeks with a wired jaw, Marshall's evident lack of concern was striking.

95 . A supervisor can also put a "letter of instruction” into an officer's file. Such letters are not considered adverse action; they are removed from the file after one year and have no negative consequences.

96 . Neither defendant Gomez nor anyone acting under his direction questioned the Warden's decision. However, defendants' own expert noted that, if he were in the CDC headquarters in Sacramento, he "would have wanted to know a great deal more about why the charges were dismissed than is actually set forth in the report.” DuBois Tr. 29-4794.

97 . The third incident involved an inmate (Richard) who suffered a fractured jaw as a result of a staff assault. In another example of a questionable IAD investigation, the IAD report found only that the Sergeant involved had slapped Richard's head while he was restrained. Although the Warden initially recommended dismissal of the Officer, and doubted that the Sergeant had been truthful with investigators, he subsequently exercised his discretion to lessen the discipline to 90 days suspension, and then 34 days suspension, because the Officer was a "good officer” with no prior adverse action who was experiencing financial difficulties. Marshall Tr. 22-3734-35, 3837-38. The record also indicates that defendant Gomez had requested that defendant Marshall “take a look at [i]t" and reduce the severity of the adverse action. Id. at 22-3735.

The fourth incident involved a Sergeant who taped an inmate's mouth shut during an escort. According to the Officer involved, the inmate refused to stop talking during an escort of several prisoners, the inmate was speaking Spanish, a language which the Sergeant did not understand, and he was concerned that a situation might develop between the Hispanic and African-American inmates. The State Personnel Board found the demotion proposed by Warden unusually severe (perhaps especially so when viewed in light of the generally tolerant attitude toward use of force at Pelican Bay), and reversed the Warden's decision.

98 . Defendant Marshall, for example, reviews the watch commander's log, notices of unusual occurrence, and incident reports; he receives copies of all Shooting Review Board and Shooting Review Team reports; he reviews Internal Affairs reports; and for much of the time that Pelican Bay has been open, has reviewed all incident report packages. Similarly, defendant Peetz reviews nearly all incident reports, notices of unusual occurrence and the watch commander's log; he commonly reviews incident report packages; he meets daily with the head of Internal Affairs and reviews their reports; he participates in preparing the recommendation for proposed discipline in instances where force was misused. Defendant Gomez and his staff receive and review the incident reports, Shooting Review Board reports, the Internal Affairs reports, and the adverse action packages proposing discipline where misconduct has been found.

99 . For example, inmates are no longer routinely subjected to full-scale cell extractions in the event a meal tray is withheld, tasers are no longer routinely used, and fetal restraints are no longer utilized with any frequency, if at all.

100 . While another inference would be simple incompetence, there is no evidence in the record to support the conclusion that defendants, with lengthy careers in corrections, could not perform their responsibilities in a competent manner.

101 . From the context of his testimony, it is clear that Martin's colloquial reference to “some folks” was not a reference to a few renegade officers but to the management level officials at the prison.

102 .Defendants have argued that Dr. Start's sample was "biased” because he examined records that plaintiffs felt reflected deficiencies in the medical care system. However, plaintiffs assert not that Dr. Start's sample was random, but that it was a focused study which highlighted systemic problems at Pelican Bay.

103 . The qualifications of both Dr. Armond Start and Dr. Jay Harness are set out in section I, supra.

104 . References herein are to Dr. Start's declaration filed on January 14, 1994.

105 . See Marshall Tr. 22-3826.

106 . MTAs are licensed vocational nurses. They are qualified, when directed and supervised by a physician, to administer medications by hypodermic injection, withdraw blood, start and superimpose intravenous fluids, administer tuberculin and other skin tests and various immunizations. Cal.Bus. & Prof.Code §§ 2860.5-2860.7. They are not specifically licensed to perform a triage function although some MTAs may have triage training. Many MTAs at Pelican Bay have worked as corpsmen in the military.

107 . Triage is a sorting system which is used to assign priorities to medical complaints based on urgency and seriousness.

108 . As Dr. Cooper noted, inmates can also call out to custodial staff — properly characterized as "screamfing] for help.” Cooper Tr. 14-2305.

109 .Head MTAs do lead "debriefing" after some emergencies; however, isolated debriefing cannot take the place of regular, structured evaluation of MTA performance.

110 . A 1991 audit of Pelican Bay commented that a written policy "appeared" to sanction MTA-written orders, and that it "appeared” that medications were routinely being renewed by MTAs. Trial Exh. P-3334 at 32538-39. Both of these activities exceed the scope of MTA and LVN licensure as well as competence.

111 . The Court finds Dr. Start’s opinion on the adequacy of medical records much more credible than that of Dr. Harness. Dr. Harness reviewed only eight files and stated that the records were sufficient, but conceded on the stand that one of the eight records was incomplete. Dr. Start, on the other hand, examined well over 100 patient files.

112 . Dr. Harness, the defendants’ medical expert, agreed with the following statement in a budget request prepared in 1992:

[A]dditionaI clerical staff is essential. The current staffing level is inadequate to provide even the normal daily coordination of medical and psychiatric appointments, purging of medical records, reviewing of intake records for compliance including appropriate follow-up, and auditing of records for completeness prior to transferring. It is imperative that all of these tasks be done timely to maintain the continuity of medical care being delivered.

Trial Exh. D-148 at 49178; Harness Tr. 19-3127.

113 .Defendants were formally warned that this practice has “a negative impact upon the care provided to inmate patients” in a 1991 audit by the Department of Corrections. Trial Exh. P-3334 at 32551.

114 . In addition to delays .within the institution, up to one quarter of new inmates arrive at Pelican Bay without their medical records. In 1992 Dr. Astorga sent a memo to Dr. Khoury, describing how records often arrive a week late and how "[t]he results of the lack of this vital information can be disastrous, creating litigation, as well as medical, problems." Trial Exh. P-4225 at 51647.

115 . For instance, Dr. Start described the medical records of inmate Harold Van Horn in his declaration. The chart shows several bouts of starting and discontinuing pharmaceutical treatment of the patient’s seizure disorder, and then shows:

”[0]n 11/14/91 Dr. Johns signs order to discontinue Dilantin and lower bunk precautions (looks like Gordon, MTA, wrote it). On 11/15/91 there is a typed note from Dr. Johns giving inmate lower bunk because 'medical condition.' Which is the correct order? On 11/22/91 the patient complains of seizure, requests medicine (Med. Order Sheet shows he's been getting it 11/4-11/30!) On 12/9/91 and 12/18/91 there are similar requests but no evidence that he did see an M.D.”

Start Decl. at 312 (references to Trial Ex. P-625 omitted). Dr. Start describes the records of another inmate, Glenn Turner, as follows:

"Low white blood cell count noted in M.D. notes. The progress note of 8/27/90 indicates that the physician was concerned about the possibility of diabetes — a chem panel was ordered. Subsequent visits to the physician on 10/1/91 and 10/24/91 [ — ] no mention is made of the chem panel which was ordered to rule out diabetes. On the 10/1/91 visit to the physician the 'patient concerned about chronic low white blood count — desires complete blood count (CBC)[;] also wants to get weighed.' On 10/5/91 the inmate was weighed. On 10/24/91 the patient was seen by the physician, 'Inmate concerned for low white blood count (WBC)-3.3. WBC discussed with patient and he is apparently reassured.'

In my opinion, the three physician notes are in violation of the California Medical Malpractice Act because they do not contain the basic elements expected of a physician encounter. There is no progress note follow up on the chem panel to rule out diabetes. There is no history regarding the reason why the patient was concerned about his WBC. Was the patient woiried about HIV infection? The medical record indicates a history of drug abuse.

Start Decl. at 309-10 (references to Trial Exh. P-623 omitted). These instances are typical of the many analyses of medical charts in the record.

116 . Dr. Start examined 130 charts which were selected because the patient had experienced some health care problem. The sample of these files was blind with respect to screenings; those selected were, if anything, more likely than a completely random sample to have been screened, since those patients had reason to come in contact with the Pelican Bay medical staff. Dr. Start collected data both on the proportion of inmate transfers to the prison in which an intake screening was performed, and on screening of inmates for contagious diseases.

117 . Dr. Start testified that two "fairly typical” examples of incomplete intake histories included "one inmate whose history had no mention of his hypertension, intravenous drug use, ulcers or the presence of a bullet lodged in his spine with nerve injury, and another who had no mention of his being HIV positive or having had extensive neck surgery.” Start Decl. at 19-20.

118 . A positive result shows that the patient has been exposed to TB. Further examination, such as an X-ray, is needed to determine whether the patient has active tuberculosis or only a latent infection.

119 . Retesting of patients who have tested positive before provides no new information and can even be dangerous: a second test can cause the patient's arm to ulcerate, a painful and disfiguring condition. Start Deck at 23.

120 . Defendants argue that until 1993, prison offi- . cials did not by law have authority to force TB testing on prisoners who refused it. However, there has been no evidence presented to support an inference that shortcomings in the 1992 testing were due to prisoners’ refusals to be tested. Nor have defendants presented any evidence that screenings since 1992 have avoided retesting any more effectively or led to more prompt treatment of those who tested positive.

121 . While defendants by law cannot test inmates without their consent, sound medical practice dictates that they at least actively promote HIV screening.

122 . For a description of the effects of seizures, see Dr. Start’s description of the treatment of Tyler Henderson, section 11(B)(2)(f) at p. 1209-1210, infra.

123 .Dr. Start conducted a survey of over 3000 sick call slips to determine the average number of days inmates waited for an appointment. He examined all of the sick call slips submitted in June and July of 1992 for Facility A (590 slips), May 1991 for Facility C (1053 slips), and May through August of 1990 for Facility D (1300 slips). The time period for each facility was chosen randomly.

Each sick call slip was examined for symptoms that Pelican Bay doctors identified as requiring immediate or same-day referral to a physician: cardiac symptoms, significant abdominal pain, and unresolved bleeding. Start Deck at 31: Dr. Start found that the average delay before a physician appointment for cardiac symptoms was 11.27 days; for unresolved bleeding, 12.75 days; for abdominal pain, 16.8 days. Id.

While these figures show abysmal delays in patient treatment, they may not reflect the current state of medical care at Pelican Bay. Access to medical care has improved since 1990, and since nearly half of the slip’s examined are from that year, the data generated by Dr. Start does not represent or indicate current periods of delay.

124 . In addition, access to specialists is limited. Most of the specialists to whom inmates are referred are located up to 90 miles away, and several types of specialists are simply not available in the community surrounding Pelican Bay. Cooper Tr. 14-2261-62.

125 . See section 11(B)(2)(d), supra.

126 . See section 11(B)(2)(b), supra.

127 . When a medical emergency arises, the medical staff informs the watch commander. The watch commander in turn calls an ambulance and assembles a transportation team, which follows the ambulance in order to prevent escape attempts.

128 . Peer review is a process through which physicians confidentially review each others' work by evaluating individual cases and discussing the medical care provided.

129 . The minutes of the committee’s meetings provide unambiguous evidence of the nature of the “review” conducted. For example, the "Orthopedics” section of the minutes entitled “Medical Peer Review Meeting, July 22, 1993” reads in part as follows:

HART C49090 Mr. Hart is referred for evaluation of a symptomatic ganglion cyst, right foot. Repeated aspiration has been performed. Patient of Dr. Winslow. This has been DEFERRED.

CRAPO H17375 Mr. Crapo is referred for evaluation of proven intraarticular chondro-malacia, symptomatic, with locking and chronic recurring pain. Patient of Dr. Cooper. This has been GRANTED.

Trial Exh. D-286 at 42637.

130 . Quality assurance is a review by someone in a supervisory capacity of care delivered in the past to determine whether it meets desired standards. For example, quality assurance may be conducted by reviewing a sample of records to evaluate the outcome of medical care provided; In fact, Dr. Harness stated that the evaluation of records performed by plaintiffs' expert, Dr. Start, was an "excellent” quality assurance analysis. Harness Tr. 19-3119.

131 . McKinsey is the Deputy Director of the Department of Corrections' new Division of Health Care Services. Health care administration has been reorganized and a new Program and Systems Development department will oversee mental health, physical medicine, public health, planning, and quality program units. McKinsey Tr. 26-4287-92.

132 . The unit will also oversee utilization review and examine litigation issues.

133 . Defendants did issue a protocol for infirmary care of diabetics on March 15, 1993 — several weeks after the discovery cutoff for this suit.

134 . Although the inmate consented to an HIV test in May of 1992, he refused to have his blood drawn when the test was about to be performed. There is no record that anyone counseled Evans about the importance of HIV testing or discussed AIDS with him further. See Trial Exh. P-451 at 28944, 29089, 29097.

135 . Defendants’ expert Dr. Harness testified that, although he disagreed with Dr. Start's overall conclusions and his interpretations of a few case studies, "when it comes to Dr. Start’s factual analyses of the medical records, I have very little disagreement with his factual reading of the rec ords and the statements that he has made in his analyses.” Harness Tr. 19-3077.

136 . Dr. Start described in detail the grossly inadequate medical care Dortch received while medical staff busied themselves attempting to minimize the ramifications of the incident. Start Decl. at 37-48.

137 . In a system that is already understaffed, for example, having to decipher disorganized records reduces the amount of time physicians can spend with each patient; if no quality control is in place, no one will stop MTAs from inappropriately "diagnosing” sick inmates as malingerers.

138 . See Harness Tr. 19-3111-3116. Dr. Harness gave protracted and evasive answers to plaintiffs' requests for his professional opinion of the policy. He did, however, conclude that "[i]t's an area that clearly needs improvement. There’s no question about that.” Id. at 3115.

139 . See discussion section H(B)(2)(d), supra.

140 .Dr. Astorga testified that he did not recall reading or receiving a copy of the letter (even though it bears a stamp indicating it was received by his office). Astorga Depo. 868. Even if this were the case, that such an important matter could go unnoticed by the Chief Medical Officer is itself a scathing indictment of the administration and supervision of the medical staff.

141 . The expert qualifications of Doctors Gras-sian, Haney, and Dvoskin are set out in section I, supra.

142 . Page references for Dr. Grassian’s declaration refer to the copy of his declaration filed January 4, 1994.

143 . At the time of this survey, Dr. Khoury held the position of Chief of Medical Services with responsibility for overseeing health care operations for the California Department of Corrections.

144 . A patient who is “psychotic" has a major mental disorder with current symptoms. The diagnosis also includes inmates who are acutely suicidal. A patient who is “psychotic in partial remission” is still psychotic but shows some improvement. Trial Exh. P-3820.

145 . This figure probably substantially underestimated the number of psychotic inmates at Pelican Bay. See Trial Exh. P-4602 at 49198 (memorandum from Warden Marshall to Dr. Khoury noting that estimate of at least 200 active psychotic inmates "does not include those inmates who are otherwise severely or moderately disabled, and those who have, thus far, remained undiagnosed”); Rose Depo. at 28 (figure was an underestimate based on his experience as staff, psychologist at Pelican Bay).

146 .Page references for Dr. Haney's declaration refer to the copy of his declaration filed September 14, 1993.

147 . Dr. Khouiy suggested that budget requests, known as "Budget Change Proposals,” simply represent hyperbolic attempts to secure additional funds and should be given little weight. Even assuming, however, that a budget request is written to highlight, rather than minimize, problems, we decline to assume that Warden Marshall submitted a request that was false in any serious respect. Rather, considering all of the evidence in this case, it is clear that the request was essentially accurate.

Dr. Khouiy also testified that he does not believe that there have ever been any inmates in the SHU who suffer from a serious mental disorder. Khouiy Tr, 10-1579. This testimony is not credible in light of the overwhelming evidence in the record to the contrary. More generally, the Court found Dr. Khouiy's attempts to minimize the mental health needs of inmates at Pelican Bay unconvincing in light of his defensive demeanor and evasive responses to many questions.

148 . Coleman refers to the related case, Coleman v. Wilson, Civ-S-90-0520 LKK (E.D.Cal.), a class action challenging the constitutional adequacy of mental health services at all California prisons, including Pelican Bay, but excluding San Quentin and the California Medical Facility at Vacaville.

149 . In his letter of resignation, Dr. Simonds stated that "I am leaving because in this new position I will not be able to provide safe and adequate services, given the limited resources, and where there exist too many situations in which my license could be place [sic] in jeopardy." Trial Exh. P-3816. Dr. Simonds expressed particular frustration over the fact that inmates often refused to cooperate in physical examinations or other examinations, thus making it difficult for him to practice psychiatry in the manner he had envisioned. Trial Exh. P-4092. At that time, as well as now, Pelican Bay had no procedures for providing involuntary medication to inmates. See section 11(C)(2)(e), infra.

150 . At trial, Dr. Ruggles stated that they are "beginning” to see people in non-crisis contexts as well and that, as of late, he is able to see people "as often as I need to.” Tr. 17-2906-08. While the level of care may have improved since Dr. Ruggles’s deposition, the Court is not convinced that the description he gave at that time does not remain largely accurate.

151 . The therapeutic benefits of psychotropic or anti-psychotic drugs is "well documented.” Washington v. Harper, 494 U.S. 210, 226 n. 9, 229-30, 110 S.Ct. 1028, 1039 n. 9, 1041, 108 L.Ed.2d 178 (1990). There are, however, potentially serious or dangerous side effects including acute dystonia, a severe involuntary spasm of the upper body, tongue, throat or eyes (which is reversible); akathesia, which is characterized by restlessness and inability to sit still; neuroleptic malignant syndrome, a relatively rare condition which causes cardiac dysfunction and possibly death; and tardive dyskinesia, a potentially irreversible neurological disorder which results in involuntary, uncontrollable movements of various muscles, especially around the face. Id.

152 .Notably, a study prepared by Scarlett Karp and Associates for the California Department of Corrections, to which Dvoskin contributed as a consultant for Scarlett Karp, recommended that the mental health staff at Pelican Bay be increased to 38, a figure which he suggested at his deposition was roughly 10 percent over the amount that would be required to provide minimally adequate mental health care. Dvoskin Depo. at 423; Tr. 27-4414.

153 . One of the courses taught at the R.A. McGee Correctional Training Center for new correctional officers is a three hour class entitled "Unusual Inmate Behavior.” Trial Exh. D-327.

154 . Inmates needing inpatient care are generally categorized as either Category I, J, or K inmates. Category I inmates are inmates who, because of serious mental illness, are unable to function within a particular prison setting and require psychiatric inpatient treatment. A determination that an inmate needs to be treated on an inpatient basis might be made “only after crisis intervention steps to try to control the symptomatolo-gy with medication and/or brief psychotherapy." Sheff Tr. 25-4106.

Category J inmates are those who have a residual or chronic mental illness and may have problems functioning in a particular prison setting; they require a supportive rather than acute inpatient treatment. Category K prisoners are those who have similar sorts of problems as Category J prisoners, but whose problems are due to mental retardation.

Clinicians at Pelican Bay can not formally designate an inmate as falling within either Category I, J, or K; that designation can only be made by California Department of Mental Health practitioners at CMF or Atascadero State Hospital. The mental health staff at Pelican Bay can only refer an inmate who they believe should be categorized as Category I, J, or K to CMF or Atas-cadero State Hospital for evaluation and treatment.

155 . For privacy purposes, specific inmates will be referred to by letter rather than by name in this section.

156 . As noted earlier, see note 151, supra, certain psychotropic drugs are extremely potent and may have serious adverse side effects. Thus, while the use of such drugs may reduce symptoms of mental illness, this is not an adequate treatment or long term solution if the mental health staff are of the opinion that the underlying source of the illness is continued exposure to conditions in the SHU.

157 . The nine clinicians currently employed at Pelican Bay see inmates for evaluation after referral in a "psychiatric line," "psychology line” or "clinical social worker line” on certain days of the week in each different section of the prison. The lines operate like a clinic, with the MTA arranging for a set number of inmates to be seen on each line. A clinician sees about eight to twelve inmates on each line, with the psychiatry lines being somewhat longer. Follow-up review and treatment is handled by the "treatment team" for either the SHU or the General Population section of the prison. The treatment team includes the Chief Psychiatrist, any clinicians assigned to a particular housing unit, the MTA assigned to the housing unit, and custody staff including correctional counselors and program administrators, who are the senior managers of the prison housing facilities.

158 . See note 151, supra. As Dr. Grassian also notes, the act of "malingering” may itself be a symptom of mental illness. An inmate "faking” a symptom complained of may well be so ill in another sense that he is desperately seeking help in any way possible. Also, a patient may both be manipulating and at the same time very ill. “Both things can go on in the same individual.... [A] person can be very ill and that could be why they're manipulating.” Grassian Tr. 12-1979. The medical records, however, do not indicate that these possibilities have been considered when an inmate is characterized as a "malingerer.”

159 . Trial Exhibit P-4220, which contains a report on Pelican Bay prepared by a Special Consultant for the California Legislature Joint Committee on Prison Construction and Operations, was circulated to the CDC and discussed at a hearing attended by high-ranking CDC officials.

160 .Dr. Khoury maintained that psychiatric staffing at Pelican Bay was not necessary during the initial phase of its operation because the Department of Corrections did not send more than a handful of mentally ill prisoners to Pelican Bay. However, the overwhelming evidence demonstrates that there was no policy, and certainly no enforced policy, of excluding mentally ill inmates from Pelican Bay. Furthermore, Dr. Khoury's contention completely fails to address the needs of those inmates who may develop mental illness after their transfer to Pelican Bay.

161 . Notably, the Legislative Analyst's Office made the following observations in its analysis of the 1992/93 Budget:

Too Little Planning, Too Much Litigation. Our review indicates that, although the CDC has made some improvements in administration of its medical programs, the programs are too often characterized by a lack of adequate long-term planning, and 'crisis management,' often brought about by litigation. For example, the CDC's justification for two significant budget proposals — restructuring of the existing psychiatric outpatient program and authorization to build a new health care facility at CIW — are based on consent decrees and court orders. ¶ The lack of long-term planning has been apparent over the years.... Trial Exh. P-3958 at 32.

162 . Statewide, there are approximately 2,300 inmates who are held in security housing units, a figure which represents roughly two percent of the California prison population (which totalled 119,000 as of November 1993). Besides the Pelican Bay SHU, defendants also operate smaller SHUs at the New Folsom and Corcoran state prisons.

163 . Earlier release from the SHU may be secured only if the inmate successfully completes what is referred to as a "debriefing process.” This process requires the inmate to furnish detailed information regarding other prison gang members or gang activity. See section 11(F), infra.

164 . As of November 1992, there were 675 prisoners serving determinate terms for disciplinary rule violations, 485 prisoners serving indeterminate terms for prison gang membership or association, 200 prisoners serving indeterminate terms due to concerns regarding their assaultive or disruptive behavior, and 62 prisoners serving indeterminate terms to avoid risks to their own safety. In October 1993, the number of inmates serving indeterminate terms for prison gang affiliation was 625.

165 . For security reasons, inmates are always restrained in handcuffs and/or waist and ankle chains any time they leave the pod area.

166 . Although the record is somewhat unclear, it appears that the religious volunteers conduct their visits while standing outside the cell door. On some occasions, the inmate may be taken to a holding cell in another area which affords greater privacy.

167 . See also Grassian & Friedman, Effects of Sensory Deprivation,in Psychiatric Seclusion and Solitary Confinement, 8 Int’l Journal of Law & Psychiatry 49 (1986); Brodsky and Scoggin, Inmates in Protective Custody: First Data on Emotional Effects, 1 Forensic Reports 267 (1988); Toch, Mosaic of Despair: Human Breakdowns in Prison, Washington, D.C.: American Psychological Association, 1992; Benjamin & Lux, Solitary Confinement as Psychological Punishment, 13 Cal.W.L.Rev. 265, 268-277 and citations therein (1977); Grassian Decl. at 15-18, and bibliography attached thereto; Haney Decl., Appendix D.

168 . Dr. Sheff also testified that he did not attribute the complex he observed to conditions in the SHU; however, he explains that he "did not take a scientific research attitude towards the symptom complex but treated the symptoms as they appeared and treated inmates as individuals, not guiding myself according to whether there was a need to remove the person from a noxious stimulus or hot.” Tr. 25-4115-16.

169 . When asked to focus on the psychological aspects of the environment, defendants’ expert undertook a similar approach. While touring the prison, he specifically tried to find the people who were most in need. It "was not meant to be a random selection survey.” Dvoskin Tr. 27-4369.

170 . For privacy purposes, specific inmates will be referred to by number rather than by name in this section.

171 . See section II(A)(1)(4), supra, for a description of a cell extraction.

172 . Dr. Haney originally interviewed 65 inmates for approximately one hour each; 40 of the 65 inmates were randomly selected while 25 were preselected. Because a high number of prisoners spontaneously complained about various negative psychological and psychiatric symptoms in the SHU, which correlated with published literature discussing extreme forms of social deprivation and psychological trauma, Dr. Haney independently decided to conduct the more systematic random study described above.

173 . The 100 SHU inmates exhibited the following symptoms: talking to self (63%), hallucinations (41%), ruminations (88%), violent fantasies (61%), oversensitivity to stimuli (86%), perceptual distortions (44%), irrational anger (88%), confused thought process (84%), emotional flatness (73%), mood/emotional swings (71%), chronic depression (77%), suicidal thoughts (27%), overall deterioration (67%), and social withdrawal (83%). Haney Decl. at 37, 41.

174 . Dr. Haney hypothesized that the relatively high response to the control question could have been caused by the fact that the lack of activity in the SHU' may have led some inmates to experience circulatory problems, one symptom of which is a tingling sensation in fingers or toes.

175 . Serious forms of mental illness are “frequently characterized by breaks with reality or perceptions of reality that lead the individual to serious disruption of normal functioning if not treated.” Sheff Tr. 24-4117.

176 .Undoubtedly, the lack of adequate mental health care has aggravated the psychiatric problems of many inmates; we are not persuaded, however, that this fully explains the serious psychiatric distress being suffered by certain inmates in the SHU. Rather, in many of the cases addressed by Dr. Grassian, the suffering is attributable, in substantial or significant part, to conditions in the SHU.

177 . Mr. Martin's expert qualifications are described in the introduction to this decision at note 7, supra. To prepare for his testimony on this claim, Martin conducted a review of Pelican Bay’s cell-housing policies, and reviewed hundreds of documents relating to cell fights at Pelican Bay. He also conducted a review of approximately 36 central files of inmates who had been involved in repeated cell fights or who had been involved in cell fights in which a weapon was used or which resulted in serious injury. Defendants did not call an expert witness to testify regarding this claim.

178 . Specifically, section 3337.1(c) provides that "[a]n 'S' suffix may be affixed by a classification committee to the inmate’s custody designation to alert staff of an inmate's need for single cell housing. The classification committee’s decision to affix the ‘S' suffix shall be based on documented evidence that the inmate may not be safely housed in a double cell or dormitory situation.”

179 . For example, an inmate can be given a single cell designation because he is on psychotropic medication, physically impaired, or involved in a "high notoriety" case.

180 . These figures likely underestimate the actual number of cell fights because not all cell fights are officially reported. If a cell fight is not otherwise detected, the victim may not report the fight out of fear of being labeled a "snitch.” Also, all participants in a cell fight are generally issued a "rales violation" which can adversely affect privileges. This further discourages the reporting of undetected cell fights.

181 .These figures include only cell fights that occurred at Pelican Bay. Thus, these inmates may have been involved in cell fights at other prisons if they were incarcerated elsewhere before coming to Pelican Bay.

182 . Prisoners with a score of 0 to 18 points are considered Level I minimum security inmates. Prisoners with point scores of 19 to 27 points are considered Level II, medium security inmates. Prisoners with point scores of 28 to 51 are considered Level III, higher security inmates, and prisoners with point scores of 52 and above are considered Level IV, maximum security prisoners. CaLCode of Regs., Tit. 15, § 3375.1.

183 . The CDC defines gangs as "any ongoing formal or informal organization, association or group of three (3) or more persons, which has a common name or identifying sign or symbol whose members and/or associates engage or have engaged, on behalf of that organization, association or group, in two or more activities which include planning, organizing, threatening, financing, soliciting, or committing unlawful acts or acts of misconduct classified as serious....” CDC DOM § 55070.16. Pertinent portions of the DOM are contained in Trial Exh. D-172.

184 . The sections of title 15 that are referred to herein are contained in Trial Exh. P-4824.

185 . In this opinion, the term "gang affiliation” refers to inmates who have become either members or associates of a prison gang pursuant to DOM § 55070.19.2 and .3.

186 . This policy, however, is not absolute. A small but unspecified number of gang members or associates are permitted to remain in the general population at the discretion of prison officials.

187 . IGIs were previously known as Criminal Activity Coordinators or "CACs."

188 . At least one of the three sources must be a direct link to a validated member, such as "a validated member or former member identifying the inmate/parolee as an associate; correspondence with a validated member; photographed with a validated member; staff or informant observations of being in company with a validated member; identified as an associate by a validated associate who has a documented direct link; etc.” DOM § 55070.19.3.

189 . Pending the SSU's review of the package, the inmate may be transferred to temporary administrative segregation in the SHU. If the in mate is retained in administrative segregation for ten days or more, he is entitled to a hearing before either a classification committee or a classification officer. Cal.Code Regs. tit. 15, § 3338(a), (b); Trial Exh. P-789 at 18059. Retaining an inmate in temporary administrative segregation for more than thirty days requires the approval of a Classification Staff Representative. Cal.Code Regs. tit. 15, § 3335(c)(1).

190 .This is not an uncommon occurrence. For example, SSU Agent Addison testified that packages frequently include items that are actually duplicative, and thus do not count as independent items: "[S]elf-admissions are a common example of that, where whoever is submitting the package will use self-admission on more than one occasion. It will show up like in a probation officer’s report, this individual admits to the probation officer that, Yeah, I’m this. Then they get to the institution and he admits to the counselor. And that will frequently be submitted to us as two separate self-admissions and used as .two separate pieces of documentation. So we’ll count that once.” Addison Depo. 153-54. Another example of a non-acceptable item of evidence is when "somebody's identified somebody as an EME sympathizer.... This inmate says this inmate is an EME sympathizer, and I believe what everybody says, but we don’t allow that kind of designation to help us in validating somebody. 'Sympathizer' is not a term we’ll use.” Addison Depo. 155.

191 . An inmate who has completed the debriefing process must be formally validated by the SSU as a "drop out.” The SSU is also empowered to change an inmate’s validation status to "delete,” which means that there is no longer a sufficient basis for concluding the inmate is a member or associate of a prison gang.

192 . California regulations allow for additional procedural protections, but these are not usually followed, at least in cases involving validated gang members. See Cal.Code Regs. tit. 15, §§ 3336(b), 3341, 3318 (providing that inmate should be permitted the help of a staff assistant or investigative employee if inmate is illiterate or the complexity of the issues impairs the inmate’s ability to collect and present evidence in support of his case); Cal.Code Regs. tit. 15, § 3338(h) (permitting inmate to present evidence and call witnesses, if doing so will not “be unduly hazardous' to the institution safety or correctional goals").

193 . Although California regulations provide that this review may occur every 180 days, see Cal. Code Regs. tit. 15 § 3341.5, the evidence in the record indicates that the routine practice is to conduct UCC reviews every 120 days. There was no evidence that reviews were delayed for 180 days.

194 . In addition to the UCC and ICC reviews, an inmate is entitled to appeal the decision to segregate him in the SHU through the prison grievance process. See Cal.Code Regs. tit. 15. §§ 3084.1, 3084.5.

195 . The Director of the CDC estimates that there are 80,000 prisoners entering the California prison system each year and approximately 70,000 who are released.

196 . Unlike the first theory, plaintiffs' second theory does not depend on proof of a pattern of excessive force; rather, plaintiffs need only show that defendants’ failure to adequately control the use of force, and their deliberate indifference to the risk to inmate safety that such failure poses, resulted in some serious inmate injuries. Plaintiffs have not identified any class-action excessive force case that has premised Eighth Amendment liability on this theory. Rather, evidence that prison administrators have failed to implement adequate use-of-force controls has been used, not to establish an independent basis for liability to the class, but to explain the underlying causes of a pattern of excessive force and to assess the defendants' state of mind. See, e.g., Fisher v. Koehler, 692 F.Supp. 1519, 1551 (S.D.N.Y.1988), aff'd, 902 F.2d 2 (2nd Cir.1990).

We note, however, that case precedent provides general support for such a theory. First, as the Ninth Circuit has ruled, "[pjrison officials have a duty to take reasonable steps to protect inmates from physical abuse" from prison guards. Hoptowit, 682 F.2d at 1250; see also Slakan v. Porter, 737 F.2d 368, 377 (4th Cir.1984), ce rt. denied, 470 U.S. 1035, 105 S.Ct. 1413, 84 L.Ed.2d 796 (1985) (prison administrators had firmly established duty to ensure that weapons were not misused against inmates).

Second, under 42 U.S.C. § 1983, municipalities can be found liable where the failure to train or supervise subordinates evinces deliberate indifference that leads to constitutional deprivations. Canton v. Harris, 489 U.S. 378, 390, 109 S.Ct. 1197, 1205, 103 L.Ed.2d 412 (1989) (“[I]t may happen that ... the need for more or different training is so obvious, and the inadequacy so likely to result in the violation of constitutional rights, that the policy makers of the city can reasonably said to have been deliberately indifferent to the need"); Ybarra v. Reno Thunderbird Mobile Home Village, 723 F.2d 675, 680 (9th Cir.1984). This doctrine of supervisory liability has also been applied in actions against state prison officials. See, e.g., Walker v. Norris, 917 F.2d 1449, 1455-56 (6th Cir.1990) (applying supervisory liability theory in case against Tennessee state prison officials accused of allowing one inmate to kill another, but finding that evidence did not justify overturning directed verdict in their favor); Slakan, 737 F.2d at 370, 372-376 (affirming judgment against prison supervisory officials on ground that they were deliberately indifferent to a known risk of harm, as evidenced by their failure to provide prison guards with adequate training and guidance); LaMarca v. Turner, 995 F.2d 1526 (11th Cir.1993), cert. denied, — U.S. —, 114 S.Ct. 1189, 127 L.Ed.2d 539 (1994) (failure to supervise guards led to impermissible levels of violence between inmates, creating supervisory liability under Eighth Amendment). Of course, in the prison context, a showing of deliberate indifference must satisfy the subjective test enunciated in Farmer, rather than the objective standard contained in Canton. Farmer, — U.S. at —, — - —, 114 S.Ct. at 1979, 1981-82.

Such supervisory liability is premised, not on a theory of respondeat superior, but rather on “a recognition that supervisory indifference or tacit authorization of subordinates’ misconduct may be a causative factor in the constitutional injuries they inflict on those committed to their care.” Slakan, 737 F.2d at 372. Indeed, just as a lack of training and supervision of municipal police officers can lead directly to unjustified and unnecessary injuries, the same is true for guards in state prisons. See, e.g., LaMarca, 995 F.2d at 1533 (failure to enforce procedures for investigating prison rapes "created an atmosphere of tolerance of rape which enhanced the risk that [such] incidents would occur”); Slakan, 737 F.2d at 376 (prison officials’ failure to provide clear policies regarding use of water hoses left matter to guards' unbridled discretion, thus inviting abuses against inmates).

197 . As used in this discussion, the phrase "pattern of excessive force” does not include a "pattern” of instances where properly trained and supervised guards end up using more force than necessary in a good faith effort to restore order. As Whitley and Hudson teach, such incidents, no matter how frequent, are justified by legitimate interests in security. Rather, the phrase “pattern of excessive force” is used here to refer to a pattern of incidents of excessive force that, can not be justified on such grounds.

198 . Defendants also suggest that the maliciousness standard must be applied to defendants in this case in light of LeMaire v. Maass, 12 F.3d 1444 (9th Cir.1993). We disagree.

LeMaire employed the malicious and sadistic test in determining liability for measures that prison administrators had taken in response to a particular inmate's extreme and dangerous conduct. Id. at 1452-53 (LeMaire’s complaint was “levelled at measured practices and sanctions either used in exigent circumstances or imposed with considerable due process and designed to alter LeMaire's manifestly murderous, dangerous, uncivilized, and unsanitary conduct") (emphasis added); see also id. at 1453-54 (noting that use of the malicious and sadistic standard in LeMaire's case is "consonant with" cases from other circuits, each of which involved use of force against a particular disruptive inmate). Thus, LeMaire does not address the state of mind requirement when plaintiffs seek to hold top-ranking prison administrators liable in a class action alleging a pattern of excessive force.

LeMaire also applied the maliciousness test in determining whether certain prison regulations authorizing certain types of force (e.g. four- or five-point restraints) violated the Eighth Amendment. Although such regulations were not adopted under any time pressure or in haste, the court relied on the general principle that prison administrators are accorded wide-ranging deference in the adoption of policies that are needed to preserve internal order, and also that the purpose of the prophylactic regulations at issue was to maintain or restore order. Id. at 1453. The question here, however, is not the facial validity of a prison regulation designed to restore order; indeed, the Court has not ruled on the facial validity of any prison regulation. Rather, the question is whether defendants have a policy of permitting and condoning a pattern of excessive force, and whether that policy is attributable to a culpable state of mind.

199 . See section 11(A)(1)(a), supra.

200 . Notably, in the context of class actions alleging a violation of the Eighth Amendment based on a pattern of inadequate medical care, a plaintiff class may prevail by showing a series of incidents of negligent medical care, none of which would be individually actionable. See, e.g., DeGidio v. Pung, 920 F.2d 525, 532-33 (8th Cir.1990); Todaro v. Ward, 565 F.2d 48, 52 (2nd Cir.1977). For as the case law makes clear, an individual plaintiff can not recover by showing that a defendant provided negligent medical care or committed malpractice in a specific instance. Estelle, 429 U.S. at 106, 97 S.Ct. at 292; Toussaint IV, 801 F.2d at 1111; Franklin v. State Welfare Div., 662 F.2d 1337 (9th Cir.1981). However a plaintiff class can rely on such instances and systemic deficiencies to show that prison administrators are deliberately indifferent to a pattern of inadequate medical care and obtain injunctive relief on a classwide basis. De-Gidio, 920 F.2d at 532-33; Todaro, 565 F.2d at 52 ("And while a single instance of medical care denied or delayed, viewed in isolation, may appear to be the product of mere negligence, repeated examples of such treatment bespeak a deliberate indifference by prison authorities to the agony engendered by haphazard and ill-conceived procedures”); Ramos v. Lamm, 639 F.2d 559, 575 (10th Cir.1980), cert. denied, 450 U.S. 1041, 101 S.Ct. 1759, 68 L.Ed.2d 239 (1981).

201 . Indicia of ''serious” medical need include "[t]he existence of an injury that a reasonable doctor or patient would find important and worthy of comment or treatment; the presence of a medical condition that significantly affects an individual's daily activities; or the existence of chronic and substantial pain....” McGuckin v. Smith 974 F.2d 1050, 1059-1060 (9th Cir.1992). As the evidence shows, plaintiffs, as a class, clearly have serious medical needs. They also have serious mental health needs, in that members of the class suffer from mental disorders and illnesses that go beyond the mere stress or anxiety that is part of the “routine discomfort” of incarceration. See Doty, 37 F.3d at 546 (ailments such as nausea and depressed appetite caused by unresolved family situational stress not serious medical need); Wellman v. Faulkner, 715 F.2d 269, 272 (7th Cir.1983), cert. denied, 468 U.S. 1217, 104 S.Ct. 3587, 82 L.Ed.2d 885 (1984) ("Treatment of the mental disorders of mentally disturbed inmates is a 'serious medical need.’ ”)

202 . See also Hudson, 503 U.S. at 16, 112 S.Ct. at 1004 (Blackmun, J., concurring) (observing that Supreme Court has not limited injury cognizable under the Eighth Amendment to physical injury and noting that "it is not hard to imagine inflictions of psychological harm — without corresponding physical harm — that might prove to be cruel and unusual punishment”); Jordan, 986 F.2d at 1531 (cross-gender body searches, which caused psychological suffering and were unnecessary and wanton, violated the Eighth Amendment); Hoptowit, 682 F.2d at 1253 (prison must provide adequate physical and mental health care); id. at 1257-58 (finding that certain conditions “createfd] an extreme hazard to the physical and mental well-being of the prisoner”) (emphases added); Spain, 600 F.2d at 199) (finding that prisons can not deprive inmates of regular physical exercise because it is important not only to physical, but mental health as well); id. at 200 (observing that the "court's judgment must be informed by current and enlightened scientific opinion as to the conditions necessary to insure good physical and mental health for prisoners”) (emphasis added). But see Newman v. Alabama, 559 F.2d 283, 291 (5th Cir.1977), cert. denied, 438 U.S. 915, 98 S.Ct. 3144, 57 L.Ed.2d 1160 (1978).

203 . This claim is distinct from plaintiffs' claim that the prison has failed to provide adequate mental health care to inmates once they are in need of mental health treatment.

204 . Defendants have presented the Pelican Bay SHU as a centerpiece of their program to decrease violence in the California prison system. However, evidence regarding the SHU's significance is inconclusive. Statistics submitted by the defendants show a consistent rate of decline in violence since 1984, five years before Pelican Bay opened its doors in December of 1989. See Trial Exh. D-80 at C003912 (1992 is "seventh year in a row that the rate of [violent] incidents has decreased”); id. at C003914 (downward trend in assaults on staff since 1984); id. at C003926-7 (the only recent significant drop in deaths per year occurred between 1987 and 1988).

205 . In Gordon, the inmate was exposed to subfreezing weather without hat or gloves for over one hour. The court noted that the conduct at issue was not as harmful as a whipping or electrical shock; nonetheless, the pain inflicted was sufficient to violate the Eighth Amendment given, inter alia, the absence of any legitimate penological interest, and defendant’s callous refusal to provide the hats and gloves although they were readily available. 800 F.Supp. at 800.

206 . Such measures may also have negative effects on security as well. For example, training materials for the CDC Correctional Training Center observe that "isolation of semi-sensoiy deprivation [sic]” and "dehumanizing incarceration” are two factors, among others, that increase violence by inmates. Trial Exh. P-3021. However, absent constitutional violations, it is for prison officials to determine what measures provide the best overall security for the prison.

207 . See also Hutto v. Finney, 437 U.S. 678, 686, 98 S.Ct. 2565, 2571, 57 L.Ed.2d 522 (1978) (length of confinement relevant to deciding whether confinement meets constitutional standards); Young, 960 F.2d at 364 (“The duration and conditions of segregated confinement cannot be ignored in deciding whether such confinement meets constitutional standards.”); Sheley, 833 F.2d at 1429 (although the court has “been hesitant in the past to apply the Eighth Amendment to claims of physical and mental deterioration by prisoners in the general prison population ... [the plaintiff’s] twelve-year confinement in [solitary confinement] raises serious constitutional questions”); Pepperling v. Crist, 678 F.2d 787, 789 (9th Cir.1982) (“The deprivations associated with an institutional lock-up, including twenty-four hour confinement, and curtailment of all association, exercise and normal vocational and educational activity, may constitute a ... violation of the Eighth Amendment, if they persist too long.”).

208 . See, e.g., Smith v. Coughlin, 748 F.2d 783, 787 (2nd Cir.1984) (district court found that plaintiff was not suffering from any psychological damage as a result of conditions of confinement); Sostre v. McGinnis, 442 F.2d 178, 193 n. 24 (2nd Cir.1971) (no evidence of psychological injury to the health of the prisoner), cert. denied, 404, 405 U.S. 1049, 978, 92 S.Ct. 719, 1190, 30, 31 L.Ed.2d 740, 254 (1972); Bruscino v. Carlson, 654 F.Supp. 609, 621 (S.D.Ill.1987) (evidence presented did not support plaintiffs' claims regarding the mental effects of challenged conditions), aff 'd, 854 F.2d 162 (7th Cir.1988), cert. denied, 491 U.S. 907, 109 S.Ct. 3193, 105 L.Ed.2d 701 (1989).

209 . As the Supreme Court has made clear, conditions of confinement may establish an Eighth Amendment violation "in combination," even when each would not do so alone, "when they have a mutually enforcing effect that produces the deprivation of a single, identifiable human need." Wilson, 501 U.S. at 304, 111 S.Ct. at 2327. Thus, even though the challenged conditions in this case may not individually deprive an inmate of any basic human need or violate the Constitution, plaintiffs may attempt to show that, in combination, they deprive plaintiffs of an identifiable human need, which in this case is the inmate’s sanity or mental health.

We also note that defendants have not disputed that the conditions at issue here are more restrictive than those that were in effect at San Quentin prison during the Toussaint litigation. For example, inmates there were permitted the opportunity for group exercise in a larger yard with equipment. Also, "movement throughout the institution was nowhere near as significantly controlled and surveilled as it is in Pelican Bay.” Haney Tr. 6-1053; see also Haney Decl. at 59.

210 . Although Helling involved a risk to an inmate’s physical health, it appears that the principles enunciated would apply in the context of mental health as well. This is consistent with the fact that courts have borrowed standards utilized in analyzing physical health care when evaluating the adequacy of mental health care. See, e.g., Doty, 37 F.3d at 546; Cody, 599 F.Supp. at 1058 (adequacy of mental health care system "is governed by the same constitutional standard which applies when determining the adequacy of a prison's medical ... system”).

211 . To the extent that other inmates not falling within these categories may develop a serious mental illness in response to conditions in the SHU, an adequate mental health care system should provide sufficient monitoring to be alert to such occurrences and have the ability to address them. See section II(A)(C)(2), supra.

212 . Notwithstanding the above authority, defendants make the novel argument that because the overall purpose of the SHU is related to the security of the prison system, the standard set forth in Whitley, 475 U.S. 312, 106 S.Ct. 1078, for evaluating the state of mind in claims of excessive force is applicable here. Thus, although this is a claim challenging the conditions of confinement rather than the use of excessive force, defendants argue, citing LeMaire v. Maass, 12 F.3d 1444 (9th Cir.1993), that no Eighth Amendment violation may be found unless plaintiffs show that defendants acted maliciously and sadistically for the veiy purpose of causing harm.

In LeMaire, the court stated that it would apply the “maliciousness" standard in evaluating specific measures undertaken to control a particular inmate confined in a '‘Disciplinary Segregation Unit” similar in purpose to the SHU. Id. The LeMaire court, however, distinguished that case from "prison condition cases”: "What LeMaire complains of are not so much conditions of confinement or indifference to his medical needs which do not clash with important governmental responsibilities; instead his complaint is leveled at measured practices and sanctions either used in exigent circumstances or imposed with considerable due process, and designed to alter Le-Maire’s manifestly murderous, dangerous [conduct].” Id. at 1452-53 (emphasis added).

Here, plaintiffs are not challenging measures designed to respond to the particular behavior problems posed by a particular inmate. Rather, they are challenging routinized, basic conditions on the ground that they adversely affect serious mental health needs. We are not persuaded that Whitley was intended to stretch so far beyond its borders so as to govern the challenge to conditions presented here.

213 . Once defendants were aware of the risk, they were obligated to consider the risk seriously, and to conduct adequate inquiries. See Farmer, — U.S. at — n. 8, 114 S.Ct. at 1982 n. 8 (defendant may "not escape liability if the evidence showed that he merely refused to verify underlying facts that he strongly suspected to be true, or declined to confirm inference of risk that he strongly suspected to exist”). Cf. Jordan, 986 F.2d at 1529 (prison officials have duty to consider issue carefully and to “afford sufficient weight to the constitutional rights of individuals”).

214 . Although Redman involved Fourteenth Amendment protections, the Ninth Circuit therein explicitly recognized that the Eighth Amendment imposes a duty on prison officials at least as rigorous. Redman, 942 F.2d at 1442-1443.

215 . Some of the inmates who were involved in only one or two fights may have had an established history of assaulting inmates at prior institutions; however, plaintiffs have not identified what percentage of such inmates would fall into this category.

216 . There is evidence in the record that requests for single cells in the SHU are common; however, there was no evidence presented as to whether such requests were prompted by a desire for the privacy offered by a single cell, by a fear of attack from cellmates, or for some other reason or combination of reasons.

217 . We note that plaintiffs have not mounted a constitutional challenge to defendants' debriefing policy. Thus, the Court has not considered or addressed this issue.

218 . The amount of process due in any given case presents a question of law for the Court to decide. Quick v. Jones, 754 F.2d 1521, 1523 (9th Cir.1985).

219 . State "law” may include statutes, codes, and regulations, including prison regulations, official policies and customs. See, e.g., Toussaint IV, 801 F.2d at 1097; Clark v. Brewer, 776 F.2d 226, 230 (8th Cir.1985); Hayward v. Procunier, 629 F.2d 599, 601 (9th Cir.1980), cert. denied, 451 U.S. 937, 101 S.Ct. 2015, 68 L.Ed.2d 323 (1981).

220 . The full text of section 3341.5(c)(3) reads as follows: "Release from SHU. An inmate shall not be retained in SHU beyond the expiration of a determinate term or beyond 11 months, unless the classification committee has determined before such time that continuance in the SHU is required for one of the following reasons: (A) The inmate has an unexpired MERD from SHU. (B) Release of the inmate would severely endanger the lives of inmates or staff, the security of the institution, or the integrity of an investigation into suspected criminal activity or serious misconduct. (C) The inmate has voluntarily requested continued retention in segregation.”

221 . In his Third Special Report, the Monitor noted that “gang membership ... is inherently virtually impossible to ascertain or discover with precision. The gang's only tangible existence is in the minds of the prisoners and prison officials. It is quite unlikely that any two individuals would independently list the same set of persons as members of the group.” Third Special Report of the Monitor at ¶ 33 ("Monitor's Report”).

222 . As the Toussaint Monitor observed, the ICCs “rely heavily upon any conclusion which the CAC has reached concerning the relationship between the prisoner and a gang.... Since at least 1984, the policy ... has been to segregate all ‘known’ prison gang members. Therefore, the CAC’s determination that a prisoner belongs to a prison gang carries with it the virtually inevitable consequence that the ICC will decide that the prisoner must remain in [segregation] ...” Monitor’s Report at ¶ 39.

223 . See also McCollum v. Miller, 695 F.2d 1044, 1049 (7th Cir.1982) ("not all prison inmates who inform on other inmates are telling the truth; some are enacting their own schemes of revenge"); Baker v. Lyles, 904 F.2d 925, 934 (4th Cir.1990) (Sprouse, J., dissenting) ("Well-known is the harshness of inprison 'justice' administered by prisoners against each other, including infamous means for the settling of scores based on jealousies, gang loyalties, and petty grievances. Unfortunately, there also still may be discrete instances wherein guards seek to retaliate against prisoners if they perceive that regular prison procedures will not adequately redress their grievances”).

224 . We would agree, however, as stated in section III(D)(1), supra, that certain conditions in the SHU appear tenuously related to the risks posed by prison gangs. For example, the particularly severe sterility of the environment, the absence of any outside view, and the refusal to provide any recreational equipment in the exercise pen, when combined with the other restrictions imposed in the SHU, appear more punitive than security related. However, as stated above, we can not say that the conditions in the SHU, when taken as a whole, are so extreme in relation to their asserted administrative purpose that we must infer that they are, in fact, imposed primarily for purposes of punishment, with respect to prison gang members.

225 . Given our conclusion that gang affiliates are transferred to the SHU for primarily administrative, rather than punitive, reasons for purposes of fourteenth amendment due process analysis, we do not reach plaintiffs' First Amendment and Eighth Amendment proportionality claims. Of course, the conditions in the SHU must still comport with Eighth Amendment requirements governing conditions of confinement. Plaintiffs' claim regarding the constitutionality of SHU conditions under the Eighth Amendment is addressed in section III(D), supra.

226 . In the Toussaint litigation, the Ninth Circuit held that suspected gang affiliates must have an opportunity to present their views to the prison's Criminal Activities Coordinator (now referred to as the IGI). See Toussaint VI, 926 F.2d at 803, 804-05 (Wiggins, J., concurring). This holding was based on the determination that it was the CAC who "effectively determine^]” gang affiliation. Id. at 805. We note, however, that the Toussaint decisions make no mention of the SSU, thus leaving its role, if any, in that case unclear. In any event, neither the district court nor the Ninth Circuit appear to have considered the necessity of a hearing before the SSU.

227 . According to IGI Briddle, this particular inmate had "104 documents in [his] file saying that he's an EME member. He’s been validated as an EME associate for 15 years. We certainly believe that it's safe to come to the conclusion prior to interviewing the inmate, especially since we know he has EME tattoos on his body, that he is an EME member. And, therefore, in that circumstance, all the documentation was prepared prior to coming into my office.”

228 . Plaintiffs also complain that prison officials do not affirmatively document in an inmate's file when there is an absence of new evidence of gang involvement over some period of time. However, given our ruling above, we are not convinced that this practice has constitutional significance. Moreover, this practice does not mean that prison officials are unaware that an inmate has remained “clean" for some period. Since it is defendants' consistent practice to add any new evidence of gang activity or association to an inmate’s file, the lack thereof necessarily demonstrates that the inmate has a “clean" record for this period.

229 . Plaintiffs have also alleged that defendants have denied inmates the constitutional right of meaningful access to the courts set out in Bounds v. Smith, 430 U.S. 817, 821, 97 S.Ct. 1491, 1494, 52 L.Ed.2d 72 (1977). We defer issuing a decision on this claim so that we can better consider recent Ninth Circuit case law on the subject, including Vandelft v. Moses, 31 F.3d 794 (9th Cir.1994), and Casey v. Lewis, 43 F.3d 1261 (9th Cir.1994).

230 . We also note that defendants’ recent policy changes relating to the use of force do not moot plaintiffs’ request for injunctive relief. In cases involving constitutional violations in prisons, courts have held that "[cjhanges made by defendants after a suit is filed do not remove the necessity for injunctive relief, for practices may be reinstated as swiftly as they were suspended.” Jones v. Diamond, 636 F.2d 1364, 1375 (5th Cir.1981), cert. dismissed, 453 U.S. 950, 102 S.Ct. 27, 69 L.Ed.2d 1033 (1981), and overruled on other grounds by Int’l Woodworkers of America v. Champion Int'l Corp., 790 F.2d 1174 (5th Cir.1986); see also Gluth, 951 F.2d at 1507; Gates v. Collier, 501 F.2d 1291, 1321 (5th Cir.1974); Santiago v. Miles, 774 F.Supp. 775, 793-95 (W.D.N.Y.1991) ("In cases involving challenges to prison practices, federal courts in this circuit have not been reluctant to issue injunctive relief in spite of substantial voluntary improvements by prison officials”); Fisher, 692 F.Supp. at 1565-66. Thus, the burden is on the defendants to prove "that the wrongs of the past could not reasonably be expected to recur.” Jones, 636 F.2d at 1375; LaMarca, 995 F.2d at 1541-42 (reforms enacted after filing of action challenging prison conditions do not preclude injunctive relief unless defendants show that the institution "would not return to its former, unconstitutionally deficient state"); Santiago, 774 F.Supp. at 794. For the reasons discussed in this Court's factual findings, the policy changes that have occurred in the use-of-force area appear not only litigation-inspired but also transitory rather than permanent. We certainly are not persuaded that, absent injunctive relief, the prison would "not return to its former, unconstitutionally deficient state.” LaMarca, 995 F.2d at 1541. Further, the policy changes at issue do not fully remedy the constitutional violations found herein.

231 . Thomas Lonergan previously served as a court-appointed Monitor for this Court, from December 1982 through June 1994, in Cherco v. County of Sonoma, which involved conditions at the Sonoma County Jail in Santa Rosa, California. He discharged his responsibilities as court Monitor in that case with utmost professionalism, integrity, diligence, and sensitivity to the concerns of all participants, and the Court has complete confidence in his ability to serve' effec tively as a Special Master in this case. He has also served as a court-appointed Special Master, Monitor or expert in 8 other cases including Jordan v. Multnomah County (United States District Court, Oregon), and Fischer v. Winter (United States District Court, N.D. California). He was employed by the Los Angeles County Sheriff's Department from 1963 until his retirement in 1980. He has a B.A. and M.A. in Political Science from California State University at Long Beach, and a Masters in Public Administration from Pepperdine University. He is a member of the American Correctional Association and served on its Legal Committee for four years.

232 . With respect to the provision of mental health care, the parties and Special Master should consider whether any remedy should be stayed pending further proceedings in Coleman v. Wilson, Civ S-90-0520 LKK (E.D.Cal). If not, any remedial plan should be carefully tailored to coordinate with proceedings in that action.