MEMORANDUM OPINION

JUSTICE, Senior District Judge.

TABLE OF CONTENTS I. INTRODUCTION 860 A. Case History 862 1 Ruiz v. Estelle 862 2. The 1992 Final Judgment 863 B. The Prison Litigation Reform Act 864 C. Post-1992 Procedural History 865

Counsel for the State of Texas, in an opening statement in this matter, declared that "[u]nder the guidance of this court, and out of a sincere desire to improve its prison system, . . . Texas has transformed its prison policies and practices over the course of the last 20 years."

There can be no doubt that since David Ruiz and the other named plaintiffs began this civil action in 1972 with allegations of unconstitutional practices and conditions in the Texas Department of Corrections' (TDC) prisons, the parties have effected remarkable changes within the prison system. In an epic trial in 1978 and 1979, the plaintiffs' evidence offered a rare glimpse behind the walls that so conveniently shielded free world society from the barbarous living conditions of many of the approximately 25,000 individuals then incarcerated in the TDC prison system. Faced with the staggering magnitude of the constitutional violations found in Texas prisons in 1980, this court regretfully acknowledged that

it is impossible for a written opinion to convey the pernicious conditions and the pain and degradation which ordinary inmates suffer within TDC prison walls-the gruesome experiences of youthful first offenders forcibly raped; the cruel and justifiable fears of inmates, wondering when they will be called upon to defend the next violent assault; the sheer misery, the discomfort, the wholesale loss of privacy for prisoners housed with one, two, or three others in a forty-five foot cell or suffocatingly packed together in a crowded dormitory; the physical suffering and wretched psychological stress which must be endured by those sick or injured who cannot obtain adequate medical care; the sense of abject helplessness felt by inmates arbitrarily sent to solitary confinement or administrative segregation without proper opportunity to defend themselves or to argue their causes; the bitter frustration of inmates prevented from petitioning the courts and other government authorities for relief from perceived injustices.

Truly, much has changed. The Texas prison system, having grown to incarcerate approximately 140,000 inmates in over 100 penal institutions, has instituted a complex web of policies and regulations designed to alleviate many, if not all, of those problems. Today, plaintiffs and defendants alike look back with horror at the way the system used to be. Plaintiffs and defendants alike may also look back with pride at how much the system has changed.

Of course, transformation does not, in and of itself, a constitutional institution make. Although a legitimate source of pride, the prison system's continuing evolution neither proves compliance with the constitution, nor lowers the basic standards of humanity that the constitution demands. To assess the existence of on-going constitutional violations in the Texas prison system, this court must consider not the system's reformation over the last two decades, but that reformation's contemporary product. It must be determined whether the policies, practices, and conditions of the Texas prison system, as they exist today, provide the humanitarian protections afforded inmates by the Constitution of the United States. Inmates, although they do give up many of their constitutional liberties and rights upon incarceration, do not live in a world void of constitutional protections. As the United States Court of Appeals for the Fifth Circuit has reminded us, "[t]here is no iron curtain drawn between the Constitution and the prisons of this country." Ruiz v. Estelle, 679 F.2d 1115, 1126 (5th Cir. 1982).

Prisons, of course, are designed to punish. Institutions of incarceration are an unfortunately necessary infrastructure of our society. And, this court has sentenced more offenders to such institutions than it cares to remember. However, an offender is sentenced to a term of imprisonment; an offender should not, and must not, be sentenced to a term of enslavement by gangs, rape and abuse by predatory inmates, or excessive force by prison employees. To allow such "punishment" would run counter to modern penologists' ambitions of corrective rehabilitation. As this court stated after the original trial in this civil action, such "iniquitous and distressing circumstances are prohibited by the great constitutional principles that no human being, regardless of how disfavored by society, shall be subjected to cruel and unusual punishment or be deprived of the due process of the law within the United States of America." Ruiz, 503 F. Supp. at 1390.

Couched in two motions to terminate its jurisdiction in this civil action, this court has before it, once again, questions of the Texas prison system's constitutionality. After determining that the termination provisions of the Prison Litigation Reform Act violate the separation of powers doctrine and due process clause of the Constitution of the United States, the court has alternatively found the Texas prison system continues to violate inmates' constitutional rights.

It is determined that TDCJ's medical and psychiatric care systems, while at times plagued by negligent and inadequate treatment of members of the plaintiff class, are not so deliberately indifferent to inmates' physical and mental health needs as to be unconstitutional. The extreme deprivations and repressive conditions of confinement of Texas' administrative segregation units, however, have been found to violate the Constitution of the United States' prohibition against cruel and unusual punishment, both as to the plaintiff class generally and to the subclass of mentally ill inmates housed in such confinement. Furthermore, members of the plaintiff class still live under conditions allowing a substantial risk of physical and sexual abuse from other inmates, as well as malicious and sadistic use of force by correctional officers. Despite its institutional awareness of these conditions, TDCJ has failed to take reasonable measures to protect vulnerable inmates from other, predatory prisoners and overzealous, physically aggressive state employees.

These findings do nothing to diminish the respect and approbation deserved by those in TDCJ who work diligently to punish, reform, and rehabilitate society's offenders. Over the course of this litigation, this court has developed a profound appreciation for the formidable task of handling Texas' massive incarcerated population. As the plaintiffs themselves have acknowledged, the Texas prison system, from its policy-makers to its correctional officers, employs many conscientious public servants-individuals who recognize that prisons must not only punish wrong-doers but also attempt to improve them. It is hoped that these same public servants, in light of the evidence produced before this court, recognize that similar improvement is required of the prison system itself.

A. Case History

1. Ruiz v. Estelle

Seeking declaratory and injunctive relief for alleged violations of their constitutional rights, David Ruiz and other inmates brought suit in 1972, pursuant to 42 U.S.C. § 1983, against the Director of TDC. In 1974, Ruiz's civil action was consolidated with a number of others, and class action status was granted to the plaintiffs, who represented all past, present, and future inmates in the Texas Department of Corrections. The inmate plaintiffs, joined by plaintiff-intervenor the United States, alleged TDC conditions and practices violated the Eighth and Fourteenth Amendments to the Constitution of the United States.

Extensive discovery continued through 1977. After several procedural and logistical delays, trial commenced in Houston, Texas, on October 2, 1978, with the undersigned judge sitting by designation in the Southern District of Texas. Trial lasted 159 days. By the end of trial, the court had heard the testimony of 349 witnesses and had received approximately 1,565 exhibits into evidence. As this court noted at the time, the trial "lasted longer than any prison case-and perhaps any civil rights case-in the history of American jurisprudence. In marked contrast to prison cases in other states, the defendant prison officials here refused to concede that any aspect of their operations were unconstitutional, and vigorously contested the allegations of the inmate class on every issue." Ruiz, 503 F. Supp. at 1390.

In its 1980 memorandum opinion, this court made separate factual findings and legal analyses of plaintiffs' claims. Detailed findings of fact were made in the areas of overcrowding, security and supervision, health care (including medical, dental, and psychiatric care, as well as the special needs of some inmates), discipline, access to the courts, and fire and work safety. Based on those factual findings, this court held that the prisons were grossly overcrowded, that sanitation and recreational facilities were wholly inadequate, that health care was inadequate, that hearing procedures for discipline were inadequate, that access to courts was inadequate, and that fire safety and sanitation standards were in violation of state law and the Constitution.

On April 20, 1981, this court issued a consent decree granting comprehensive injunctive relief. See Ruiz, 679 F.2d 1115, Appendix A. The order obligated TDC to reduce, inter alia, the inmate population in each prison unit to alleviate overcrowding, to increase the security guard and support staff, to furnish adequate medical and psychological care, and to bring all living and working environments into compliance with state health and safety standards. Also, a special master was appointed to supervise and monitor the defendants' effectuation of this court's orders. Ruiz, 503 F. Supp. at 1389.

After a series of motions and appeals by defendants, this first phase of Ruiz case finally culminated with the Fifth Circuit's affirmation of this court's determination that TDC imposes cruel and unusual punishment on inmates in its custody as a result of the totality of conditions in its prisons. The Fifth Circuit also affirmed this court's finding that some of TDC's practices deny inmates due process of law. The Fifth circuit further affirmed this court's decision that remedial measures were necessary. The appellate court, however, narrowed the scope of some of the remedies imposed by this court on TDC to bring the system into compliance with the Constitution. Ruiz, 679 F.2d 1115.

2. The 1992 Final Judgment

During the rest of the 1980's, the parties continued to negotiate various remedial measures, at times coming back before this court with stipulations and motions. See Ruiz v. Lynaugh, 811 F.2d 856, 857 (5th Cir. 1987). In March 1990, this court ordered the parties to negotiate a comprehensive settlement of all remedial issues. That agreement was submitted to the court in July of 1992.

After an evidentiary hearing, this court adopted the parties' agreement and issued both a memorandum opinion to that effect, and a separate order approving the judgment, on December 11, 1992. This Final Judgment vacated and replaced numerous detailed orders and compliance plans. It resulted in termination of the court's jurisdiction in nine substantive areas and continuing permanent injunctive orders on eight substantive issues. The remaining substantive areas were staffing, discipline, administrative segregation, use of force, access to courts, crowding, health services, and death row. In that memorandum opinion, this court praised prison system officials for their continued attempts to improve TDCJ-ID:

TDCJ-ID has remade itself into a professionally operated agency whose goals are to achieve the highest standards of correctional excellence.

Equally important, the measures taken by TDCJ-ID officials to meet their constitutional obligations have been memorialized and institutionalized in numerous internal rules and regulations that have replaced this court's orders as the agency's `road map' to success. The court is satisfied that the defendants not only will maintain and implement these rules and regulations, but also will continue to strive to improve on them and their implementation despite the absence in many areas of detailed court orders.

The parties have caused remarkable and palpable changes to occur within TDCJ-ID, and for that the court is grateful.

It is the continued viability of this Final Judgment that is at issue before the court today.

B. The Prison Litigation Reform Act

President Clinton signed the Prison Litigation Reform Act of 1995 ("PLRA") into law on April 26, 1996, as part of an omnibus appropriations bill for fiscal year 1996. Pub.L. No. 104-134, §§ 801-810. According to its proponents, the dual purpose of the legislation was to limit federal-court intervention in the operation of correctional facilities and to curb frivolous inmate litigation.

To that end, the PLRA's amendments to 18 U.S.C. § 3626 include several provisions regarding the "termination of prospective relief." 18 U.S.C. § 3626(b). Two of these provisions are at issue here. The first is the "immediate termination" provision of the PLRA. 18 U.S.C. § 3626(b)(2). This provision calls for immediate termination of any prospective relief in any civil action with respect to prison conditions, if such relief does not contain a finding that the relief is narrowly drawn, extends no further than necessary to correct the violation of the Federal right, and is the least intrusive means necessary to correct the violation of the Federal right. 18 U.S.C. § 3626(b)(2). The procedure for implementing this "immediate termination" provision is set forth in 18 U.S.C. § 3626(e), which provides that the district court shall promptly rule on such motions and calls for an automatic stay of the prospective relief beginning on the thirty-first day after such motion is filed.

Such an "immediate termination" motion, brought pursuant to 18 U.S.C. § 3626(b)(2), is subject, however, to the qualification of § 3626(b)(3). That provision states that relief should not be terminated if such relief remains necessary to correct "a current and ongoing violation of the Federal right, extends no further than necessary to correct the violation of the Federal right, and . . . is narrowly drawn and the least intrusive means to correct the violation." 18 U.S.C. § 3626(b)(3); see Fifth Circuit Order in No. 96-21118, Aug. 6, 1997, at 22-24.

In addition to this "immediate termination" clause, the "two-year termination" provision of the PLRA is now at issue before this court. That provision authorizes termination of a final judgment (that was entered before the passage of the PLRA) two years after the enactment of the statute. 18 U.S.C. § 3626(b)(1)(iii). The "two-year termination," or "(b)(1) provision," is subject to the same caveat as the "immediate termination" provision of (b)(2)-relief should not be terminated if it remains necessary to correct constitutional violations. 18 U.S.C. § 3626(b)(3).

Defendants filed a motion to terminate under the "immediate termination" provision on September 5, 1996. Further, defendants filed a motion to terminate under the "two-year termination" provision on May 5, 1998. Both motions were at issue in the hearing held before this court beginning on January 21, 1999.

C. Post-1992 Procedural History

Although this hearing was held pursuant to the PLRA, the procedural track leading to this decision actually began before that law was enacted. On March 25, 1996, pursuant to Fed.R.Civ.P. 60(b)(5), the state defendants filed a Motion to the Vacate Final Judgment, and thereby sought to nullify their four-year old court-adopted settlement agreement with the plaintiff class. The defendants argued that their "compliance with the Final Judgment, the public's interest and the State of Texas' desire to exercise autonomy over its institutions, mandate that any remaining vestiges of court involvement-however passive-with the prison system, now be vacated." Defendants claimed to operate a constitutional prison system and acknowledged that "no practical effect would be felt by the vacating of the Final Judgment. . . ." ( Def. Mot. to Vacate Final J., Mar. 25, 1996).

One month later, on April 26, 1996, the United States Congress passed the Prison Litigation Reform Act (PLRA). 18 U.S.C. § 3626.

In hopes of assessing the continued necessity, if any, of the 1992 Final Judgment, on May 31, 1996, this court appointed counsel for the plaintiff class and ordered that both parties "shall meet and confer within forty-five days . . . and attempt to (1) narrow the issues in dispute, (2) establish a joint discovery plan, (3) discuss what role, if any, the Special Master in this cause could play in the efficient development of a factual record, and (4) propose to the court an agreed-upon schedule for a hearing on the defendants' motion." By August 8, 1996, the defendants reported back to this court that initial discovery had begun and that the parties would agree on a hearing date once the need for further discovery was assessed.

On September 5, 1996, defendants supplemented their Rule 60(b)(5) motion to vacate the judgment with a motion to terminate under the "immediate termination" provision of the PLRA. 18 U.S.C. § 3626(b)(2). Defendants asserted that this court's 1992 Final Judgment must immediately terminate because it did not contain the findings proscribed in the PLRA-that the relief is narrowly drawn, that it extended no further than necessary to correct the violation of the Federal right at issue, and that it was the least intrusive means necessary to correct the violation of the Federal right. 18 U.S.C. § 3626(b)(2). Plaintiffs served their response to this supplemental motion to vacate on September 20, 1996. Therein, plaintiffs contended that the PLRA cannot be applied retroactively to invalidate the Final Judgment, that the 1992 Final Judgment does, "in substance and effect," make the required findings, that a fact-finding hearing would be necessary to resolve the defendants' motions, that defendants were estopped from invoking the "automatic stay" provisions of the PLRA, and that the PLRA's invalidation of the 1992 Final Judgment would violate the United States Constitution for a number of reasons.

Affirming plaintiffs' concerns regarding the automatic stay provisions' constitutionality, on September 24, 1996, the United States filed its Response to Defendants' Supplemental Motion to Vacate Final Judgment, in which the Justice Department acknowledged that a literal reading of the PLRA would render the automatic stay provision unconstitutional under the separation of powers doctrine. 1996 WL 932104 (S.D.Tex.). The United States suggested that its constitutionality may be preserved by reading the automatic stay provision to allow time for a "court's deliberative processes." Id. at *19. Although it took no position on whether this court should vacate the Final Judgment, the United States did express "some continued concern over issues of medical and mental health care in the Texas prison system, based on numerous complaints [the federal government] [had] received in the last few years." Id. at *1.

On September 25, 1996, this court addressed the "automatic stay" provision's constitutionality. It was determined that "the PLRA `automatic stay' provisions violate the Separation of Powers and due process of law."

Pursuant to the PLRA's mandate that the existence of "ongoing constitutional violations" be considered, this court also ordered discovery to develop a record on the alleged existence of on-going constitutional violations in TDCJ. Further, this court stated that it would "proceed to give due consideration to both of defendants' motions when the parties are ready for a hearing on them."

Defendants then began what would prove to be an extended campaign to avoid such a hearing. On October 24, 1996, defendants moved for a stay of this court's September 25, 1996, discovery order pending the resolution of their appeal before the Fifth Circuit. This court denied that motion on November 14, 1996. On December 31, 1996, the Fifth Circuit denied defendants' motion to stay the discovery. The Fifth Circuit also noted the doubtful appealability of such an order.

Defendants petitioned the Fifth Circuit on January 24, 1997, for the writ of mandamus ordering this court to stay discovery and to decide their motion to terminate without any further investigation of "current or ongoing" constitutional violations. After hearing oral arguments on June 3, 1997, the Fifth Circuit, in a decision dated August 6, 1997, denied defendants' petition for writ of mandamus, and determined that this court's discovery order was unappealable.

Some discovery, in the form of document production and site inspections, began in 1996. In light of defendant's assertion in their recent Motion for Judgment that "[p]laintiffs have had two-and-a-half years to scour the Texas prison system to identify its imperfections," however, it is important to note that throughout 1997, defendants resisted and delayed plaintiffs' discovery in this civil action. During January of 1997, defendants held the position that discovery should not be conducted pending the Fifth Circuit's decision in defendants' appeal. Defendants informed plaintiffs' counsel that neither they, nor their experts and staff, would be allowed to enter any prison facility without an order of this court. ( Decl. of Donna Brorby in Supp. of Mot. to Compel, Oct. 17, 1997.)

During the summer of 1997, defendants resisted discovery in this civil action on the grounds that they were entitled to have their motion for immediate termination of the Final Judgment decided based on the record previously established, without any further discovery at all. ( See Def. Obj. to Request to Permit Entry; Def. Obj. to Request for Prod., July 31, 1997.) Even after the Fifth Circuit denied defendants' appeal and application for the writ of mandamus on August 6, 1997, defendants initially informed plaintiffs that defendants would not produce any more documents or permit inspections without a court order. ( Decl. of Donna Brorby in Supp. of Mot. to Compel, Oct. 17, 1997, at 7-8.) It was in the context of these on-going discovery disputes that it was determined by this court that:

plaintiffs have diligently sought discovery in this matter since the Summer of 1996. Audits of health and treatment conditions in the prison system were arranged in December 1996. Inspections by counsel and experts were scheduled in January 1997. These attempts at discovery, both formal and informal, have been resisted by defendants. As late as July 31, 1997, defendants refused plaintiff's formal requests to produce documents or to permit entry of plaintiff's counsel and experts into the prison system. Defendants object to further discovery in this matter, because they believe the issue can be resolved by review of the documents recently produced by them concerning a number of system-wide statistics.

(Order, Oct. 31, 1997.)

Meanwhile, after this court set the date for a management conference for the fact-finding hearing, defendants moved the court on September 23, 1997, for a "prompt" ruling on their motion to terminate, and again invoked the automatic stay provision of the PLRA. Thus, during the month of October, 1997, defendants were ironically arguing for a "prompt ruling" on their motion while simultaneously resisting the discovery necessary to resolve that motion.

This court issued a discovery order on October 21, 1997, allowing plaintiffs to conduct security and medical site visits as required by the PLRA. This order required that plaintiffs' counsel be allowed to inspect twenty units, with forty-five days total access to those units. The order also required separate access to twenty units by medical and psychiatric experts, in addition to review of the medical records for deceased inmates by those experts and access to all five regional medical facilities.

On December 8, 1997, a management conference was held in which both parties (and prospective intervenors) presented their preliminary legal and factual positions to the court. The disputed necessity for further discovery was discussed.

On January 28, 1998, this court held the automatic stay provision of the PLRA, as amended in 1997, unconstitutional. Also, the defendants' motion for a prompt ruling was granted and defendants' motion to stay this civil action was denied. Defendants gave notice of their appeal of these orders on February 23, 1998.

On May 5, 1998, the second motion currently before this court was filed. Defendants, having filed a motion to terminate pursuant to the "immediate termination" provision of the PLRA on September 5, 1996, then moved to terminate the Final Judgment pursuant to the "two-year" provision ("(b)(1) provision") of the PLRA, 18 U.S.C. § 3626(b)(1)(A)(iii).

Two days later, defendants moved this court to reconsider and vacate its earlier discovery motion. Defendants argued that the security site visits have "taken a tremendous toll of the state's resources, occupying the time and energy of hundreds of people, and requiring the production or inspection of thousands of pages of miscellaneous documentation on site." Defendants further complained that the scope of plaintiffs' discovery was exceeding the scope of the Final Judgment.

In June 1998, the legal and factual arguments of the parties began to take shape in the record. Defendants, in their supplement to their motion to terminate, filed June 17, 1998, provided the court with a collection of affidavits and data purportedly showing the constitutionality of the Texas prison system. At the same time, plaintiffs filed a consolidated memorandum opposing the motion to vacate which included a collection of declarations and exhibits purportedly showing the unconstitutionality of various aspects of the Texas prison system. Plaintiffs also outlined their constitution-based challenges to the PLRA itself.

On July 21, 1998, in its denial of defendants' motion to reconsider and vacate its discovery order, this court reasserted its intention to fulfill its mandate under the PLRA to consider whether or not on-going constitutional violations exist in the Texas prison system. This court, faced with defendants' ironically disingenuous objection to the pace and extent of plaintiffs' discovery inspections, again admonished defendants for their resistance to discovery:

Plaintiffs' efforts at discovery were blocked by defendants for a significant portion of the nineteen months referred to. It was the unfavorable resolution of defendants' own appeal which ended their intransigence as to the plaintiffs' right to conduct discovery. Furthermore, defendants have canceled site inspections and delayed the rescheduling. Defendants have also taken considerable time to produce death records requested by the plaintiffs. Despite these delays, plaintiffs have notified defendants that site inspections should be completed in September, 1998. After examining the declarations of the parties, it is found that the plaintiffs are making a prompt effort to complete discovery as ordered by the United States Court of Appeals for the Fifth Circuit and this court.

(Order, July 21, 1998, p. 2.) In the same order, this court reasserted that "[n]o order will issue on the motion to terminate relief until a factual inquiry to determine its applicability is completed." ( Id. at 1.)

In their objections to that order, defendants again expressed their disapproval of the decision to hold an evidentiary hearing on their motion to terminate, and argued that the issues should be resolved by summary judgment. Defendants supplemented these objections with a number of affidavits rebutting plaintiffs' allegations concerning the constitutionality of the Texas prison system.

The defendants next petitioned the Fifth Circuit on October 28, 1998, for the writ of mandamus compelling this court (1) to rule immediately on its pending motion to terminate on-going prospective relief and (2) to terminate extra-constitutional aspects of the 1992 Final Judgment. On November 4, 1998, this court set the fact-finding hearing on defendants motions to terminate for January 21, 1999. Although the Fifth Circuit expressly denied the petition for writ of mandamus, it ordered this court to act within thirty-one days of the evidentiary hearing set for January 21, 1999, and to rule on defendants' motions not later than March 1, 1999. On November 20, 1998, the Fifth Circuit ordered this court to grant defendant-intervenors' motion to intervene.

Thus, this court arrived at a fact-finding hearing, albeit necessarily truncated, to consider allegations of on-going constitutional violations in the Texas prison system. It proved regrettable that an issue as complex and massive as the constitutionality of various policies and practices of a system incarcerating approximately 140,000 inmates had to be truncated and compressed into a few weeks. In order to comply with the Fifth Circuit's March 1, 1999, deadline, this court limited each party to fifty hours of testimony. Testimony began on January 21, 1999, and ended on February 12, 1999.

II. SUMMARY AND CLARIFICATION OF ISSUES

A. The Parties' Positions

The essence of defendants' position is that the PLRA, 18 U.S.C. § 3626, requires termination of this court's involvement with the Texas prison system. As outlined in the joint pre-trial order submitted to this court on January 5, 1999, defendants contend that they are entitled to relief from the Final Judgment under 18 U.S.C. § 3626(b)(1) and (b)(2) for the following reasons: (1) the 1992 Final Judgment does not comply with the requirements of 18 U.S.C. § 3626(a)(1) and (b)(3), in that it does not contain findings of any ongoing or current violation of a Federal right; (2) the 1992 Final Judgment contains extra-constitutional relief; (3) the plaintiffs have not established the existence of current and ongoing violations of a Federal right; and (4) the prospective relief contained in the 1992 Final Judgment is not limited to the end that it extend no further than necessary to correct the violations of a Federal right and is not limited so as to be narrowly drawn and the least intrusive means to correct the violation of a Federal right. Defendants further argue that the absence of a finding of a current and ongoing violation of a Federal right in the Final Judgment causes it to be barred by the jurisdictional limits of the Tenth and Eleventh Amendments of the Constitution of the United States. Finally, defendants reject plaintiffs' contentions that the termination provisions of the PLRA are impermissibly retroactive, that defendants have waived their ability to seek the 1992 Final Judgment's termination, and that the PLRA violates the Constitution of the United States.

The gravamen of plaintiffs' position is that the presence of ongoing constitutional violations in the Texas prison system requires continued oversight by this court. Plaintiffs begin their contentions with legal objections to the PLRA itself. Plaintiffs argue that the PLRA does not, and may not, apply retroactively to the 1992 Final Judgment. Were it construed to be applicable to that order, plaintiffs maintain, the PLRA would be unconstitutional as a violation of the separation of powers doctrine, as a denial of due process, and as a deprivation of equal protection.

If the PLRA is constitutional and applicable to this civil action, plaintiffs insist that they are entitled under the PLRA to continuing prospective relief that "remains necessary to correct a current and ongoing violation of [their] Federal right[s], extends no further than necessary to correct the violation[s] of the[ir] Federal right[s] and . . . that is narrowly drawn and the least intrusive means to correct the violations." 18 U.S.C. § 3626(b)(3). More specifically, defendants are further accused by plaintiffs of deliberate indifference to prisoners' serious physical and mental health care needs. Plaintiffs also complain that defendants deprive administratively segregated prisoners of the minimum necessities of life and punishment of inmates without due process. Defendants are also accused by plaintiffs of failing to meet their Eighth Amendment duty to ensure prisoners reasonable safety from other prisoners and to curtail malicious and sadistic uses of force by correctional officers. Under the plaintiffs' theory, crowding and understaffing prevent defendants from curing these constitutional violations.

B. The Proper Scope of this Decision

Defendants have consistently argued that the issues before the court in this matter should be limited to those outlined in the 1992 Final Judgment. Defendants point out that, as a result of the 1992 Final Judgment, this court lifted all injunctive relief relating to the following areas: classification, maintenance of facilities, Spanish-English interpreters, provision of necessities, major structural deficiencies, the special master, programs for physically handicapped inmates, programmatic and recreational activities, monitoring by plaintiffs' counsel, work health and safety, and mentally retarded offenders programs. The 1992 Final Judgment retained injunctive relief in the following areas: staffing, access to courts, health services, support services inmates, contact visitation, psychiatric services, discipline, crowding, death row, administrative segregation, internal monitoring and enforcement, and use of force. As defendants acknowledge, these are areas in which plaintiffs could properly offer testimony in the fact-finding hearing. This court wholeheartedly agrees with, and accepts, the Fifth Circuit's admonition that the district court's decision

should not entail a general overall examination of the prison system, but should simply be focused on those continuing injunctive orders (concerning the `eight substantive issues') contained in the December 1992 judgment. Ruling on the motion is not an occasion to examine other areas of prison conditions or practices, neither the "nine substantive areas" as to which the December 1992 judgment "resulted in complete relief from judgment and termination of the court's jurisdiction" nor areas not dealt with one way or the other in the December 1992 judgment.

(5th Cir. Order, Aug. 6, 1997.) After reviewing the sum total of the evidence presented to the court, it is determined that plaintiffs have in fact limited their presentation of evidence to the areas of injunctive relief remaining after the 1992 Final Judgment.

In light of defendants' contentions, however, the propriety of one area of testimony merits further clarification. Defendants have urged the court to disregard evidence concerning TDCJ's unit classification committees on the grounds that the 1992 final judgment contains no remaining relief relating to classification. In their Motion for Judgment, defendants state:

Plaintiffs have put on numerous inmate witnesses who have complained that their requests to the unit classification committees (UCCs) for transfer or reclassification to safekeeping or protective custody were improperly denied, subjecting them to alleged inmate assaults. Plaintiffs' experts have also testified that these alleged improper classification decisions may not have occurred but for the loss of 200-300 classification case manager positions in the 1995 budget.

( Def. Mot. For J., Feb. 4, 1999, at 7.)

Defendants are correct to point out that orders regarding the prisoner classification system were vacated with the 1992 Final Judgment. This court was led to believe that the classification system in place at that time would remain in place in some form or another, and made factual findings to that effect:

The prisoner classification system developed by the Institutional Division of the Texas Department of Criminal Justice ("TDCJ-ID") has become an important and entrenched part of the operation of TDCJ-ID. James A. Collins, Director of the TDCJ-ID, testified that, like security staffing, classification is one of the primary ingredients for operating a prison system successfully. The thrust of his testimony was that TDCJ-ID intends to maintain its current classification system, although that system undoubtedly will evolve over time as conditions change and prison managers improve and redefine it.

( 1992 Mem. Op.) It was found that TDCJ had, in actuality, greatly improved its classification system, which at the time of the original trial in this case focused exclusively on age, size, and the number of prior incarcerations. By the time of the settlement negotiations in the early 1990's, the classification system provided prison managers with detailed information on prisoners and established reasonable criteria for grouping prisoners based on their needs, security and custody levels, medical condition, and relevant physical characteristics. In light of those circumstances, it was concluded that "TDCJ-ID is likely to continue to employ this system of classification, or one equally appropriate for the operation of safe and secure prisons." ( 1992 Mem. Op.).

Due to circumstances beyond prison officials' control, TDCJ's system of classification of prisoners has apparently changed dramatically in recent years. The classification system contemplated by the 1992 Final Judgment effectively no longer exists. In 1995, there were 328 unit classification staff, with a ratio of prisoners to unit classification staff of 233 to one. That is, each unit classification officer was responsible for maintaining safe and appropriate placement for 233 inmates. Today, there are approximately 60 such classification staff. Each unit classification officer is responsible for approximately 1,250 inmates' placement. Pl.Ex. 11.

As defendants point out, the simple fact that the structure of the classification has changed, even if to the detriment of inmates' interests, does not reopen those orders and agreements regarding classification that were vacated by the 1992 Final Judgment. Defendants have correctly noted that "classification," per se, is not an issue properly before the court.

Defendants' further assertion that plaintiffs' testimony regarding the improper denial of their requests for safekeeping or protective custody should be ignored by this court is, however, incorrect. In making this claim, defendants have mistakenly attempted to pigeon-hole large portions of plaintiffs' evidence into the category of "classification" when such evidence is actually appropriately viewed as evidence related to "crowding" and "staffing"-areas in which the court has in fact retained injunctive relief. The various titles for the viable issues in the 1992 Final Judgment are not rigidly exclusive. It is possible, and highly probable, that much of the testimony in this matter, such as that regarding inmate requests for protection, could fall under several of those headings. Over-crowding and understaffing, in the event such conditions were found, could very easily cause defendants to fail and refuse to consider and act reasonably on information inmates give them about substantial risks of serious harm they or other inmates face. In that event, the constitutionality of that condition would have to be assessed.

Furthermore, plaintiffs have not put the classification system, in itself, directly at issue. Rather, the loss of the two to three hundred classification case managers has been offered as a rational explanation for the alleged failure of the prison system to provide reasonable safety to inmates. For these reasons, defendants' contention that "failure to protect claims are not properly at issue in this hearing" is without merit.

III. LEGAL AND CONSTITUTIONAL CHALLENGES TO THE PLRA

Plaintiffs first oppose termination of the consent decree with a series of legal contentions aimed at removing the 1992 Final Judgment from the scope of the PLRA. First, they argue that the PLRA does not apply retroactively to consent decrees issued before the statute's date of passage. The plaintiffs go on to assert that if the PLRA does, in fact, apply retroactively-and thereby applies to the 1992 Final Judgment in this civil action-it violates the separation of powers doctrine, due process, and equal protection under the Constitution. Defendants contest all of plaintiffs' legal arguments.

A. Retroactivity

Plaintiffs' threshold assertion-that the PLRA does not apply to court orders signed before its enactment-is without merit because it is contradicted by the plain language of the statute. The PLRA expressly addresses the orders "issued on or before the date of enactment of the Prison Litigation Reform Act . . ." 18 U.S.C. § 3626(b)(1)(iii). Plaintiffs cite the Supreme Court's decision in Landgraf v. USI Film Products as creating a presumption against retroactive legislation. 511 U.S. 244, 114 S.Ct. 1483, 128 L.Ed.2d 229 (1994). That presumption, however, has no bearing here. In determining whether a statute will be applied retroactively, a "court's first task is to determine whether Congress has expressly prescribed the statute's proper reach." Landgraf, 511 U.S. at 280, 114 S.Ct. 1483. Congress has done so in the PLRA. Plaintiffs' argument fails, therefore, as the termination provisions of the PLRA at issue here are expressly retroactive. It is precisely this retroactivity, however, that gives additional substance to plaintiffs' next legal argument.

B. Separation of Powers

The crux of plaintiffs' separation of powers claim concerns the effect of the legislatively-enacted PLRA termination provisions on the judicially-entered 1992 Final Judgment in this civil action. Under the separation of powers principle, final judgments of the judicial branch are protected from legislative interference. Plaut v. Spendthrift Farm, Inc., 514 U.S. 211, 227-28, 115 S.Ct. 1447, 131 L.Ed.2d 328 (1995). Congress may neither reopen a court's final judgment, id. at 227, 115 S.Ct. 1447, nor prescribe the outcome of a decision in a particular civil action without changing the underlying substantive law applicable to the case, U.S. v. Klein, 13 Wall. 128, 80 U.S. 128, 20 L.Ed. 519 (1871). Because the PLRA could overturn, by "immediate termination," the 1992 Final Judgment of this court, the statute's constitutionality is clearly called into question.

The final judgment at issue here is a consent decree that developed as a settlement of plaintiffs' previously validated claims of unconstitutional practices and conditions in the Texas prison system. See Ruiz, 503 F. Supp. 1265. Consent decrees are a common judicial remedy in prison litigation. See, e.g., Carty v. Farrelly, 957 F. Supp. 727 (D.Virgin Islands 1997); Hadix v. Johnson, 933 F. Supp. 1360 (E.D.Mich. 1996), rev'd, 144 F.3d 925 (6th Cir. 1998); United States v. Michigan, 989 F. Supp. 853 (W.D.Mich. 1996); Benjamin v. Jacobson, 935 F. Supp. 332, 342 (S.D.N.Y. 1996), aff'd in part and rev'd in part, 124 F.3d 162 (2d Cir. 1997). Consent decrees are essentially court-adopted settlement agreements. Prison officials and prisoners often negotiate solutions to prisoners' complaints in order to avoid not only litigation, but also the potential that a court will adjudicate that state prisons are being operated unconstitutionally. If the presiding judge approves the agreement, he or she executes it, so that it may be enforced as a judicial order. Consent decrees, therefore, offer defendants a means of avoiding litigation without admitting wrongdoing. Just such a settlement process occurred in this civil action. The 1992 Final Judgment targeted by the PLRA represents this court's adoption of, as a final judgment, the parties' own negotiations in Ruiz v. Estelle.

1. Separation of Powers Challenges to the PLRA

At least six times, federal district courts have determined that the "immediate termination" provision of the PLRA is unconstitutional as a violation of the separation of powers doctrine. Denike v. Fauver, 3 F. Supp.2d 540 (D.N.J. 1998); Gavin v. Ray, 1996 WL 622556 (S.D.Iowa 1996), rev'd sub nom. Gavin v. Branstad, 122 F.3d 1081 (8th Cir. 1997); Hadix v. Johnson, 947 F. Supp. 1100 (E.D.Mich. 1996), rev'd, 133 F.3d 940 (6th Cir. 1998); Inmates of the Suffolk Co. Jail v. Sheriff of Suffolk Co., 952 F. Supp. 869, 882 (D.Mass. 1997) ( alternative holding); McClendon v. Albuquerque, 29 F. Supp.2d 1267 (D.N.M. 1996); Taylor v. Arizona, 972 F. Supp. 1239 (D.Ariz. 1997), aff'd, 143 F.3d 1178 (9th Cir. 1998), reh'g granted, opinion withdrawn, 158 F.3d 1059 (9th Cir. 1998). Thus far, the only circuit court to uphold a district court's finding that the termination provision of the PLRA is unconstitutional, however, was the United States Court of Appeals for the Ninth Circuit in a decision that has since been withdrawn. Taylor v. United States, 143 F.3d 1178 (9th Cir. 1998), reh'g granted, opinion withdrawn, 158 F.3d 1059 (9th Cir. 1998) On somewhat varying theories, the First, Second, Sixth, Eighth and Eleventh Circuits have upheld the constitutionality of the PLRA termination provisions. See Benjamin v. Jacobson, 124 F.3d 162 (2nd Cir. 1997); Dougan v. Singletary, 129 F.3d 1424, 1426 (11th Cir. 1997); Gavin v. Branstad, 122 F.3d 1081 (8th Cir. 1997); Hadix v. Johnson, 133 F.3d 940 (6th Cir. 1998); Inmates of Suffolk County Jail v. Rouse, 129 F.3d 649 (1st Cir. 1997); Plyler v. Moore, 100 F.3d 365 (4th Cir. 1996), cert. denied, 520 U.S. 1277, 117 S.Ct. 2460, 138 L.Ed.2d 217 (1997). And yet, perhaps the most recent court to consider the issue, the United States District Court for the District of New Jersey, adjudged that § 3626(b)(2) of the PLRA "constitutes a violation of separation of power principles." Denike v. Fauver, 3 F. Supp.2d 540, 548 (D.N.J. 1998).

While its ramifications for the PLRA are obviously highly contested, the importance of the separation of powers doctrine is not. The separation of powers doctrine is a fundamental tenet of our democratic government. The doctrine is "basic and vital, namely, to preclude a commingling of the . . . essentially different powers of government in the same hands." O'Donoghue v. U.S., 289 U.S. 516, 530, 53 S.Ct. 740, 77 L.Ed. 1356 (1933) ( citations omitted). Although it is never explicitly referenced in the Constitution, the axiom that the judicial, legislative, and executive branches must not usurp each other's power finds a reverential home in the decisions of the Supreme Court. See Steel Co. v. Citizens for a Better Environment, 523 U.S. 83, 118 S.Ct. 1003, 140 L.Ed.2d 210 (1998); INS v. Chadha, 462 U.S. 919, 951, 103 S.Ct. 2764, 77 L.Ed.2d 317 (1983); National Mut. Ins. Co. v. Tidewater Transfer Co., 337 U.S. 582, 590-91, 69 S.Ct. 1173, 93 L.Ed. 1556 (1949).

2. Congress's Inability to Reopen an Article III Court's Final Judgment

In light of the collective wisdom of a number of Supreme Court decisions, the PLRA's constitutional shortcomings are clear. The first, Plaut v. Spendthrift Farm, Inc., holds that an Article III court's final judgment may not be reopened by Congress. 514 U.S. 211, 229, 115 S.Ct. 1447, 131 L.Ed.2d 328 (1995).

Plaut involved Congress's attempted reinstatement of certain securities fraud actions that had been dismissed as time-barred under a previous Supreme Court decision. Four years prior to the Plaut decision, in Lampf, Pleva, Lipkind, Prupis Petigrow v. Gilbertson, the Supreme Court held that litigation under § 10(b) of the Securities and Exchange Act and the SEC's Rule 10b-5 "must be commenced within one year after the discovery of the facts constituting the violation and within three years of such violation." 501 U.S. 350, 364, 111 S.Ct. 2773, 115 L.Ed.2d 321 (1991). This decision obviously imposed a shorter-than-expected limitations period on many litigants. In 1991, Congress enacted a statute that not only changed the statute of limitations for § 10(b), but also purportedly applied retroactively, so as to reverse the numerous judicial dismissals on the merits that occurred in the wake of Lampf. Plaut, 514 U.S. at 214, 115 S.Ct. 1447. Hence, the legislation attempted to reopen the courts' final judgments. Id. at 214-215, 115 S.Ct. 1447.

Citing Thomas Jefferson, James Madison, Alexander Hamilton, Sir William Blackstone and Abraham Lincoln, id. at 219-26, 115 S.Ct. 1447, the Supreme Court in Plaut noted that "[t]he Framers of our Constitution lived among the ruins of a system of intermingled legislative and judicial powers," id. at 219, 115 S.Ct. 1447, and felt a "sense of a sharp necessity to separate the legislative from the judicial," id. at 221, 115 S.Ct. 1447. From their own words, it is abundantly clear that the Framers sought to insulate final judgments from legislative revision. Id. at 240, 115 S.Ct. 1447.

To this day, "it remains a basic principle of our constitutional scheme that one branch of the Government may not intrude upon the central prerogatives of another." Loving v. U.S., 517 U.S. 748, 757, 116 S.Ct. 1737, 135 L.Ed.2d 36 (1996). "Judgments within the powers vested in courts by the Judiciary Article of the Constitution may not lawfully be revised, overturned or refused faith and credit by another Department of Government." Chicago Southern Air Lines, Inc. v. Waterman S.S. Corp., 333 U.S. 103, 113, 68 S.Ct. 431, 92 L.Ed. 568 (1948). "[N]o decision of any court of the United States can, under any circumstances, . . . be liable to a revision, or even suspension, by the Legislature itself, in whom no judicial power of any kind appears to be vested." Hayburn's Case, 2 U.S. (2 Dall.) 408, 413, 1 L.Ed. 436 (1792).

Noting that its previous decisions had "uniformly provided fair warning that such an act exceeds the powers of Congress," the Supreme Court in Plaut declared that "[w]hen retroactive legislation requires its own application in a case already finally adjudicated, it does no more and no less than `reverse a determination once made, in a particular case.'" 514 U.S. at 225, 115 S.Ct. 1447 ( citations omitted) ( quoting THE FEDERALIST No. 81, at 545 (Hamilton) (J. Cooke, ed., 1961)). Under the Constitution of the United States, Congress simply has no power to set aside a court's final judgment.

The PLRA, if applied to the 1992 Final Judgment in this civil action, would affect just such an unconstitutional reversal of a judicial final judgment. The two termination provisions of the PLRA would mandate the revision of a final judgment entered by this court four years before the statute was signed into law. Like the Supreme Court, this court "know[s] of no previous instance in which Congress has enacted retroactive legislation requiring an Article III court to set aside a final judgment, and for good reason. The Constitution's separation of legislative and judicial powers denies it the authority to do so." Plaut, 514 U.S. at 240, 115 S.Ct. 1447.

3. The Consent Decree as a Final Judgment

The several circuit courts that have declined to find the PLRA termination provisions unconstitutional have done so not by disputing the separation of powers doctrine, but instead by questioning whether a consent decree entered in prison litigation qualifies as a "final judgment" for the purposes of that doctrine. Dougan v. Singletary, 129 F.3d 1424, 1426 (11th Cir. 1997); Gavin v. Branstad, 122 F.3d 1081 (8th Cir. 1997); Hadix v. Johnson, 133 F.3d 940 (6th Cir. 1998); Inmates of Suffolk County Jail v. Rouse, 129 F.3d 649 (1st Cir. 1997); Plyler v. Moore, 100 F.3d 365 (4th Cir. 1996), cert. denied, 520 U.S. 1277, 117 S.Ct. 2460, 138 L.Ed.2d 217 (1997). This court respectfully disagrees with these determinations. See Denike, 3 F. Supp.2d 540. A consent decree has all the finality of a final judgment, and is therefore immune to legislative tampering.

The Supreme Court clearly established consent decrees' status as a final judgment in Rufo v. Inmates of Suffolk County Jail, 502 U.S. 367, 391, 112 S.Ct. 748, 116 L.Ed.2d 867 (1992). The Supreme Court held that a consent decree "is subject to the rules generally applicable to other judgments and decrees." Id. at 378, 112 S.Ct. 748.

In a situation not unlike the one before this court today, that decision involved a consent decree entered by a federal court to settle inmates' claims of constitutional violations. As in the Ruiz litigation, the Rufo district court had previously found constitutional violations in a detention facility. Id. at 372-73, 112 S.Ct. 748. The Supreme Court was considering the legitimacy of proposed modifications to the consent decree. Id. Before establishing the standard for modification of a consent decree, the Court addressed the somewhat nebulous nature of consent decrees, generally:

A consent decree no doubt embodies an agreement of the parties and thus in some respect is contractual in nature. But it is an agreement that the parties desire and expect will be reflected in, and be enforceable as, a judicial decree that is subject to the rules generally applicable to other judgments and decrees.

Hence, in accordance with the separation of powers doctrine, such a "reopening" in the interest of equity may be effected only by a court, not by Congress. Rule 60(b) of the Federal Rule of Civil Procedure expressly allows a court to reopen a final judgment in certain delineated situations and other "extraordinary circumstances." Fed.R.Civ.Pro. 60(b). That rule, however, far from granting Congress any exemption from the separation of powers doctrine, "merely reflects and confirms the courts' own inherent and discretionary power. . . ." Plaut, 514 U.S. at 233, 115 S.Ct. 1447. The PLRA, therefore, represents an unconstitutional intrusion on the role of Article III courts.

4. The Final Judgment's Protection of Inmates' Private Constitutional Rights

The crux of the discrepancy between courts that have upheld and those that have struck down the termination provisions of the PLRA is the relative "finality" of a consent decree. In the face of Plaut's clear declaration that "[t]he separation of powers violation here, if there is any, consists of depriving judicial judgments of the conclusive effect that they had when they were announced . . .," 514 U.S. at 228, 115 S.Ct. 1447, proponents of the termination provisions' constitutionality have seized on an overly-narrow interpretation of the 1855 decision Pennsylvania v. Wheeling Belmont Bridge Co., 59 U.S. (18 How.) 421, 15 L.Ed. 435 (1855). Citing Wheeling Bridge, the Plaut Court did in fact distinguish the unconstitutional statute before it from "legislation . . . that altered the prospective effect of injunctions entered by Article III courts." 514 U.S. at 232, 115 S.Ct. 1447. In the context of prison litigation consent decrees, Wheeling Bridge is, therefore, sometimes cited for the proposition that Plaut's final judgment analysis is inapplicable to prospective relief, because such relief is not considered a "final judgment." See, e.g., Hadix, 133 F.3d 940. A careful review of the Wheeling Bridge decision, however, reveals that it supports the finality of the 1992 Final Judgment in the Ruiz litigation.

In Wheeling Bridge, the Supreme Court was faced with a petition for an injunction to prevent the rebuilding of a low bridge over a navigable river. 59 U.S. at 429. In response to an earlier case brought by the State of Pennsylvania in which a court granted injunctive relief requiring that the bridge be restructured so as to allow free navigation under it, Congress passed a statute declaring the bridge lawful. Id. When the bridge was destroyed in a storm, a new petition was filed pursuant to the original injunction to stop the bridge's reconstruction. Id. The Court's decision to refuse the second injunction was based, in part, on a distinction between money judgments and executory injunctions. Wheeling Bridge, 59 U.S. at 431-32. A money judgment, said the Court, is unaffected by the subsequent law. Id.

"But that part of the decree, directing the abatement of the obstruction, is executory, a continuing decree, which requires not only the removal of the bridge, but enjoins the defendants against any reconstruction or continuance. Now, whether it is a future existing or continuing obstruction depends upon the question whether or not it interferes with the right of navigation. If, in the meantime, since the decree, this right has been modified by the competent authority, so that the bridge is no longer an unlawful obstruction, it is quite plain the decree of the court cannot be enforced. There is no longer any interference with the enjoyment of the public right inconsistent with law. . . ."

Two aspects of this decision clearly distinguish it from the situation at hand. First, faced with the changed law in Wheeling Bridge, the Supreme Court determined that because Congress had established the right of free navigation in the first place, it could in fact alter those standards. 59 U.S. at 431. There is no question that Congress can change its own laws. And, when those laws are the foundation for an injunction's prospective relief, that relief must be revisited by the issuing court.

As regards the 1992 Final Judgment, however, the Constitution-not Congress-is the source of the underlying rights that the consent decree remedied. All injunctive relief entered by this court since those findings of constitutional violations, has been predicated on correcting those violations. The 1992 Final Judgment was entered by this court as a consent decree, at the request of the parties, as a means of resolving plaintiffs' allegations of constitutional violations without further litigation. Under Wheeling Bridge, Congress may revise its own laws, thereby affecting the viability of a court's prospective relief. Congress may not, however, revise the Constitution.

Second, the Wheeling Bridge Court distinguishes between "public" and "private" rights. Id. at 431-32. Rather than declare that a judgment's finality depends only on whether it is prospective relief or a money judgment, the Wheeling Bridge Court explicitly emphasizes the private/public right distinction. Id. A number of courts, including the Supreme Court, have interpreted Wheeling Bridge as making this distinction. See, e.g. Hodges v. Snyder, 261 U.S. 600, 603-04, 43 S.Ct. 435, 67 L.Ed. 819 (1923); Johnston v. Cigna Corp., 14 F.3d 486, 492 (10th Cir. 1993); Daylo v. Administrator of Veterans' Affairs, 501 F.2d 811, 818 (D.C. Cir. 1974). In Hodges, the Supreme Court explained that

it is true . . . the private right of parties which have been vested by the judgment of a court cannot be taken away by subsequent legislation, but must be thereafter enforced by the court regardless of such legislation. . . . This rule, however, as held in the Wheeling Bridge case, does not apply to a suit brought for the enforcement of a public right, which, even after it has been established by the judgment of the court, may be annulled by subsequent legislation and should not be thereafter enforced. . . .

Prospective relief of a public right-one established in the first place by Congress, such as the right to navigate a river-may, of course, be altered by Congress's revision of that underlying right. Wheeling, at 431. Congress has the power to alter prospective relief only where it has plenary power over the underlying law-the realm of "public" rights. The consent decree at issue before this court, however, involves private constitutional rights-those that Congress may not revise. The Wheeling Bridge Court itself stated:

it is urged, that the act of congress cannot have the effect and operation to annul the judgment of the court already rendered, or the rights determined thereby. . . . This, as a general proposition, is certainly not to be denied, especially as it respects adjudication upon the private rights of parties. When they have passed into judgment the right becomes absolute, and it is the duty of the court to enforce it.

[There is] a critical difference between rights created by federal statute and rights recognized by the Constitution. . . . [S]uch a distinction seems to us to be necessary in light of the delicate accommodations required by the principle of separation of powers reflected in Art. III. The constitutional system of checks and balances is designed to guard against "encroachment or aggrandizement" by Congress at the expense of other branches of government. But when Congress creates a statutory right, it clearly has the discretion, in defining that right, to create presumptions, or assign burdens of proof, or prescribe remedies. . . . No comparable justification exists, however, when the right being adjudicated is not of congressional creation. In such a situation, substantial inroads into functions that have traditionally been performed by the Judiciary cannot be characterized merely as incidental extensions of Congress' power to define rights that it has created. Rather, such inroads suggest unwarranted encroachments upon the judicial power of the United States, which our Constitution reserves for Art. III courts.

Thus, a consent decree based on private constitutional rights, even containing prospective relief, becomes final upon adoption by the court. There can be no doubt that "[e]ven when a branch does not arrogate power to itself, . . . the separation-of-powers doctrine requires that a branch not impair another in the performance of its constitutional duties." Loving v. U.S., 517 U.S. 748, 757, 116 S.Ct. 1737, 135 L.Ed.2d 36 (1996). The separation of powers doctrine, therefore, prohibits Congress from overturning the 1992 Final Judgment entered in this civil action. With the termination provisions of the PLRA, "Congress has exceeded its authority by requiring the federal courts to exercise `the judicial Power of the United States,' in a manner repugnant to the text, structure and traditions of Article III." See Plaut, 514 U.S. at 217-218, 115 S.Ct. 1447 ( quoting U.S. Const., Art. III, § 1).

5. Congress's Inability to Decide a Discrete Group of Cases

Independently from, and in addition to, the Plaut axiom that Congress may not reopen a court's final judgment, the PLRA violates the separation of powers principle that Congress may not prescribe a rule of decision in a discrete group of Article III cases. United States v. Klein, 80 U.S. (13 Wall.) 128, 20 L.Ed. 519 (1871). That is, the PLRA would go further than simply reopening a court's final judgment, it would actually usurp the court's decision in the controversy. The Supreme Court, in Klein, expressly rejected such legislative adjudications. Id.

Klein grew out of a 1863 statute that allowed individuals to recover property that was seized by the government during the Civil War only upon showing proof that they did not aid or comfort the South during the war. Id. at 131-33. In response to a Supreme Court decision that a presidential pardon fulfilled the statute's requirement, Congress legislated that a pardon was actually proof that the person did in fact aid the South. Klein, 80 U.S. at 133. The statute thereby stripped the jurisdiction of any court hearing such a case if proof of a pardon was presented. Id. The Supreme Court struck down the law as unconstitutional. Id. at 146-47. And, in words appropriate to the PLRA's relation to this court's 1992 Final Judgment, the Supreme Court stated:

It seems to us that this is not an exercise of the acknowledged power of Congress to make exceptions and prescribe regulations to the appellate power. . . . What is this but to prescribe a rule for the decision of a cause in a particular way? . . . Can we do so without allowing one party to the controversy to decide it in its own favor? Can we do so without allowing that the legislature may prescribe rules of decision to the judicial department in the cases pending before it? . . . We think not. . . . We must think that Congress has inadvertently passed the limit which separates the legislative power from the judicial power.

The unacceptable situation before the Supreme Court in Klein was that a judicial institution was required to dismiss a case upon proof of certain evidence: in that case, a pardon. Similarly, under the PLRA, the judicial institution must terminate relief upon proof of certain evidence: here, the absence of particular findings. See 18 U.S.C. § 3626(b). The limitations on termination in 18 U.S.C. § 3626(b)(3), rather than preserve the court's involvement, only further remove the court from the judicial process and replace it with legislation. Because in most cases there was no record made at the time of the final judgment, it is impossible for a court to later make such findings. Under the PLRA, if there were actual violations of the consent decree by the defendants, the plaintiff class would have to start from scratch demonstrating, anew, constitutional violations in the prison system. Through its intrusive provisions, the PLRA effectively prohibits meaningful discovery, hearing, briefing and decision-making, thereby violating the separation of powers doctrine as presented by Klein.

6. The PLRA's Ambiguous Drafting

A number of courts have commented on the PLRA's poor drafting. And, having attempted to make sense of the statute, this court agrees that "[t]he PLRA is not a paragon of clarity." Inmates of Suffolk Co. Jail v. Rouse, 129 F.3d 649, 654 (1st Cir. 1997). In its definition sections, the PLRA "contradicts the usual understanding of both `relief' and `consent decree.'" Id. According to Black's Law Dictionary, a consent decree is a "judgment entered by consent of the parties whereby the defendants agree to stop alleged illegal activity without admitting guilt or wrongdoing," 6th ed. 1990 at 410, and a "judgment" is a "final decision of the court resolving the dispute and determining the rights and obligations of the parties." Id. at 841-42. In the PLRA, "Congress conflated the two terms when it described consent decrees as a form of relief rather than as a judgment that engenders relief." Rouse, 129 F.3d at 654. In light of these drafting problems, it is beyond dispute that "[t]he statute's ambiguity is daunting because its consequences are significant." Benjamin, 124 F.3d at 168. Ironically, to some respected jurists, it is precisely this ambiguity that preserves the PLRA's constitutionality.

The Second Circuit, in Benjamin v. Jacobson, capitalized on the ambiguity in the PLRA to avoid the troubling constitutional questions presented by it. Id. That court asserted that the PLRA termination provisions could be read either to "limit the jurisdiction of federal courts so that these courts could not in the future enforce past consent decrees . . .," or to "render null and void all past federally approved prison consent decrees unless these decrees met the requirement of being narrowly tailored to a federal right." Id. at 166-167 ( emphasis added). The former of these readings preserves the statute's constitutionality because Congress may alter federal courts' jurisdiction in some situations. See Yakus v. U.S., 321 U.S. 414, 64 S.Ct. 660, 88 L.Ed. 834 (1944). However, this interpretation forces the court to read the termination provisions, which call for "termination of prospective relief," as somehow preserving the consent decree while removing it from federal jurisdiction. The PLRA nowhere distinguishes between a federal court's jurisdiction over prospective relief and the consent decree giving rise to that relief. On the other hand, the latter of the two readings (which would make the consent decree null and void) would, at the very least, "raise issues of when and how the legislature can affect past judgments in the light of fundamental principles of separation of powers." Benjamin, 124 F.3d at 168.

Thus, the Second Circuit's analysis exposes the true conundrum of the PLRA termination provisions. A court must either utilize tortured statutory interpretation to reach a far-fetched legal fiction, or strike down the provisions as violations of the Constitution of the United States. As the price for maintaining the law's constitutionality, the Second Circuit chose to somehow transform the PLRA's termination provisions into jurisdictional ones.

Based on the plain language of the statute, this court must agree with those courts that have found no fair basis for such an interpretation. See Dougan, 129 F.3d at 1426 n. 4 (11th Cir. 1997). And, to assume that the plaintiffs could enforce a validly negotiated consent decree in state court after the issuing federal court "terminated prospective relief" is, at best, unrealistic. Thus, given the fact that the PLRA's constitutionality "depends on [an] interpretation of the termination provision as doing no more than limiting the remedial jurisdiction of the federal courts," id. at 173, this court agrees with those who have determined that the termination provisions of the PLRA violate the separation of powers doctrine of the Constitution of the United States.

7. The PLRA's Unambiguous Legislative History

Faced with ambiguity in the meaning of a statute, it is not uncommon for a court to look for guidance to the legislative history of the law. Unfortunately, the legislative history of the PLRA is relatively sparse. What record there is, however, does not indicate that Congress meant the PLRA as merely a limitation on federal jurisdiction.

Eventually signed into law as an attachment to an omnibus appropriations bill, the law was conceived in the House of Representatives as the Violent Criminal Incarceration Act of 1995. H.R. 667. The sum total of the documented consideration of the proposal is little more than a "single Senate hearing before the Judiciary Committee, one substantive House Report, and some floor debate . . ." Benjamin, 935 F. Supp. at 340, aff'd in part, rev'd in part, 124 F.3d 162 (2d Cir. 1997).

It is clear from that record, however, that the troubling separation of powers implications of the PLRA were well known to its drafters. In July of 1995, Associate Attorney General John Schmidt warned the Senate Judiciary Committee that the PLRA was subject to attack on separation of powers grounds. Associate Attorney General Schmidt explained that:

The application of these restrictions to such relief raises constitutional concerns under the Supreme Court's recent decision in Plaut v. Spendthrift Farm, Inc. . . . . The Court held in that case that legislation which retroactively interferes with final judgments can constitute an unconstitutional encroachment on judicial authority. . . . [T]he application of proposed 18 U.S.C. § 3626(b) to orders in pre-enactment final judgments would raise serious constitutional problems.

142 CONG. REC. S2299 (daily ed. Mar. 19, 1996). By letter, Senators Fred Thompson, James Jeffords, Edward Kennedy, Joseph Biden, Jr., and Jeff Bingaman recommended that the "Administration negotiate changes in the PLRA that remedy the serious . . . problems outlined by Mr. Schmidt and other experts." See 142 Cong. Rec S2297 (daily ed. Mar. 19, 1996).

The Act's proponents, however, ignored those concerns and, rather, trumpeted the Act's potential to curb frivolous prisoner lawsuits and to end federal micro-management of state prisons. See Overhauling the Nation's Prisons, Testimony July 28, 1995, Before Senate Judiciary Committee, 104th Con. (1995) (statement of Sen. Hatch). Senator Dole characterized the PLRA provisions as restraining "liberal Federal judges who see violations of constitutional rights in every prisoner complaint and who have used these complaints to micromanage state and local prison systems" in fashioning remedies that raise conditions beyond minimal constitutional standards. 141 CONG. REC. S14,414 (daily ed. Sept. 27, 1995) (statement of Sen. Dole). Senator Abraham trumpeted the PLRA as "limiting judicial remedies in prison cases." 141 CONG. REC. S14316 (daily ed. Sept. 26, 1995). In the words of Senator Kay Baily Hutchison of Texas, "[p]risons exist for the protection of society-not for the comfort and convenience of criminals. Interference by the federal courts has put the interests of criminals ahead of the interests of victims and law-abiding citizens." Benjamin, 935 F. Supp. at 340 ( quoting Hearing on Prison Reform Before the Senate Comm. On the Judiciary, 104th CONG. 1th Sess. (July 27, 1995)).

From this sparse legislative history of the PLRA, there can be no doubt that Congress not only knew of the constitutional problems with the statute, but passed the statute with the purpose of reopening and deciding judicially developed final judgments. Congress's clear intent to set aside judgments made by federal courts in prison litigation only validates the unconstitutionality of the PLRA under the Supreme Court's decisions in Plaut and Klein.

8. Conclusion

As the Supreme Court of the United States' jurisprudence makes clear, Congress is prohibited from reopening a final judgment of an Article III federal court, and from prescribing decisions in particular cases. "The prohibition is violated when an individual final judgment is legislatively rescinded for even the very best of reasons, such as the legislature's genuine conviction (supported by all the law professors in the land) that the judgment was wrong; and it is violated 40 times over when 40 final judgments are legislatively dissolved." Plaut, 514 U.S. at 228, 115 S.Ct. 1447. The PLRA's termination provisions, by effectively terminating the consent decree that was negotiated as a resolution of the plaintiffs' valid allegations of unconstitutional abuse, neglect, and mistreatment in Texas prisons, violates the separation of powers doctrine of the United States Constitution.

C. Due Process

Plaintiffs also argue that application of the termination provisions to the 1992 Final Judgment would violate due process by undermining vested rights created the judicial decision. As both contentions turn on the "finality" of the 1992 Final Judgment, the viability of this argument is directly related to the plaintiffs' separations of powers argument. "In essence, the vested rights doctrine is really only the due process analogue of the separation-of-powers doctrine that prevents Congress from reopening final judgments of Article III courts." Gavin v. Branstad, 122 F.3d 1081, 1091 (8th Cir. 1997).

If the consent decree actually qualifies as a "final judgment" for separation of powers purposes, plaintiffs would have vested rights in its protections. See Hodges, 261 U.S. at 603, 43 S.Ct. 435. "It is not within the power of a legislature to take away rights which have been once vested by a judgment. Legislation may act on subsequent proceedings, may abate actions pending, but when those actions have passed into judgment the power of the legislature to disturb the rights created thereby ceases." McCullough v. Virginia, 172 U.S. 102, 123-24, 19 S.Ct. 134, 43 L.Ed. 382 (1898). It has been determined, based on the Supreme Court's decisions in Plaut, Rufo, Wheeling Bridge, and Northern Pipeline, that the 1992 Final Judgment in this civil action does qualify as a "final judgment" for the purpose of protection against legislative interference. The Due Process Clause protects the judgment rights of its beneficiaries, subject only to the court's inherent equitable powers as reflected in Rule 60(b). For the same reasons that this court granted plaintiffs' separation of power arguments, the court finds that the PLRA violates plaintiffs' due process rights by interfering with their vested rights in the decree.

Plaintiffs make another, similar argument that the termination provisions of the PLRA unconstitutionally impair their contract with the State-a contract that was carefully negotiated to avoid litigating the constitutionality of the system itself. Plaintiffs argue that retroactive legislation should not, and cannot, nullify the hard-earned agreement represented by the consent decree:

Had the parties known of the PLRA requirements when they negotiated the Final Judgment, they could have stipulated to PLRA findings so the Judgment would not be immediately terminable. If defendants had refused to stipulate, plaintiffs would have been free to seek more stringent relief than provided by the Judgment, based on explicit PLRA findings. But sweeping away the settled expectations they did have because of a totally unanticipated later-enacted statute plainly violates plaintiffs' due process rights.

( Pl. Mem. in Opp. Mot. for J., Feb. 8, 1999, at 10).

It is well-settled that "[a] consent decree no doubt embodies an agreement of the parties and thus in some respects is contractual in nature." Rufo, 502 U.S. at 378, 112 S.Ct. 748. A number of courts have considered consent decrees to be "hybrids," having "attributes both of contracts and of judicial decrees." Local No. 93, International Association of Firefighters v. City of Cleveland, 478 U.S. 501, 519, 106 S.Ct. 3063, 92 L.Ed.2d 405 (1986).

As a contract, the consent decree at issue in this civil action is protected from legislative impairment by the due process clause of the Fifth Amendment. See Pension Benefit Guaranty Corp. v. R.A. Gray Co., 467 U.S. 717, 733, 104 S.Ct. 2709, 81 L.Ed.2d 601 (1984). However, when a federal statute impairs a private contract, judicial scrutiny is "quite minimal." National R.R. Passenger Corp. v. Atchison, Topeka Santa Fe R.R. Co., 470 U.S. 451, 105 S.Ct. 1441, 84 L.Ed.2d 432 (1985). Furthermore, the statute violates the Due Process Clause only if Congress has "acted in an arbitrary and irrational way" in passing the PLRA. See Id. It is found that under the specific facts before the court here, such a finding is not warranted. Unlike their due process claim based on their vested rights in the judgment, the plaintiffs' contract-based due process claims, therefore, are without merit.

D. Equal Protection

The plaintiffs further argue that the termination provisions of the PLRA are unconstitutional, because they deny equal protection to prisoners by restricting the fundamental rights of a disfavored and unpopular class of citizens. The first determination for such an equal protection claim is the applicable level of scrutiny. Heightened scrutiny is appropriate when a statute uses a classification that is constitutionally suspect, McLaughlin v. Florida, 379 U.S. 184, 192, 85 S.Ct. 283, 13 L.Ed.2d 222 (1964), or when such a statute burdens a fundamental right. San Antonio Indep. Sch. Dist. v. Rodriguez, 411 U.S. 1, 33-34, 93 S.Ct. 1278, 36 L.Ed.2d 16 (1973). The plaintiffs do not contend that prisoners are a suspect class, but rather that the PLRA restricts fundamental rights of prisoners and therefore is subject to strict scrutiny.

While 18 U.S.C. § 3626(b)(2) and (3) circumscribe a district court's ability to provide some prospective relief to prisoners, this restriction is allowable only in the absence of violations of Federal rights. 18 U.S.C. § 3626(b)(3). Therefore, because the termination provisions of the PLRA also permit prisoners to establish violations of their federal constitutional rights in federal courts, it cannot be said that the termination provisions restrict their fundamental rights. A statute that neither abridges a fundamental right nor operates against a suspect class receives rational basis review when it is challenged under the Equal Protection Clause. Romer v. Evans, 517 U.S. 620, 631, 116 S.Ct. 1620, 134 L.Ed.2d 855 (1996).

In determining whether there is a legitimate purpose for a legislative classification, the "`burden is on the one attacking the legislative arrangement to negative every conceivable basis which might support it.'" Heller v. Doe, 509 U.S. 312, 320-21, 113 S.Ct. 2637, 125 L.Ed.2d 257 (1993) ( quoting Lehnhausen v. Lake Shore Auto Parts Co., 410 U.S. 356, 364, 93 S.Ct. 1001, 35 L.Ed.2d 351 (1973)). The rational-basis review in equal protection analysis "is not a license for courts to judge the wisdom, fairness, or logic of legislative choices." FCC v. Beach Communications, Inc., 508 U.S. 307, 313, 113 S.Ct. 2096, 124 L.Ed.2d 211 (1993). "In the ordinary case, a law will be sustained if it can be said to advance a legitimate government interest, even if the law seems unwise or works to the disadvantage of a particular group, or if the rationale for it seems tenuous." Romer, 517 U.S. at 632, 116 S.Ct. 1620, 134 L.Ed.2d 855 (1996). "[C]ourts are compelled under the rational standard basis review to accept a legislature's generalizations even when there is an imperfect fit between means and ends." Heller, 509 U.S. at 321, 113 S.Ct. 2637. The PLRA satisfies this deferential standard. The plaintiffs have failed to show that the PLRA is not rationally related to a legitimate government purpose.

E. Conclusion

It is determined that the Supreme Court's jurisprudence clearly prohibits the legislature from reopening and deciding issues resolved by this court's 1992 Final Judgment. As an unconstitutional statute, the PLRA cannot be applied to this civil action by this court. Having found the termination provisions of the Prison Litigation Reform Act to be in clear violation of the separation of powers doctrine and due process clause of the Constitution of the United States, the plaintiffs' allegations of unconstitutional practices and conditions in the Texas prison system will be considered as alternative bases for the decision in this matter.

IV. THE FACT-FINDING HEARING

A. Procedure for the Truncated Hearing

Pursuant to the Court of Appeals for the Fifth Circuit's mandate that an order in this matter issue no later than March 1, 1999, the fact-finding hearing, which, as previously stated, had been expected to last between two and three months, was compressed into three-and-a-half weeks (including two Saturday sessions). Each party was granted one hour for opening statements and one hour for closing statements. Each party was granted fifty hours for the presentation of evidence, cross-examination of the other parties' witnesses, and objections. Plaintiffs used virtually all of their allotted time. Defendants used all but four hours and twenty minutes of their fifty hours. Closing statements were heard on February 12, 1999.

B. Procedure for Evidentiary Objections

In the interest of limiting interruptions during the short time available for testimony, it was determined that evidentiary objections would be unnecessarily time-consuming and disruptive during this fact-finding hearing. As explained to both parties before the hearing, this court would, and has, relied only on admissible and relevant evidence to come to its conclusions. That is, during trial, all evidence was provisionally admitted. During deliberation by the court, inadmissible and irrelevant evidence was not considered.

V. OVERVIEW OF THE EVIDENCE

During the nineteen days of testimony, the court heard from over 60 witnesses and admitted over 330 exhibits. The plaintiffs' witness list was made up of nine well-credentialed experts (most of whom had reviewed some aspect of the TDCJ-ID system) and a great many current and former inmates who had purportedly experienced first-hand the unconstitutional conditions and practices complained of by the plaintiff class. Defendants presented their case primarily through current or former TDC employees, some of whom were proffered as expert witnesses. Several experienced experts from outside the TDC system testified on behalf of the defendants as well. One prisoner was called by the defendants.

Plaintiffs' evidence may be organized into several categories specifying allegations of unconstitutional practices and conditions in the Texas prison system. First, plaintiffs allege that TDCJ-ID is deliberately indifferent to prisoner's medical and psychiatric care needs. A number of medical and psychiatric experts testified on behalf of the plaintiff class. Plaintiffs' experts Drs. Dennis Jurczak, Roberta Stellman, Jeffrey Metzner, Steven Jenison, and John Robertson all alleged that their independent review of prison medical or psychiatric services revealed serious insufficiencies in TDCJ-ID's treatment of ill prisoners. A number of inmate witnesses testified to their own experiences with what they believe is a deficient system of medical and psychiatric care. Defendants responded with TDCJ-ID officials David Smith, M.D., who testified to the extensive medical policies and infrastructure of the TDCJ-ID medical system, Glenda Adams, M.D., who addressed some specific inmates' medical records, and Suzanne Ducate, M.D., who outlined the policies and procedures of TDCJ-ID psychiatric care. Defendants also called expert witness Robert Jones, M.D., who reported on his own evaluation of the Texas medical and psychiatric care system.

Through the testimony of penological psychology expert Craig Haney, as well as experts Breed and Riveland, and a number of psychiatrists, plaintiffs further allege that prisoners in administrative segregation, especially those with psychiatric illnesses, are suffering cruel and unusual punishment by being deprived of a minimal measure of civilized life's necessities. Plaintiffs further accuse TDCJ-ID of practicing a widespread pattern of warehousing mentally ill prisoners in administrative segregation. Defendants's psychiatric and medical experts, primarily those within the TDCJ-ID system, reject those allegations.

Through inmate and expert witnesses, plaintiffs also accuse TDCJ-ID of failing to ensure the reasonable safety of prisoners incarcerated in TDCJ. Plaintiffs' experts Allen Breed and Chase Riveland, both of whom have extensive experience in prison management, testified to systemic deficiencies in the system's attention to the needs of sexually and physically victimized inmates. Many, if not most, of the inmates who testified for the plaintiffs presented disturbing personal stories of their own victimization at the hands of other inmates, as well as their frustration with a system that, they allege, has failed to protect them despite repeated pleas for help. Defendants responded through a number of TDCJ officials, including officers in the cellblocks and Director of TDCJ-ID, Gary Johnson. The court also heard defense testimony rebutting plaintiffs' safety allegations from two non-TDCJ prison experts who evaluated the Texas system.

Plaintiffs finally allege that TDCJ-ID subjects prisoners to malicious and sadistic force by prison guards. Expert witnesses Breed and Riveland made condemnatory conclusions as to the prevalence of unnecessary and excessive force and intimidation of inmates by correction officers in their day-to-day interaction with them. Inmate witnesses repeatedly described their own purported experiences with, and resulting injuries inflicted by, abusive guards. Defendants countered those charges with essentially the same witnesses as utilized to rebut the inmate protection allegations. A number of individuals well-familiar with prison life testified that there is no pattern of excessive use of force in TDCJ-ID.

VI. OVERVIEW OF PRISONERS' RIGHTS UNDER THE EIGHTH AMENDMENT

The Constitution "does not mandate comfortable prisons," Rhodes v. Chapman, 452 U.S. 337, 349, 101 S.Ct. 2392, 69 L.Ed.2d 59 (1981), but it also does not permit inhumane ones, Farmer v. Brennan, 511 U.S. 825, 832, 114 S.Ct. 1970, 128 L.Ed.2d 811 (1994). While prisoners certainly lose many of the freedoms the rest us of enjoy, they continue to live under the protection of the Eight Amendment, 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, 50 L.Ed.2d 251 (1976); see Helling v. McKinney, 509 U.S. 25, 31, 113 S.Ct. 2475, 125 L.Ed.2d 22 (1993).

Prisoners are not made inhuman by virtue of their incarceration. "[H]aving stripped [prisoners] of virtually every means of self-protection and foreclosed their access to outside aid, the government and its officials are not free to let the state of nature to take its course." Farmer, 511 U.S. at 833, 114 S.Ct. 1970. The Eighth Amendment 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 the jailhouse cells." Madrid v. Gomez, 889 F. Supp. 1146, 1245 (N.D.Cal. 1995).

The Eighth Amendment, which applies to the states through the Due Process Clause of the Fourteenth Amendment, prohibits the infliction of "cruel and unusual punishments" on those convicted of crime. Robinson v. California, 370 U.S. 660, 666-67, 82 S.Ct. 1417, 8 L.Ed.2d 758 (1962). While there is no static "test" by which the courts determine whether conditions of confinement are cruel and unusual, Rhodes, 452 U.S. at 346, 101 S.Ct. 2392, the Eighth Amendment prohibits punishments that are "incompatible with `the evolving standards of decency that mark the progress of a maturing society' . . . or which `involve the unnecessary and wanton infliction of pain.'" Estelle, 429 U.S. at 102-103, 97 S.Ct. 285 ( citations omitted). The Eighth Amendment prohibits punishments such as physical and sexual assault, physically abusive guards, or deprivation of medical care. See Farmer, 511 U.S. at 832, 114 S.Ct. 1970. "Being violently assaulted in prison is simply not `part of the penalty that criminal offenders pay for their offenses against society.'" Id. at 834, 114 S.Ct. 1970 ( quoting Rhodes, 452 U.S. at 347, 101 S.Ct. 2392).

The Eighth Amendment requires that prison officials provide inmates with such minimum essentials as adequate food, shelter, clothing, medical care, and officials must also take reasonable measures to ensure the safety of inmates. Helling, 509 U.S. at 32, 113 S.Ct. 2475; Hudson v. Palmer, 468 U.S. 517, 526-27, 104 S.Ct. 3194, 82 L.Ed.2d 393 (1984).

An inquiry into an Eighth Amendment claim alleging cruel and unusual punishment has two components. The first component is an objective one: "Was the deprivation sufficiently serious?" Wilson v. Seiter, 501 U.S. 294, 298, 111 S.Ct. 2321, 115 L.Ed.2d 271 (1991). "[A] prison official's act or omission must result in the denial of `the minimal civilized measure of life's necessities.'" Farmer, 511 at 834, 114 S.Ct. 1970 ( quoting Rhodes, 452 U.S. at 347, 101 S.Ct. 2392.) This objective component is contextual and responsive to "contemporary standards of decency," Estelle, 429 U.S. at 103, 97 S.Ct. 285, and it focuses on discrete and essential human needs such as health, safety, food, warmth, and exercise. Wilson, 501 U.S. at 304, 111 S.Ct. 2321. And, whether or not a claimant is successful in an Eighth Amendment claim is an issue of law for the court to decide. Hickey v. Reeder, 12 F.3d 754, 756 (8th Cir. 1993).

The second component of an Eighth Amendment inquiry involves intent and is subjective: "Did the officials act with a sufficiently culpable state of mind?" Wilson, 501 U.S. at 298, 111 S.Ct. 2321 (1991). "The source of the intent requirement is not the predilections of this Court, but the Eighth Amendment itself, which bans only cruel and unusual punishment. If the pain inflicted is not formally meted out as punishment by the statute or the sentencing judge, some mental element must be attributed to the inflicting officer before it can qualify." Id. at 300, 111 S.Ct. 2321 ( emphasis in original). In prison-conditions cases, the state of mind required is "deliberate indifference" to the inmate's health and safety. Id. at 303, 111 S.Ct. 2321. In excessive force cases, plaintiffs must show that "officials used force with `a willingness that [harm] occur.'" Farmer, 511 U.S. at 836-37, 114 S.Ct. 1970 ( quoting Hudson v. McMillian, 503 U.S. 1, 7, 112 S.Ct. 995, 117 L.Ed.2d 156 (1992) ( internal quotation marks and citations omitted)).

Mere negligence is not enough to support an Eighth Amendment claim in prison-condition cases. See Estelle, 429 U.S. at 106-107, 97 S.Ct. 285. Rather, the test for deliberate indifference is the criminal law standard of subjective recklessness. Farmer, 511 U.S. at 839-840, 114 S.Ct. 1970. That is, prison officials must have consciously disregarded a substantial risk of serious harm to prisoners. Id. at 839, 114 S.Ct. 1970 ( internal quotations omitted) ( citing MODEL PENAL CODE § 2.02(2)(c)). This state of mind inquiry "is a question of fact subject to demonstration in the usual ways, including inference from circumstantial evidence. . . ." Id. at 842, 114 S.Ct. 1970. "[A] factfinder may conclude that a prison official knew of a substantial risk from the very fact that the risk was obvious." Id.

At the same time, "a prison official cannot be found liable under the Eighth Amendment for denying an inmate humane conditions of confinement unless the 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 837, 114 S.Ct. 1970. "[A]n official's failure to alleviate a significant risk that he should have perceived but did not, while no cause for commendation, cannot under [Supreme Court] cases be condemned as the infliction of punishment." Id. at 838, 114 S.Ct. 1970.

To defend against an accusation of an Eighth Amendment violation, prison officials can "prove that they were unaware even of an obvious risk to inmate health and safety." Id. at 844, 114 S.Ct. 1970. Also, "prison officials who actually knew of a substantial risk to inmate health or safety may be found free from liability if they responded reasonably to the risk, even if the harm was not ultimately averted." Id. This is because prison officials have the "unenviable task of keeping dangerous men in safe custody under humane conditions." Id. ( quoting Spain v. Procunier, 600 F.2d 189, 193 (9th Cir. 1979)).

Nevertheless, this standard does not require that the plaintiffs "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, 511 U.S. at 842, 114 S.Ct. 1970. This also does not mean that "prison officials will be free to ignore obvious dangers." Id. A prison official will "not escape liability if the evidence showed that he merely refused to verify underlying facts that he strongly suspected were true, or declined to confirm inferences of risk that he strongly suspected to exist. . . ." Id. at 843 n. 8, 114 S.Ct. 1970. "The long duration of a cruel punishment condition may make it easier to establish knowledge and hence some form of intent. . . ." Wilson, 501 U.S. at 300, 111 S.Ct. 2321.

While prison-conditions cases are governed by the mental element of deliberate indifference, "whenever prison officials stand accused of using excessive physical force in violation of the Cruel and Unusual Punishments Clause, 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 6-7, 112 S.Ct. 995. The factors that should be considered in an excessive force claim include the extent of the injury suffered, the need for application of force, the relationship between the need and the amount of force used, the threat "reasonably perceived by the responsible officials," and "any efforts made to temper the severity of a forceful response." Hudson, 503 U.S. at 7, 112 S.Ct. 995 ( quoting Whitley v. Albers, 475 U.S. 312, 321, 106 S.Ct. 1078, 89 L.Ed.2d 251(1986)).

VII. PROVING SYSTEMIC CONSTITUTIONAL VIOLATIONS

A. A Matter of "Logic and Judgment"

If the PLRA applied to the 1992 Final Judgment, prospective relief would continue if this court made written findings that such relief remains necessary to correct a current and ongoing violation of a Federal right. 18 U.S.C. § 3626(b)(3). In the evidentiary hearing that began January 21, 1999, the plaintiffs sought to show, by a preponderance of the evidence, such current and ongoing violations of a Federal right.

One of the persistent arguments by defendants in the current phase of this litigation is that plaintiffs have been unable to show the types of systemic constitutional violations that justify broad-scale, system-wide relief. Defendants correctly cite the Supreme Court's decision in Lewis v. Casey for the proposition that such system-wide relief can only be justified by findings of a system-wide violation. 518 U.S. 343, 359, 116 S.Ct. 2174, 135 L.Ed.2d 606 (1996). This court does not agree, however, with defendants' apparent contention that plaintiffs would have to evaluate every single one of the over 100 institutions in TDCJ-ID to make their case. Such a requirement would virtually immunize a system as large as Texas' from constitutional scrutiny. Rather, whether evidence rises to a level of system-wide violation is based on the finding of a persistent pattern as determined by the logical evaluation of the trial judge. See, e.g., Orantes-Hernandez v. Thornburgh, 919 F.2d 549 (9th Cir. 1990). "Although systemwide injunctive relief may not be predicated on individual misconduct that `is not part of a pattern of persistent and deliberate official policy,' an institutional practice may be sufficiently prevalent to warrant such relief even though it is not embodied in regulations or imposed by official fiat. Executives have a duty to control their subordinates." Ruiz, 679 F.2d at 1154 (citations omitted).

The Fifth Circuit spoke directly to this point in earlier stages of this very litigation:

Whether the data observed are sufficient to warrant a general conclusion must be determined by logic and judgment. The human events of the past cannot be sampled so perfectly as to permit a statistically valid analysis. This kind of conclusion is not a mixed question of fact and law or one of legal inferences from the facts. It is instead one of the sufficiency of an evidentiary basis for a factual conclusion. Although that type of conclusion does not rest on the credibility of witnesses or the advantage of firsthand observation, it is equally a question of fact, to be accepted by us unless clearly erroneous.

B. Defendants' Objections to Plaintiffs' Experts' Methodology

In direct contrast to this Fifth Circuit language, defendants objected to what they believe are not statistically-valid methods employed by plaintiffs' experts in their "sampling" of charts, cases, and inmates. Defendants allege that plaintiffs' experts have no reliable bases for their conclusions, "which are based primarily on their unsupported, subjective opinion." Defendants contend that "[n]ot one of the Plaintiffs' experts have conducted anything like a scientific inquiry into the status of the Texas prison system. . . . [I]t appears that the experts cherry-picked from the general prison population those cases which appear to be the most egregious in order to show some type of violation." ( Def.'s Mot. for J., 9-10.) The defendants assert that "[e]ven if these experts located isolated incidents of potential constitutional violations . . ., their anecdotal testimony is insufficient to establish a current and ongoing systemic violation." ( Id. at 10.)

Defendants called statistics expert Daniel Freeman, Jr., Ph.D., to impeach the statistical validity of several of the plaintiffs' experts' methods. Dr. Freeman, by document, and through in-court testimony, verified the statistical validity of defendants' audits and criticized the methods used by a number of plaintiffs' experts, particularly Robertson and Breed. See Feb. 8 Tr. at 96-166. Dr. Freeman testified that Breed's methods "do not meet the minimum standard for statistical work and they're fatally flawed from a statistical point of view." Id. at 108. Dr. Freeman also criticized the audit tool used by Dr. Robertson, id. at 128-132, and stated that the methodological shortcomings of Dr. Robertson's process undermined the validity of his conclusions, id. at 135-144.

Plaintiffs presented their own statistics expert to testify to the validity of Dr. Robertson's methods. While Dr. Catarina Kiefe emphasized that "[a]ssessing quality of health care is an inexact yet necessary and ubiquitous science," Pl.Ex. 1626; see Feb. 4 Tr., she endorsed the validity of Dr. Robertson's methods and results, id. at 15-21. She also positively reviewed Dr. Jenison's evaluations. Id. at 21-23. As a frequent reviewer for academic and medical journals, she testified that Dr. Robertson's research is of publishable quality in quality improvement literature. Id. at 30.

It is determined that defendants' Daubert-esque objections to plaintiffs' experts are without merit for three reasons. First, an expert's evaluation of a prison system's quality of medical care, use of force, or protection of inmates is not the type of testimony that necessarily implicates Daubert's requirement of scientific methodology. Second, even statistically valid methodology cannot cure faulty underlying data. Third, even if plaintiffs' experts' methodology is statistically invalid, such a determination goes to the weight of the evidence, not its admissibility. Each of these points will be addressed in turn.

First, Federal Rule of Evidence 703 provides experts with "wide latitude to offer opinions, including those that are not based on firsthand knowledge or observation." Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 592, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). "Trial courts have broad discretion in rulings on the admissibility of expert opinion evidence. . . ." Hayter v. City of Mount Vernon, 154 F.3d 269, 273-74 (5th Cir. 1998); see also Eiland v. Westinghouse Elec. Corp., 58 F.3d 176, 180 (5th Cir. 1995).

The defendants rely on several Fifth Circuit cases to argue that the plaintiffs' experts' methodology is fatally biased and unscientific. In one, Watkins v. Telsmith, the Fifth Circuit held that the application of Daubert in determining admissibility of expert testimony is not limited to "scientific knowledge" and expert testimony based on "novel" scientific evidence. 121 F.3d 984, 991 (5th Cir. 1997). The court in Watkins urged that a district court should "evaluate the methods, analysis, and principles relied upon in reaching the [expert] opinion." Id. In a second case, Moore v. Ashland Chemical Inc., the Fifth Circuit held that the proponent of expert testimony "need not prove to the judge that the expert's testimony is correct, but she must prove by a preponderance of the evidence that the testimony is reliable." 151 F.3d 269, 276 (5th Cir. 1998). This court is guided by these principles.

In Watkins, the Fifth Circuit affirmed the exclusion of the testimony of a purported expert who the district court found to possess little expertise of the sort required by the case. 121 F.3d 984, 988 (5th Cir. 1997). In Moore v. Ashland Chemical Inc., the Fifth Circuit upheld the exclusion of evidence from a doctor who opined about the causal relationship between a worker's exposure to industrial chemicals and his pulmonary illness. 151 F.3d 269, 278 (5th Cir. 1998). The exclusion of expert testimony in each of these cases is inapposite to the instant action. First, both of those cases concern expert testimony that addresses causation. Causation is not at issue in the present case. Second, in each case the expert testimony was excluded because, among other things, the expert did not have the expertise required by the circumstances of the case. Moore, 151 F.3d at 279; Watkins, 121 F.3d at 988. In the present case, expert testimony considered by the court came from eminent experts with extensive experience in the areas in which they testified.

Expert Allen Breed's decades of experience evaluating prisons all across the country, for example, cannot be disregarded based on quibbles over his methodology. In evaluating the conditions of a prison system, having extensive experience in and around the cellblocks of this country is a paramount qualification.

Defendants contend that, in the absence of statistical proof of violations in TDCJ-ID units, the plaintiffs have not and cannot show system-wide violations. This reliance by the defendants on statistics as the sole methodology for establishing system-wide violations in TDCJ-ID is misplaced. First, statistics has its limits when used in the courtroom. Whereas in science statistics plays a deductive role, for example, to test hypotheses that are developed from testable theories, in law statistics occupies more of an inductive role. The result, of course, is that underlying assumptions of long-run frequency and infinite reproducibility are usually absent from the courtroom. This can serve to limit the effectiveness of statistics in the courtroom. Second, statistics can be misleading and even misused. Nevertheless, the court agrees that statistics as a methodology can be useful. Statistical procedures can increase the court's confidence in making inferences from a given set of data. However, this is an issue of weight, rather than admissibility of a given methodology. Statistical models are simply not the only method for making general inferences from specific data.

Second, it seems inefficacious to reject inferences based on disagreements regarding the statistically validity of their authors' methods, when the underlying data on which such inferences are being made has been called into question as well. The self-reported numbers of use of force and sexual assaults quoted by defendants, for example, are almost certainly incomplete. It would be naive to assume that all of the sexual assaults in prison are reported to authorities. The medical charts reviewed by Dr. Robertson, Dr. Adams, and Dr. Ducate, for example, are subject to human error and neglect as well. An axiom of statistical analysis is that utility of statistically valid conclusions is wholly dependent on the trustworthiness of the underlying data.

Third, in a situation such as this, statistically valid methodology in the development of an expert's opinion is a luxury, not a necessity. While the court would be better served by having before it valid and reliable studies and conclusions, the court is also able to assess the credibility of less carefully verified inferences. The fact that the inferences may be based on a nonscientific sampling of records does diminish their weight in relation to a statistically validated inference. It does not, however, require the evidence be disallowed. The fact that 30 records show excessive uses of force does not change because the records were selected non-randomly.

Based on its own "logic and judgment," as mandated by the Fifth Circuit, this court has evaluated and found to be sound the methods, analysis, and principles relied upon by the plaintiffs experts in reaching their opinion. Ruiz, 679 F.2d at 1133. See Watkins, 121 F.3d at 991. Once again, "the human events of the past cannot be sampled so perfectly as to permit a statistically valid analysis." Id. The defendants' criticisms of the plaintiffs' experts' methods of gathering data and making inferences have been factored into the credibility of the experts' conclusions.

This court has not relied upon unsupported speculation by either party. See Moore v. Ashland Chemical Inc., 151 F.3d at 275 ( quoting Daubert, 509 U.S. at 589-90, 113 S.Ct. 2786). Nor has this court avoided scrutinizing the expert witnesses of either party because of the experts' experience and training. Watkins, 121 F.3d at 991. Rather, the plaintiffs' expert testimony is precisely the type of testimony contemplated by the Rules of Evidence.

C. Constitutional Violations As Legal Judgments

The defendants also assert, in what they characterize as their most serious charge, that "none of the experts set forth the proper constitutional standard or explained how it was violated." ( Def.'s Mot. for J., 9-10.) As defendants are no doubt aware, the court does not rely on expert testimony to establish constitutional standards. "[O]pinions of experts as to desirable prison conditions suffice to establish contemporary standards of decency. . . ." Rhodes, 452 U.S. at 348 n. 13, 101 S.Ct. 2392. Nevertheless, such opinions "`simply do not establish the constitutional minima.'" Id. ( quoting Bell v. Wolfish, 441 U.S. at 543-44 n. 27, 99 S.Ct. 1861, 60 L.Ed.2d 447 (1979)). Whether prison conditions are sufficiently harmful to establish an Eighth Amendment violation is a purely legal determination for the court to make. See Hickey, 12 F.3d at 756 ( citing Hudson, 503 U.S. at 6, 112 S.Ct. 995); Madrid, 889 F. Supp. at 1246. In fact, experts are entitled to little weight in determining whether a condition is "cruel and unusual punishment" under the Eighth Amendment. See Toussaint v. McCarthy, 801 F.2d 1080, 1107 n. 28 (9th Cir. 1986); Madrid, 889 F. Supp. at 1159. Nevertheless, "expert opinion may be properly considered in assessing the effects of challenged conditions or practices." Id. at 1159 ( citing Helling v. McKinney, 509 U.S. 25, 36, 113 S.Ct. 2475, 125 L.Ed.2d 22 (1993)).

VIII. MEDICAL AND PSYCHIATRIC SERVICES

A. Compendium

There can be no doubt that vast improvements in TDC's provision of medical and psychiatric care to inmates have been made over the course of this litigation. Where there once were actually inmates performing surgery on other inmates, there now are two of the State of Texas's finest medical teaching institutions, the University of Texas Medical Branch at Galveston (UTMB) and Texas Tech University Health Sciences Center, giving treatment to inmates. Unfortunately, a number of systemic deficiencies often prevent the high quality health care provided at UTMB and Texas Tech's hospitals from translating into sufficient medical treatment for inmates in the cellblocks across the state. Many of the inmate witnesses' accounts of their grossly inadequate medical and psychiatric treatment were both credible and disturbing. Along with the plaintiffs' experts' health care audits, they reveal a health care system that stops well-short of providing inmates adequate medical care.

The constitutionality of the system is saved, however, not by TDC's provision of health services, but rather by the extremely high bar set by the Supreme Court of the United States for inmates attempting to prove that their medical care is unconstitutional. With the evidence before the court, the plaintiffs were simply unable to meet this inordinately high standard of proving defendants' "deliberate indifference" to their physical and mental health needs.

B. Medical Testimony and Findings of Fact

In 1994, the Texas Legislature created the six-member Correctional Managed Health Care Advisory Committee to take over management of TDCJ's health care delivery system. Tex. Gov't Code 501.059. To fulfill its mandate to establish "a managed care network of physicians and hospitals . . ." and to "integrate the managed health care provider network with the public medical schools of this state," id., the committee contracts with UTMB and Texas Tech for the provision of health care. See Pl.Ex. 1607 at 6. Through this process, the day-to-day operations of the medical system are delegated to UTMB and Texas Tech Health Science Center. Linthicum, Jan. 15 Dep. at 8-11. With almost no exceptions, the health care provided to inmates by these schools' teaching hospitals was praised during the fact-finding hearing. The high-quality of the health services provided by those in-patient hospitals appears to be exemplary.

In an attempt to see beneath the veneer of well-developed policies and infrastructure of the TDCJ health care system, plaintiffs' medical experts, through a number of audits, evaluated the actual quality of care received by inmates in the various units of the system. These experts focused on the most prevalent causes of death in TDCJ-cardiac disease, cancer, and HIV. Care for diabetes was also examined.

In sum, medical experts reviewed charts of living inmates for 106 cardiac patients, 98 diabetes patients, and 62 HIV-positive patients. Death charts from several periods of time were evaluated, including 51 cardiac and cancer death charts, and 33 HIV death charts from 1998. Having both performed a number of audits themselves, and having reviewed the audits performed for this litigation by other experts, Dr. Robertson and Dr. Jenison testified as to the audits' results.

Defendants responded with expert witnesses of their own, some of whom are intimately involved in the design and implementation of TDCJ-ID medical policies. Defendants called Dr. David R. Smith, President of the Texas Tech Health Science Center, to the stand to respond to plaintiffs' allegations regarding medical care. Dr. Smith explained the structure of the prison health care system, and the medical schools' relation to TDCJ-ID. See Def. Ex. 57, 63. He took the court through a long list of exhibits, pointing out the policies and accreditations of the prison's medical services. See Def. Ex. 67, 3, 59, 5, 38, 66, 43, 64, 65, 64, 68-71, 99. On cross-examination of Dr. Smith, it was revealed that he had spent only eight to ten hours in TDCJ prisons in his life and had discussed medical care with very few inmates or staff in the units. Feb. 2 Tr. at 43-44. He also stated that he had reviewed only a couple of medical charts of inmates. Id. Dr. Smith acknowledged on redirect that the best use of his skills was "to take a look at that global system, the policies and procedures that are in place, to effectively work with the Legislature and other State agencies on behalf of the criminal justice healthcare system. . . ." Feb. 2 Tr. at 105-106.

Defendants also called Robert Jones, M.D., to give a general assessment of the state of medical care in TDCJ-ID and rebut plaintiffs' experts' assertions. Dr. Jones spent time in Texas prisons as a surveyor for the National Commission on Correctional Health Care. Defendants further called TDCJ-ID employee Dr. Glenda Adams to discuss individual inmates' medical charts. See Feb. 3 Tr.

1. Quality of Care Audits-Cancer, Cardiac Disease, and HIV

John Robertson, M.D., spent in excess of 300 hours evaluating the quality of medical care in the Texas prison system by analyzing the data culled from death charts of 59 inmates who died in 1998. Dr. Robertson identified what he believed were a number of deficiencies in the health care provided inmates in TDCJ-ID.

Using a standardized audit tool, Dr. Robertson reviewed 35 of the 40 reported "cardiac" deaths in TDCJ in 1998, and 16 of the 33 "cancer" deaths reported-an audit representing 56% of the deaths from those two categories. Pl.Ex. 1510 at 4. Each chart was rated on a scale of one to five, with one being "very poor," two being "poor," three being "satisfactory," four being "good," and five being "excellent." Id. at 5. Furthermore, each death chart was evaluated on the basis of whether or not the death was "preventable." Id. Dr. Robertson's analyses found that 16 of the 59 deaths were "preventable" or "possibly preventable." Jan. 28 Tr. at 151-152. Nine of those sixteen deaths were evaluated as having had inadequate oversight of nursing or mid-level staff and/or inadequate follow-up. Pl.Ex. 1510 at 10.

From the cardiac cases, Dr. Robertson concluded that "viewed collectively, [they] identify a consistent problem in multiple medical encounters of failure to adequately evaluate significant and serious disease processes. This occurred in terms of level of medical practitioner, timeliness, and/or appropriateness of subsequent treatment and referral." Pl.Ex. 1510 at 7. From the cancer cases, of which two were considered "potentially preventable," Dr. Robertson concluded that "[t]hese two individuals experienced significant delays in definitive diagnosis and treatment that may have negatively impacted upon prognosis." Id. at 8. In the "Summary Findings" of his report, Dr. Robertson concluded:

This review of deaths presents a troubling pattern of systemic problems in the health care delivery to inmates in the Texas Department of Criminal Justice. Of a total of fifty-nine charts reviewed, twenty (34%) were found to have received poor to very poor medical care. To assure that this finding was balanced a very conservative approach was taken, namely, only when a significant clinical outcome could be demonstrated was a score lower than 3 [(below "satisfactory")] given. Of particular concern was the finding that sixteen of the deaths (27%) could be deemed as `preventable'. Among the potentially preventable deaths, nine of sixteen had serious concerns regarding oversight and/or follow-up while seven of the 43 deaths that were not evaluated as preventable demonstrated the same deficits.

Dr. Jenison found that while many inmates receiving HIV care at the Stiles and Clements units were receiving the recommended combination therapy, many inmates were not. Dr. Jenison was particularly concerned that the Texas prison health system was not effectively lowering the viral load of its HIV-positive inmates. In response, defendants' witness Dr. Smith testified that TDCJ-ID's HIV drug therapy was still an evolving science and that experts from the Centers for Disease Control and the Texas Department of Health were brought in to establish TDCJ-ID's policies. Feb. 2 Tr. at 3. Dr. Smith also testified that the American Medical Association had approved TDCJ's HIV drug policy. Id. at 3. Defendants emphasized the problems inherent in ensuring inmate compliance with the complicated combination therapy drug regimens and further argued that the AIDS numbers in Texas prisons are improving. See Feb. 1 Tr. at 220-223 and Def. Ex. 99.

In contrast to Dr. Smith's testimony regarding the declining rates of tuberculosis in Texas prisons since 1994, Feb. 2 Tr. at 5; Def. Ex. 100, Dr. Jenison found that TDCJ was poorly managing tuberculosis infection in its HIV-positive population. Jan. 25 Tr. at 88-89. He cited several cases in which prisoners with tuberculosis were exposed to other inmates and staff. Jan. 25 Tr. at 95-96, 137-38; Pl.Ex. 1524; P.Ex. 1525. One inmate, for example, was shown to be likely infectious by a UTMB test. During the time that he had symptoms of progressive tuberculosis, other inmates were exposed to his illness. Dr. Jenison maintained that such exposure mandates an immediate "public health intervention" to determine who has been exposed. Jan. 25 Tr. at 99-100.

The testimony of Dr. Robertson and Dr. Jenison was collectively criticized by defendants. Dr. Smith criticized both the methodology and the conclusions of Dr. Robertson and Dr. Jenison, asserting that the experts did not look at a large enough sample. Feb. 2 Tr. at 27-33. In fact, defendants repeatedly questioned both doctors' "sampling" procedures in the charts they reviewed. Feb. 8 Tr. at 144-148. As discussed above, defendants' expert witness Dr. Freeman addressed what he believes are the statistical shortcomings of plaintiffs' medical experts' audits. Id. at 125-145. Defendants criticized Dr. Robertson and Dr. Jenison's primary use of chart reviews as the bases for a medical care audit. See Id. at 105-107; Smith, Feb. 2 Tr. On cross-examination, defendants emphasized the narrow scope of the plaintiffs' experts' medical care audits and questioned possible bias in the way the charts were collected. Dr. Jenison acknowledged that he was unaware if the validity of his audit tool had been tested. He also acknowledged the difficulty of ensuring compliance in a prison setting. Jan. 25 Tr. at 132-135.

Dr. Jones criticized Dr. Robertson's survey of the audits on the grounds that a number of the expert surveys Dr. Robertson reviewed contained positive comments that were not found in Dr. Robertson's report. Feb. 9 Tr. at 57. Dr. Smith further disputed some of Dr. Robertson's evaluations of particular inmates' treatment. Feb. 2 Tr. at 31-32. Both Dr. Robertson and Dr. Jenison acknowledged that the population they work with in New Mexico is relatively small compared to Texas' prison population.

Defendants also argued that the review of death charts, in and of itself, is biased because the reviewer never sees those who present with bad symptoms and do not die, only those who do. Several experts, however, testified as to death chart reviews' appropriateness as indices of a system's quality of care. See Kiefe, Feb. 4 Tr. Defendants point, however, in so far as it implies that a more thorough analysis of medical care of all patients would be ideal, is well taken. While a broader study would have been useful, it is determined that Dr. Robertson's analyses did reveal significant, even deadly, inadequacies in the level of care provided to ill inmates.

2. Non-physician Health Care

Plaintiffs made several inferences as to the reasons for the poor medical care in TDCJ. Dr. Robertson, in particular, contends that in TDCJ-ID too many non-physicians are delegated decision-making powers beyond their expertise. In a number of cases, for example, pharmacists seemed to be doing all the decision making for an inmates' medical treatment. Jan. 28 Tr. at 232. By its own admission, the State of Texas has been adjusting to higher inmate populations by hiring staff in less sophisticated medical staff categories. See An Audit Report on Managed Health Care at the Texas Department of Criminal Justice, Office of the State Auditor (January 1998), Pl.Ex. 1607 at 91. Dr. Robertson indicated that in the New Mexico prison system, a physician reviews and countersigns all midlevels' charts. In Texas, he testified, the medical director reviews a mere 10% of the charts. Jan. 28 Tr. at 17, 174-75. Defendants point to the improving ratio of inmates to health care providers as evidence of an improved medical care system. Feb. 1 Tr. at 175-77.

Plaintiffs contend that a number of harrowing examples of inmate illness and death may be attributed to TDCJ-ID's heavy dependence on non-physician health care givers. Inmate Arthur Heinz died at Garza West in 1998 after repeatedly returning to the infirmary with highly elevated blood pressure. During half of his visits, he saw only a nurse. Medication was prescribed several times, but no follow-up was scheduled. On his final visit, record was made of the possibility of previously unrecognized myocardial infarction, but it was not addressed or followed up. Inmate Heinz died of cardiac arrest five days later. Jan. 28 Tr. at 183-88; Pl.Ex. 1510. Inmate Brown, in the Stiles Unit, died from a stroke and myocardial infarction after a physician's assistant gave a phone order to administer two doses of a particular drug. Dr. Robertson has banned the drug in question in the New Mexico system, because it lowers blood pressure at an unpredictable rate. Jan. 28 Tr. at 176-80. Another inmate, Jason Wimberly, visited medical personnel five times in a day and a half with what Dr. Robertson asserts were obvious signs of appendicitis. According to the inmate's medical chart, only nurses evaluated him and they did not recognize the illness. On the fifth visit, inmate Wimberly was rushed to a hospital with a perforated, gangrenous appendix. Id. at 195-97, Pl. Ex. 683.

Defendants do not deny that much of TDCJ's health care is provided by non-physicians. Both Dr. Smith and Dr. Ducate argued, however, that midlevels and nurses are well-supervised by physicians, and that no care giver provides more sophisticated care than they are allowed to under state law. Feb. 2 Tr. at 20-22.

3. Inadequate Evaluation and Referral

Plaintiffs presented the court with a long list of cases, like inmate Wimberly's, that appear to form a pattern of inadequate evaluation and referral. William Bishop's broken arm went untreated for two weeks in the Hughes unit. Pl.Ex. 87. Alta Mae Jacobson, in the Sycamore Unit, also waited two weeks for treatment of her broken wrist. Pl.Ex. 249. In another case, defense witness Dr. Glenda Adams acknowledged that a prison doctor failed to diagnose a heart attack of Loren Kelly, shortly before inmate Kelly died. Feb. 3 Tr. at 40-41. These are just a few of the disturbing, and largely unrefuted, stories of inadequate evaluation and referral presented by the plaintiffs.

Dr. Robertson found that the referral system has trouble delineating between levels of need and is characterized by unacceptable delays. Jan. 28 Tr. at 217. Evidence of the negative effects of the delays was frequently presented. Pl.Ex. 593; Pl.Ex. 1111, 1112; Pl.Ex. 1466. A routine referral can take one to six months in TDCJ-ID. An expedited referral can take up to one month. Jan. 28 Tr. at 215. Dr. Robertson discovered one case at Stiles Unit involving an inmate who had been bleeding internally for one week waiting for the referral to come through. Id. at 215-216.

4. Failure to Follow-up

Dr. Robertson also credibly faulted the prison health system for lack of follow-up for at-risk patients. A growing mass in Gilberto Cristo's lung was not diagnosed, despite a number of x-rays, until it was inoperable metastatic cancer. Id. at 231. Patients receiving certain cardiac disease treatments, who are thereby prone to bleeding and illness, were not scheduled for follow-up. Jan. 28 Tr. at 231. One patient in particular was diagnosed as having a possibly life-threatening growth in his kidney and was scheduled for surgery in two to three weeks. This inmate was not sent for the procedure for six months, by which time the growth had led to metastatic disease. The inmate died. Id. at 219-220; Pl.Ex. 1510. These are simply a few of Dr. Robertson's examples of poor follow-up in the TDCJ-ID health system.

5. Staff Indifference

Dr. Robertson's analysis revealed numerous accounts of staff indifference. Inmate Collins Gentry showed obvious symptoms of an acute heart-attack, but was sent back to his cell. Jan. 28 Tr. at 190-92; Pl.Ex. 1509. Inmate Ophelia Rangel spent five days not eating and suffering from psychotic episodes and severe diarrhea, but she was not treated. She died. Id. at 192-94; Pl.Ex. 478, 479, 1668, 1510. Michael Bias, who fell into a coma after a suicide attempt, was left to lie in one position, so that he developed infected pressure sores that led to massive loss of skin, breakdown of muscle, and kidney failure. Jan. 28 Tr. at 199-201; Jan. 29 Tr. at 14-15; Pl.Ex. 1603, 1604. Presenting what Dr. Robertson considered to be obvious symptoms of metastatic prostate cancer, inmate Robert Lee was treated by a physician assistant as malingering. The inmate finally became paralyzed and was sent to emergency radiation therapy to regain control of his legs and bladder. Jan. 28 Tr. at 217-218; Pl.Ex. 1510.

Dr. Jenison reviewed the death chart of one 24 year-old woman who entered prison shortly after a free world hospitalization for meningitis. Medical personnel at the prison also were aware of her having HIV. She died after two months in prison. No therapy for the meningitis was ever initiated. Jan. 25 Tr. at 56-57. Another female inmate, Layovonda Alexander, was sent back to her cell after reporting significant weight loss and coughing up blood-symptoms of her HIV infection. Pl.Ex. 1510 at 29-30. Rather than immediately treat her, a nurse and midlevel documented that she was "refusing to work" and was made to return to work. Layovonda Alexander died within five months. Pl.Ex. 1510, at 29-30.

6. Poor Treatment of Diabetes

Problems with treatment for diabetes were repeatedly brought to the court's attention. Perhaps the most revealing was that of inmate DeWayne Webb who testified that he has recurring problems getting an escort to take him to the unit infirmary to receive the twice-daily insulin shots prescribed by medical staff. Jan. 25 Tr. at 241. Webb stated that he has filed a number of grievances about this problem, but to no avail. Jan. 25 Tr. at 228. See, e.g., Pl.Ex. 604. Webb's highly fluctuating and poorly managed blood sugar levels have subjected him to a number of seizures. Jan. 25 Tr. at 239, 256. Defendants, through Dr. Adams, respond that inmate Webb has at times failed to cooperate with medical staff in developing diabetic care. Feb. 3 Tr. at 81-82. Dr. Adams also pointed out that Webb was admitted to the infirmary on four occasions so he could be more closely monitored. Id. at 84. These admissions, however, were in 1994 and 1995, while most of inmate Webb's complaints involved more recent events.

Dr. Robertson, in his evaluation, also encountered instances of poor diabetes care resulting from security staffs failure to report medical needs. Jan. 28 Tr. at 223-225; see Pl.Ex. 487, 1107, Dr. Robertson was of the opinion that diabetics suffered not only from the indifference of correctional officers, but also the failure of medical staff to follow-up with treatments. Jan. 28 Tr. at 227; Pl.Ex. 487.

With particular reference to diabetes treatment, Dr. Robertson found widespread failure to perform basic, simple tests to prevent well-known complications. Of the 98 diabetes charts he reviewed, only 29% had received an eye examination within the preceding year. Such a yearly examination is needed to prevent blindness. Jan. 28 Tr. at 201-202; Pl.Ex. 1508. Twenty-eight percent of those charts reviewed had not had a standard hemoglobin A1C test in the preceding six months. Jan. 28 Tr. at 202-203. Annual foot examinations, also important aspects of preventing serious complications with diabetes, had apparently not been performed in 68 of the 98 charts. Jan. 28 Tr. at 204-205. These results represent only some of Dr. Robertson's findings. See Id. at 205-210. It should be noted that Dr. Robertson's physiological conclusions regarding a number of inmates were criticized by defense expert Dr. Smith. See Feb. 2 Tr. at 90.

7. The Lack of Satisfactory Communication Between Hospitals and Prisons

Dr. Robertson and Dr. Jenison both found that while inpatient treatment by the university hospitals was generally good, a dangerous lack of communication between the hospitals and the unit medical care facilities put inmates at risk upon their returning to prison. See Pl.Ex. 1506; Jan. 25 Tr. at 57-59. For example, 29 year-old Edward Coleman had difficulty receiving the suppressive antibiotic therapy prescribed by UTMB Galveston, to keep his toxoplasma encephalitis from recurring. He died after 18 months in prison. Id. at 57. Dr. Jenison did note, however, that the communications problems had improved between the time he performed his analysis of 1994-1996 deaths and 1998. Id. at 93. In the worst cases, the disparities between treatment in the hospital and in the prison unit seems to have been not the result of miscommunication, but of unit-based doctors ignoring the recommendations made at the hospitals. Jan. 28 Tr. at 221; Jan. 25 Tr. at 93-95; Pl.Ex. 1490, 1511.

Dr. Jenison found that poor access to medications was a recurrent problem throughout the system. Jan 25 Tr. at 57-59, 147-48. Given the risks of relapse and the development of drug-resistant strains, the problem has particularly severe consequences for HIV patients. Id. at 130-31. And, Dr. Jenison noted a number of inmates erratically received drug therapy. Id. at 132; Pl.Ex. 1490.

Further complicating the provision of medical needs are the sometimes necessary lockdowns of prison units. Unfortunately, these lockdowns often interrupt medical services, to the detriment of inmates' health. See Jan. 27 Tr. at 58, 63-65; Pl.Ex. 157, 1625, 916, 1510 at 15-16, 483, 354, 722, 713, 719, 168.

Other times the "communication" problems seems to be between medical staff and security staff. Inmate Webb (discussed previously) was housed on the third floor despite a medical housing restriction to the first floor because of his seizures. Jan. 25 Tr. 258-259; see also Pl.Ex. 127. Tonya Fountain, who was being used to train tracking dogs at the time and wearing a protective suit, broke her leg after being told to jump out of a tree into the dog pack by a prison "boss." Jan. 25 Tr. 167-168. She had surgery to insert a pin into her leg. Not only did she have to endure the pain and indignity of pulling herself on and off buses on her buttocks (because she was not allowed crutches on the bus), but also her mobility restrictions were not honored at the unit, and she was required to walk to chow hall and pill lines. On one occasion, her crutches slipped and she re-injured her leg so badly that she had to have another reconstructive surgery. Jan. 25 Tr. at 178-185.

Revealing a number of the most disquieting problems of the medical system, inmate Fountain's problems did not end there. In October 1997, inmate Fountain was assigned to the Murray unit, where she began to suffer severe internal pain and abnormal vaginal bleeding. She waited from October of 1997 to April of 1998 for an obstetrical/gynecological appointment. Id. at 188-189. When she did finally have an appointment, she was placed on birth control pills and the bleeding worsened. Id. at 189-190. On May 12, 1998, inmate Fountain passed a mass of tissue "about the size of a tennis ball." Id. at 191-192. She was instructed to bring the tissue to the infirmary with her, where she was asked if the tissue was "a fetus." Id. at 192. Two weeks after her next OB/GYN appointment on May 27, 1998, her uterus and both ovaries were removed. Id. at 193-194. Although she testified before the court that she had wanted to have children, defense evidence shows that inmate Fountain apparently signed a release stating that she wanted the surgery and did not plan to have children. See Adams, Feb. 3 Tr.

8. Medically Contraindicated Work

Plaintiffs brought before the court a number of inmates whose medically-based work restrictions were apparently ignored by correction officers on the units and in the work fields. See Pl.Ex. 299, 1119-1125, 1510, 1237, 377, 1696. Alfa Turner, for example, testified about being made by the "boss" to lift heavy objects after her wrist surgery. Jan. 28 Tr. at 110. Quite often such problems arose when inmates on psycho tropic and photosensitive drugs were sent to work in the hot sun, which physicians described as contraindicated. Pl.Ex. 195, Pl.Ex. 1290, 1293, 1294, 105, 1180-1183, 684-685, 1366-1367, 1225, 1226, 1229, 1716. One such situation involved Timothy Stonecipher who, by defendants' own admission, was misassigned to work in the fields despite his drug regimen. Inmate Stonecipher received disciplinary referrals for refusing to work, although his own statement indicated that he could not keep up. Pl.Ex. 541. Two-weeks after a psychologist wrote a memorandum to the warden about the problem, Mr. Stoncipher had a stroke-like episode in the fields. See McGhee, Jan. 25 Tr. at 207.

Defendants emphasized TDCJ's policies against assigning medically contraindicated work. See Feb. 1 Tr. at 163-167; Feb. 2 Tr. at 82-87. Neither Dr. Smith, Deputy Director Cockrell, nor expert Dr. Jones knew of any security personnel who had been disciplined for violating such policies, although the Director could remember one. Feb. 2 Tr. at 118-119; Feb. 10 Tr. at 64; Feb. 9 Tr. at 98-99.

9. The Lack of Self-Monitoring

Citing TDCJ's failure to monitor its own problems and failure to respond efficiently to diagnoses, Dr. Robertson concluded that the medical problems in TDCJ-ID were, in fact, system-wide. Jan. 28 Tr. at 115-18, 234-239. An independent State Auditor's report confirms Dr. Robertson's and Dr. Metzner's assertions that the TDCJ-ID medical system has no formal tracking and reporting system to monitor and evaluate health care. See Pl.Ex. 1607 at 29. According to that report, "[a] threshold of acceptable compliance and defined measurable performance standards has not been established for the operational review process." Pl.Ex. 1607 at 30. Dr. Smith acknowledged the general accuracy of this report. Feb. 2 Tr. at 61. Defendants respond that a number of self-monitoring processes are in place, and that plaintiffs' experts did not consider them. Id. at 26-28.

One of the few tangible examples of self-monitoring by the medical system before the court was a 1996 peer review of 24 HIV-related deaths at the Stiles Unit. See Pl.Ex. 1523. Of the Stiles deaths reviewed, two-thirds were found to be "improper." Feb. 2 Tr. at 71-72; Jan. 25 Tr. at 66. Dr. Smith acknowledged that this study reflected something seriously wrong at the Stiles Unit in delivery of care in 1996. Feb. 2 Tr. at 72-73. (As plaintiffs point out, Stiles was an NCCHC accredited institution at the time.)

10. Accreditation

Defendants implied, and at times explicitly stated, that their medical units' accreditation by the National Commission on Correctional Health Care (NCCHC) effectively proves the medical care systems' constitutionality. Jones, Feb. 9 Tr. at 77-78; see Smith, Feb. 2 Tr. at 26-27. It is true, and defendants justifiably take great pride in the fact, that all of the medical units at TDCJ-ID are accredited by the NCCHC. The NCCHC recently cited TDCJ as having five of the ten best correctional health care facilities in the United States. As the highest ranking administrator for Health Services in TDCJ testified, however, the standards by which NCCHC evaluated the medical units did not actually measure the actual standard of medical care provided by the prison system:

The National Commission on Correctional Health Care standards to me looked more at facilities and processes to see if they're adequate for health care delivery. They fall short in actually looking at quality of care. They do have the one standard that requires systems to have a quality improvement/quality management system, but that's the extent of it.

Further undermining defendants' attempt to use NCCHC accreditation as a proxy for a certification of the constitutionality of its medical care is the fact that at least two of the plaintiffs' experts who testified about profound shortcomings in the quality of care in TDCJ-ID also work as NCCHC accreditors. Thus, some of the very experts who accredit these institutions based on NCCHC standards may have serious misgivings about the actual level of medical care received by the prison population.

One of the recurring themes of the evidence presented in the hearing was that constitutional policies do not necessarily ensure constitutional practices. While NCCHC accreditation does bolster defendants' claims that its medical care system is functioning constitutionally, the accreditation simply cannot be dispositive of such a conclusion. It is beyond question that the true measure of a medical care system's constitutionality is not its brilliantly crafted policies and state of the art facilities, but its accessibility by, and quality of service to, real people in need of actual medical service.

C. Psychiatric Testimony and Findings of Fact

In a fashion similar to their physical medical evidence, plaintiffs used expert witnesses and audits of the psychiatric services in TDCJ-ID to argue that mental health care falls below a constitutional standard. Both Jeffrey L. Metzner, M.D., P.C., and Robert Stellman, M.D., evaluated the quality of psychiatric care in TDCJ. Dr. Dennis Jurczak also testified regarding psychiatric care.

Dr. Metzner assessed the prison system's psychiatric services by reviewing prison policies and reviewing audits of various psychiatrists concerning mental health services at thirteen of TDCJ-ID's institutions. Dr. Metzner also performed his own audit of the Estelle Unit. Further, he reviewed reports relevant to inmates who had experienced heat-related illnesses. Pl.Ex. 1505.

Dr. Metzner concluded that the problems he found at the Estelle Unit, including the restrictive formulary medications list, the young, inexperienced guards, the lack of monitoring for compliance, and lack of analyses of multiple admissions to psychiatric wards, were encountered by all of the experts whose work he reviewed. Feb. 3 Tr. at 190-195. In general, Dr. Metzner found "systemwide deficiencies" in the mental health care system.

The problems, as related by Dr. Metzner, include "not recognizing or minimizing symptoms indicative of major mental illnesses by either over-diagnosing malingering or `no Axis I diagnosis.'" Pl.Ex. 1505. Dr. Metzner detailed one case (mentioned above) in which a woman who was incontinent and smearing feces-signs of both organic brain disorder and mental disorder-was not diagnosed and treated. The inmate died. Feb. 3 Tr. at 206-207; Pl.Ex. 1505. Defendants represented that health care workers were fired following that incident. Ducate, Feb. 5 Tr. at 146; Feb. 6 Tr. at 52-53; Jan. 29 Tr. at 5.

Dr. Metzner was particularly bothered by the apparent practice that Axis II diagnoses (developmental disabilities, mental retardation, etc.) may be used as a means of avoiding treating inmates, in violation of policy. One inmate, for example, who had a history of smearing himself with feces, complaining of auditory and visual hallucinations, and claiming to be the Messiah, was initially diagnosed as schizophrenic. Pl.Ex. 856, 855, 853. However, a Dr. Taylor later determined the inmate was malingering and wrote, "I feel certain that we will have some grandiose display of abnormal behavior, perhaps reaching a magnitude of once again smearing feces. I would be interested to observe if the patient consumes his feces or whether he just smears his feces." Pl.Ex. 855 (Bates 157866-157867).

In a related point, Dr. Metzner found a significant over-diagnosis of malingering. He related the case of Adan Garza in particular to illustrate this point. Feb. 3 at 218-219. Inmate Garza entered the system with a history of suicide attempts, self-mutilations, hallucinations, and hospitalizations. Pl.Ex. 172. His medications were discontinued and he was diagnosed as having no Axis I illness. After a brief visit to Skyview, he was discharged with Dr. Tchokoev recommending no medication and heavy work in the field. The same day he returned to Beto, he cut himself and then attempted to hang himself. He is now in a vegetative state. Id. Inmate Paul Ferguson testified about the events he observed on the day that inmate Garza cut himself. Inmate Ferguson alleges that Officer Smith told inmate Garza she would call for psychiatric help, but she apparently failed to do so, as no help came. Jan. 22 Tr. at 4-10.

On February 9, 1998, inmate Edward Coleman went to Skyview after a long history of psychiatric treatment for a number of Axis I and Axis II diagnoses, both in the free world and TDCJ. Once there, Dr. Taylor discontinued his medication and wrote, "[t]his patient has a history of acting out at this facility when he is `found out' and is aware of the fact that he will be returned to his unit of assignment." Pl. Ex. 719 at 144581. By the end of the month, he had returned to Crisis Management and received two more, different Axis I diagnoses. Id. at Bates 144579 and 144578.

Evidence was presented before the court that called into question the correctional officers' ability, and willingness, to recognize psychiatric needs. One inmate, who had a history of self-mutilation, gave unrefuted testimony that a guard provided a razor blade when the inmate threatened to self-mutilate. Jan. 27 Tr. at 31-32. According to the inmate, he cuts himself "because I can't handle my depression, or when I get mad and I cut, it makes me feel better." Id. at 29. Several experts testified that self-mutilating inmates need mental health intervention and that simply returning them to their cells until they cut again is not appropriate. Jurczak, Jan. 26 Tr. at 177.

Dr. Metzner also found a system-wide practice of inexplicably changing diagnoses. Feb. 3 Tr. at 213-214. He identified the source of these problems as including staffing shortages, inadequate assessment procedures (like cell-side assessments), staff education issues, and probable clinical biases. Pl.Ex. 1505. Dr. Jurczak questioned the adequacy of TDCJ-ID's policy of primarily using medications to treat those mentally ill patients who are treated. Pl.Ex. 1493. He also disagreed with the use of older and less expensive (and far less effective) drugs. Id.

Regarding heat-related illness, Dr. Metzner found that three of the 16 people who became ill because of heat had died. He was particularly disturbed that inmates using psycho tropic drugs, which can be photosensitive, were not being more aggressively protected from sun exposure. Feb. 3 Tr. 201-203.

On cross-examination, Dr. Metzner was questioned about his involvement in the accreditation of a TDCJ facility for the NCCHC. Dr. Metzner emphasized that he was one of a team of people to collect data and that NCCHC does the accrediting. Feb. 3 Tr. at 239-240. Dr. Metzner stated, "I have been involved in accreditations in which I thought the place was bad and it got accredited." Id. at 242. Dr. Metzner also admitted that he had not performed a systemwide statistical analysis of the mental health needs of the inmates, but he emphasized that such an analysis is difficult, if not impossible, when the system does not have record of its own caseload. Feb. 3 Tr. at 265.

Plaintiffs' expert Roberta Stellman, M.D., a Board Certified psychiatrist who works under contract for the New Mexico Department of Corrections, came to conclusions very similar to those of Dr. Metzner. She completed audits of psychiatric care in the Gatesville, Jester IV, Skyview, Stiles and Woodman Units of TDCJ- Id. Pl.Ex. 1514, 1515, 1516, 1517, and 1518. (Her work served as a partial basis for Dr. Metzner's conclusions.) After interviewing staff, reviewing policies, touring facilities, reviewing medical records, and interviewing inmates at some of the units, Dr. Stellman made separate findings in a number of areas, including the adequacy of the medical records, the adequacy of response to emergencies, the management and assessment of inmates in isolation, and medication management. See id.; Feb. 1 Tr. Dr. Stellman expressed disapproval of nurse practitioners making serious diagnostic and prescription decisions. She also gave detailed testimony about the inadequacy of the older drugs used to treat mental illness in TDCJ. Feb. 1 Tr. at 20-30.

As summary conclusions, based on all of her audits, Dr. Stellman had "major concerns about patients being underdiagnosed, not having access to psychiatric assessment, even after repeated requests by the inmates themselves, so that there is a failure to identify and treat treatable mental illnesses, as well as conditions that are resulting in severe functional impairment throughout the system." Feb. 1 Tr. at 86. She also found that mentally ill patients were being inadequately treated, in that they were not being seen frequently enough, or were prematurely dropped from care. Id. at 86. She criticized the fact that nonformulary antipsychotic medications were only available at the hospital units themselves, and that even then older drugs were being used. Feb. 1 Tr. at 88. On cross-examination, Dr. Stellman acknowledged that she did not interview inmates on all of the units she evaluated. Feb. 1 Tr. at 100-101. She acknowledged her lack of knowledge about the NCCHC and the American Correctional Association. Id. at 112.

Defendants responded to plaintiffs' accusations of unconstitutional psychiatric care primarily through the testimony of Suzanne Ducate, M.D., Director of Mental Health Services for UTMB/TDCJ Correctional Managed Care in Huntsville. Dr. Ducate began with an overview of the psychiatric services system and process. She testified that defendants have in place a number of psychiatric screening processes, and that within two-weeks of arrival, all new inmates are screened for mental health needs. They are further evaluated each time they are moved to a different unit. Feb. 5 Tr. at 78-82.

Dr. Ducate made frequent references to the Final Judgment, providing examples of policies and practices that ostensibly bring the system in compliance with this court's previous orders. Her testimony showed that a broad range of psychiatric services are at least theoretically available in TDCJ. See Ducate, Feb. 5 Tr. She praised the NCCHC standards that TDCJ institutions achieved for accreditation. Id. at 106-107. She testified to a whole range of policies, including infection control processes, staff credentialing, staff recruitment, and quality improvement planning processes. Id. at 104-110. Dr. Ducate led the court through a series of exhibits documenting these policies and practices. See Def. Ex. 202, 203, 48, 75, 72, 76, 66.

Dr. Ducate also directly rejected Dr. Stellman and Dr. Metzner's evaluations of the psychiatric care system:

So these individuals, I don't think, had a clear picture of just how extensive our inpatient services are, because I don't understand how they could state that we are not caring for patients with the state-of-the-art facilities that we have, with the extensive programming we have, with the inpatient facilities, the special sheltered housing programs, the stepdown units, the extensive development we have done of specialized programs like the psychiatric inpatient program.

At least one plaintiffs' expert found the psychiatric staffing level in TDCJ-ID "wholly inadequate." Jurczak, Jan. 26 Tr. at 171. That expert found that one psychiatrist serves about 7,000 inmates at the Michael, Terrell, and Powledge Units. Id. According to Dr. Ducate, one of the reasons that plaintiffs' complaints about the number of psychiatrist in the TDCJ-ID system are faulty is that TDCJ-ID depends on telepsychology to allow psychiatrist and psychologists' contact with inmates all over the state. Id. at 110.

It was also ascertained from Dr. Ducate that, even though she was treating psychiatrist for three institutions before she took her current position, she spent only one day a week at one unit, has spent only one day total at another, and has never been to a third. Feb. 6 Tr. at 7-8. Furthermore, Dr. Ducate explained that "based on the job description of the psychiatrists in our system, a caseload of up to 700 patients could easily be cared for by one psychiatrist. . . ." Id. at 62.

D. Legal Analysis and Conclusions

It is found by a preponderance of the evidence that the TDCJ-ID medical and psychiatric care services, while superb in their primary sources, are not adequately reaching the unit infirmaries. Medical orders in the UTMB and Texas Tech hospitals are, in many cases, ignored or neglected in the units. A number of inmates perform work or activities contraindicated by their doctors. Systemic problems of poor evaluation and lack of follow-up have been shown. Distressing incidences of staff indifference to inmates' medical needs have been chronicled. Defendants' carefully developed policies and procedures notwithstanding, it is determined that the plaintiffs' experts' assessment of poor implementation of those policies and procedures is both substantially credible and a matter of extreme concern. Simply stated, large numbers of inmates throughout the TDCJ system are not receiving adequate health care.

It is incontrovertible that the state is obligated to provide medical care for those punished by incarceration. Estelle, 429 U.S. at 103, 97 S.Ct. 285. Nor is there any question that the constitutionally mandated medical care in prison includes psychiatric care. Hoptowit, 682 F.2d at 1253. "An inmate must rely on prison authorities to treat his medical needs; if the authorities fail to do so, those needs will not be met." Estelle, 429 U.S. at 103, 97 S.Ct. 285. The infliction of pain and suffering resulting from the denial of medical care "is inconsistent with contemporary standards of decency," because "`[i]t is but just that the public be required to care for the prisoner, who cannot by reason of the deprivation of his liberty, care for himself.'" Id. at 103-04, 97 S.Ct. 285 ( quoting Spicer v. Williamson, 191 N.C. 487, 490, 132 S.E. 291 (1926)).

The standard for evaluating the constitutionality of medical care in prisons is, however, unduly low. That is, the plaintiffs' burden in this area is inordinately high. Plaintiffs must show not that defendants are merely negligent in their provision of health care, but that defendants have shown "deliberate indifference" to inmates' medical needs. "[D]eliberate indifference to serious medical needs of prisoners constitutes the `unnecessary and wanton infliction of pain' proscribed by the Eighth Amendment." Estelle, 429 U.S. at 104, 97 S.Ct. 285 ( quoting Gregg v. Georgia, 428 U.S. 153, 182-83, 96 S.Ct. 2909, 49 L.Ed.2d 859 (1976)). "This is true whether the indifference is manifested by prison doctors in their response to prisoner's needs or by prison guards in intentionally denying or delaying access to medical are or intentionally interfering with the treatment once prescribed." Id. 104-05, 97 S.Ct. 285 (citations omitted.)

A medical accident cannot alone be characterized as wanton infliction of unnecessary pain. Id. Nor can the inadvertent failure to provide adequate medical care constitute an unnecessary and wanton infliction of pain. Id. at 106, 97 S.Ct. 285. "Medical malpractice does not become a constitutional violation merely because the victim is a prisoner." Id. Moreover, "[p]risoners do not have a constitutional right to any particular type of treatment." Long v. Nix, 86 F.3d 761 (8th Cir. 1996). Rather, prisoners' medical rights have been pared down to treatment that is not "deliberately indifferent" to their "serious" medical needs. Id. at 106, 97 S.Ct. 285.

The court was deeply disconcerted by the inadequate and negligence medical and psychiatric treatment exposed by the plaintiffs and their experts. Plaintiffs demonstrated time and again fact patterns that would likely make winning malpractice suits in civil court. Regrettably, the Supreme Court has stated in no uncertain terms that for an inmate, this is not enough. The inmates must show a systemic pattern of intentional indifference to known medical needs. A court must find that the defendants actually know of, and yet disregard, an excessive risk to inmate health or safety. Farmer v. Brennan, 511 U.S. 825, 114 S.Ct. 1970, 128 L.Ed.2d 811 (1994). Although it is found that the medical and psychiatric care systems in TDCJ are frequently grossly wanting, and that plaintiffs may have in fact shown deliberate indifference in individual cases or institutions, it cannot be stated, under present-day law, that TDCJ officials are systemically and deliberately indifferent to inmates' medical and psychiatric needs.

It can be hoped that "evolving standards of decency" will eventually lead the Supreme Court to a modification of its contemporary standards for cruel and unusual punishment regarding medical treatment for prison inmates. As the law stands today, the standards permit inhumane treatment of inmates. In this court's opinion, inhumane treatment should be found to be unconstitutional treatment.

IX. ADMINISTRATIVE SEGREGATION

A. Compendium

Unlike most of TDCJ-ID's facilities and services, the administrative segregation units of the Texas prison system deprive inmates of the minimal necessities of civilized life. While the court recognizes and appreciates the formidable task of those public servants saddled with the task of dealing with problematic, violent inmates, even those inmates who must be segregated from general population for their own or others' safety retain some constitutional rights. Texas' administrative segregation units violate those rights through extreme deprivations which cause profound and obvious psychological pain and suffering. Texas' administrative segregation units are virtual incubators of psychoses-seeding illness in otherwise healthy inmates and exacerbating illness in those already suffering from mental infirmities.

B. Testimony and Findings of Fact

1. The Psychological Effects of Administrative Segregation

According to TDCJ-ID policy, administrative segregation ("ad-seg") is a "non-punitive status involving separation of an offender from the general population for the purpose of maintaining safety, security and order among general population offenders and correctional personnel and within the prison institution." Def. Ex. 17 at 1. By policy, "[a]t no time is Administrative Segregation to be used as punishment for misconduct." Id. Rather, inmates are to be housed in administrative segregation to protect the safety of others. Sometimes administrative segregation is used to protect inmates from harm in the general population.

TDCJ-ID's administrative segregation cells are separated into three levels. In the highest-security levels, Level II and Level III, inmates are stripped of virtually all personal property, including hygienic materials. Inmates in administrative segregation may be subject to a range of deprivations, including exercise, social activity, and outside contact. In some cases, inmates may be put on "food loaf restriction" and "paper gown restriction."

Through the expert testimony of Allen Breed, Chase Riveland, and Craig Haney, plaintiffs allege that defendants are violating the Eighth Amendment rights of inmates in ad-seg by depriving them of a minimal civilized measure of life's necessities. Plaintiffs emphasized the deprivations that characterize ad-seg. Inmate's lives are virtually void of property, personal contact, and mental stimulus. Few rehabilitative programs are available to them. Breed, Jan. 27 Tr. at 158.

The most persuasive evidence presented by the plaintiffs, however, regarded the behavior of inmates locked away in administrative segregation. These experts revealed a world in which smeared feces, self-mutilation, and incessant babbling and shrieking are almost everyday occurrences. See Pl.Ex. 1391; Pipkin, Feb. 9 Tr. at 256. The most compelling testimony on the appalling world of ad-seg came from Dr. Craig Haney. Feb. 4 Tr. at 101-107. Craig Haney, Ph.D., J.D., is perhaps the nation's leading expert in the area of penal institution psychology. Having both a Ph.D. in Psychology and a J.D. from Stanford, Dr. Haney has spent his career studying the psychological aspects of systems of incarceration. In his work, he has evaluated the psychological effects on inmates of many prisons and prison systems across the country, including between 20 and 30 ad-seg units, and has testified in a number of prison litigation cases. Pl.Ex. 1480. While he began his testimony with a discussion of the psychological effects on inmates in all prisons, Dr. Haney also specifically addressed the conditions in Texas' administrative segregation units.

Based on his inspections at Eastham, Beto, and the high security unit at Estelle, Dr. Haney found "high numbers of prisoners were living in psychological distress and pain" and that these inmates live under a "high risk of psychological harm . . ." Feb. 4 Tr. at 101. While acknowledging that he is not a psychiatrist and cannot diagnose mental illness, Dr. Haney stated that "there appear to be people who are housed there who are manifesting signs and symptoms of some form of psychological disorder." Id. The picture he painted of some inmates' behavior in ad-seg was harrowing:

I'm talking about forms of behavior that are easily recognizable and that are stark in nature when you see them, when you look at them, when you're exposed to them. In a number of instances, there were people who had smeared themselves with feces. In other instances, there were people who had urinated in their cells, and the urination was on the floor. . . . There were many people who were incoherent when I attempted to talk to them, babbling, sometimes shrieking, other people who appeared to be full of fury and anger and rage and were, in some instances, banging their hands on the side of the wall and yelling and screaming, other people who appeared to be simply disheveled, withdrawn and out of contact with the circumstances or surroundings. Some of them would be huddled in the back corner of the cell and appeared incommunicative when I attempted to speak with them. Again, these were not subtle diagnostic issues. These were people who appeared to be in profound states of distress and pain.

According to Dr. Haney, of the perhaps dozens of prisons he has visited and studied in his career, Texas' ad-seg units were "as bad or worse as any I've ever seen." Feb. 4 Tr. at 104. When asked how Texas' ad-seg units were worse than others around the country, Dr. Haney responded:

The bedlam which ensued each time I walked out into one of those units, the number of people who were screaming, who were begging for help, for attention, the number of people who appeared to be disturbed, the existence, again, of people who were smeared with feces, the intensity of the noise as people began to shout and ask, Please come over here. Please talk to me. Please help me. It was shattering. And as I discussed this atmosphere with the people who worked here, I was told that this was an everyday occurrence, that there was nothing at all unusual about what I was seeing.

Dr. Haney also testified to the circular, self-fulfilling purpose of ad-seg. Inmates in ad-seg are suffering such psychological deprivation, he testified, that their behavior becomes worse and they become less able to conform to prison rules. Feb. 4 Tr. at 116. Dr. Haney testified that even psychologically healthy people living in ad-seg under these kinds of deprivations would be negatively effected:

Whether they rise to the level of mental illness or not is in a way beside the point. For many of them, being subjected to this level of deprivation and this level of desperation, it begins to bring out the worst in them. They begin to decompensate not only in the sense of not being able to control themselves but also not to be able to control their actions. They are easily provoked and provocative as they interact with other people, because many of them get extremely on edge, they're upset, they're desperate.

On cross-examination, defendants pointed out that Dr. Haney was neither a psychiatrist, nor a practicing lawyer. Dr. Haney acknowledged that it was possible that some of the inmates he saw exhibiting bizarre behavior could have been undergoing some continuing psychiatric treatment. Feb. 4 Tr. at 140. He also acknowledged the possibility of malingering on the part of inmates, or inmates use bizarre behavior for secondary gain. Id. at 141-142. Dr. Haney explained that in some cases he had reported strange behavior to a psychiatrist and/or correctional officers, but not in all. Id. at 145-147. However, according to Dr. Haney, the extreme deprivations of ad-seg foster bizarre, if not psychotic, behavior. Id. at 104-110.

Breed, whose exemplary qualifications and expertise are discussed in the discussion of inmate safety below, visited the ad-seg units of each of the units he toured. Jan. 27 Tr. at 157. He found Texas's ad-seg system unique among prison systems across the country, in that the large majority of inmates being moved into the ad-seg system begin in Level III, the level with the most severe restrictions. Id. at 163. Breed found "no correctional justification whatsoever" for a system that immediately places a prisoner in a situation with nothing left to lose. Id. at 159-160.

Charles Riveland also explored TDCJ-ID's administrative segregation system and testified to the "stark and severe conditions" imposed on prisoners in Level II and Level III. Jan. 26 Tr. at 65-67. See also Jurczak, Jan. 26 Tr. at 130. Riveland, too, had profound concerns regarding Texas' policy of starting out inmates in the highest level of ad-seg. Jan. 26 Tr. at 66-67. Nor could Riveland identify any correctional purpose in denying individuals such personal items of personal property as books, soap, and deodorant. Id. at 67.

2. Mentally Ill Inmates in Administrative Segregation

Even more disturbing to the court are plaintiffs' allegations that administrative segregation is used to warehouse mentally ill patients who need medical and psychiatric attention. The most compelling evidence of this practice came from plaintiffs' expert witness, Dr. Dennis Jurczak. Dr. Jurczak, a psychiatrist, forensic psychologist and university professor, audited psychiatric services at the Estelle, Byrd, Goree, Huntsville, Ramsey I, Eastham, Beto, Michael and Terrell Units of TDCJ. Dr. Jurczak found a number of the same "extremely repressive" conditions reported by Haney, Breed, and Riveland. See Jan. 26 Tr. at 130. Although Dr. Jurczak found no problems in medical or psychiatric care in the diagnostic units (Byrd and Goree), he did find that "the psychiatric care and treatment provided to offenders in the Texas Department of Criminal Justice institutions is medically inadequate." Pl.Ex. 1493.

Riveland also testified to the confinement of inmates who were mentally ill according to their psychiatric record, or who acted in extremely abnormal ways. Jan. 26 Tr. at 69-70. Such inmates were more often kept in Level II or Level III, and were often subject to punitive food loaf restrictions. Id. at 69-71. Dr. Jurczak was also most concerned by the large number of mentally ill patients in administrative segregation. In his report, Dr. Jurczak concluded that:

Insofar as mentally ill patients in Administrative Segregation are concerned this inadequacy or absence of care combined with the conditions of confinement contributes to an increase in the patient's suffering from the ravages of his disease, interferes with the patient's ability to participate in rehabilitative programming and worsens the long-term chances for successful treatment in the future.

Pl. Ex 1493. In Dr. Jurczak's opinion, "there was something desperately wrong with a system that would have people this ill sitting in segregation and not being recognized by the mental health staff as needing assistance." Jan. 26 Tr. at 153.

Emphasizing that he did not report on borderline mentally ill patients but only on inmates who would be found mentally ill by any reasonable psychiatrist, Dr. Jurczak testified that he had identified 15-20 "floridly psychotic" individuals in administrative segregation. Jan. 26 Tr. at 203; See id. at 132; Pl.Ex. 1493. "The patients demonstrated unequivocal signs and symptoms of a serious mental disorder, predominately that of schizophrenia." Pl.Ex. 1493; see Jan. 26 Tr. at 132. Dr. Jurczak found that many of these individuals were not being followed by the mental health staff and many were not identified as mentally ill. Pl.Ex. 1493. Dr. Jurczak also encountered a number of mentally ill patients in administrative segregation had been identified by TDCJ. Some were receiving care, some not. Pl.Ex. 1493.

In response, defendants' witness Dr. Ducate had the 20 inmates examined by the former chief psychiatrist for TDCJ. Ducate, Feb. 5 Tr. at 164. Dr. Ducate testified that the reviewing psychiatrist "did not find any one individual who was either grossly psychotic or severely under treated. He did identify one patient that he thought would be better served at the Skyview inpatient facility, and that patient was subsequently transferred there." Id. On cross-examination, however, plaintiffs presented evidence that a number of those inmates were, in fact, found by TDCJ-ID officials to be mentally ill. Id. at 242

Dr. Jurczak led the court through descriptions of his encounters with a number of the inmates in ad-seg that he identified as mentally ill. See Pl.Ex. 1491, 1494, 1495, 1497, 1498, 1501, 1502, 1555. A number of the prisoners had free world records of mental illness. The behaviors he reported included paranoid thoughts, looseness of association, and "pressured speech." See Jan. 26 Tr. at 130-150. One inmate scrubbed to remove the bugs from his skin. Jan. 26 Tr. at 144. Others incessantly talked to themselves. Others were "frequent fliers"-those that frequently attempted suicide and made repeated trips to inpatient suicide watch. See id. at 176. Another was a young man referred to him by other inmates. The inmate had been smearing his feces for several years. When Dr. Jurczak tried to interview him, the inmate was incoherently giggling, mumbling, and looking around. According to his records, this inmate was not identified as mentally ill and was not being treated. Pl.Ex. 1502; Jan. 26 Tr. at 155-56. Dr. Jurczak recognized the possibility that such behaviors were for secondary gain, but stated, "I think with 30 years of practicing psychiatry in many, many prisons, if these guys pull the wool over my eyes, they're pretty darn good." Jan. 26 Tr. at 199.

Dr. Jurczak also explained that, in his expert opinion, Texas's administrative segregation system harmed mentally ill inmates. Jan. 26 Tr. at 154-156. Such inmates, he said, need contact and social stimuli. Id. More generally, Dr. Jurczak testified that the ad-seg system is destructive to all its occupants: "I think it's a very destructive system. And I've been in many, many systems . . . and I've never seen one as repressive as I have seen [in TDCJ]." Id. at 161.

On cross-examination, defendants questioned Dr. Jurczak about his involvement in surveying Texas medical systems for NCCHC accreditation. Jan. 26 Tr. at 185. Dr. Jurczak reiterated his concern that psychiatric services was suffering from a "dumbing down" of care givers. Id. at 195-96. Dr. Ducate, however, expressly rejected this contention. Feb. 5 Tr. at 211-213.

Defendants also argued that "mental health services is committed to have every patient who is psychotic or chronically psychotic and cannot function in an outpatient setting placed in an inpatient setting." Def. Post-Hearing Br. at 21. See Cockrell, Feb. 9 Tr. at 282. According to policy, mental health patients who are in ad-seg are seen on rounds by a mental health clinician at least once per week and at least five times per week if they are in disciplinary segregation. Plaintiffs question the efficacy of those meetings, however, alleging that they often involve little more than a single question: See Jurczak, Jan. 26 Tr. at 168-171. Defense witness Nurse Patsy Pipkin, a health care provider at the Estelle high security facility, described her interview procedure for the court, explaining that she tries to ask questions of each inmate that wants to talk to her. Feb. 9 Tr. at 248-249. She insinuated that each care provider may do things differently, however. Id. She also stated that sometimes it is hard to hear the inmates over the din that is common on the wing. Id. at 255. Nurse Pipkin also verified some of the bizarre behaviors described by plaintiffs' experts. Id.

3. TDCJ Policies

Dr. Ducate, Director of Mental Health Services for UTMB, testified to a number of policies designed to ensure that inmates in ad-seg receive psychiatric care if needed. Feb. 5 Tr. at 99-103. See Def. Ex. 47. She represented to the court that some degrees of mental illness can be and are treated while the inmate is in administrative segregation. Ducate, Feb. 5 Tr. at 158. Among the programs available for such inmates is the Program for Aggressive Mentally Ill offenders (PAMIO). See Def. Ex. 50. Furthermore, if an inmate needs greater care, she testified that "no person will be turned away from crisis management if referred, no one." Ducate, Feb. 5 Tr. at 159. However, some inmates refuse treatment, she explained, likening such inmates to mentally ill homeless people who cannot be forced into treatment due to respect for their civil rights. Id. at 170. Dr. Ducate also acknowledged that ad-seg is not conducive to mental health. Id. at 161.

Defendants also introduced into evidence a "Report of Mental Illness and Mental Health Treatment Among Offenders Housed in Administrative Segregation." Def. Ex. 200. That document reviews policies aimed at treating mentally ill inmates in ad-seg with the stated purpose of demonstrating "our commitment to this mission." Id. at 1. Based on these policies and this review, Dr. Ducate concluded that "Mental Health Services is definitely taking care of offenders in administrative segregation."

Defendants' evidence shows that inmates being assigned to ad-seg are given pre-segregation evaluation. Such a screening process is designed to ensure that a patient's condition will not be worsened by placement in ad-seg. Furthermore, nurses in the wings are trained in the recognition of mental health issues and can, and do at times, make referrals. See Pipkin, Anderson, Feb. 9 Tr. Several of plaintiffs' experts, however, expressed doubts as to the success of that screening process. See Jurczak, Jan. 26 Tr.

C. Legal Analysis and Conclusions

It is found by a preponderance of the evidence that inmates in administrative segregation, particularly those in Levels II and III, are deprived of even the most basic psychological needs. More than mere deprivation, however, these inmates suffer actual psychological harm from their almost total deprivation of human contact, mental stimulus, personal property and human dignity. The scene revealed by the plaintiffs' experts, one largely unrefuted by defendants' emphasis on policies and procedures, is one of a frenzied and frantic state of human despair and desperation. Furthermore, plaintiffs submitted credible evidence of a pattern in TDCJ of housing mentally ill inmates in administrative segregation-inmates who, to be treated, would have to be removed to inpatient care. These inmates, obviously in need of medical help, are instead inappropriately managed merely as miscreants. It is determined that TDCJ officials are well aware of both these conditions and these inmates' ensuing pain and suffering. Whether because of a lack of resources, a misconception of the reality of psychological pain, the inherent callousness of the bureaucracy, or officials' blind faith in their own policies, TDCJ has knowingly turned its back on this most needy segment of its population.

Conditions of confinement, alone or combination, may deprive prisoners of the minimal civilized measures of life's necessities. Rhodes v. Chapman, 452 U.S. 337, 347, 101 S.Ct. 2392, 69 L.Ed.2d 59 (1981). It is clear that under American jurisprudence, deprivation of activity, exercise, mental stimulation and activity, at some level, is the denial of a basic human necessity. See Wilson v. Seiter, 501 U.S. 294, 304, 111 S.Ct. 2321, 115 L.Ed.2d 271 (1991). "Some conditions of confinement may establish an Eighth Amendment violation `in combination' when each would not do so alone, but only when they have a mutually enforcing effect that produces the deprivation of a single, identifiable human need. . . ." Wilson v. Seiter, 501 U.S. 294, 304, 111 S.Ct. 2321, 115 L.Ed.2d 271 (1991).

The test for determining "cruel and unusual" conditions of confinement is a dynamic one-dependent not on hard, fast rules, but "civilized standards, humanity and decency." Estelle, 429 U.S. at 102, 97 S.Ct. 285. A court must turn then to the "evolving standards of decency that mark the progress of a maturing society." Rhodes, 452 U.S. at 346, 101 S.Ct. 2392. This court consider, then, an evolving and maturing society that recognizes the harsh reality of psychological pain.

In the past, courts faced with horrendous conditions of confinement have focused on the basic components of physical sustenance-food, shelter, and medical care. See Farmer v. Brennan, 511 U.S. 825, 114 S.Ct. 1970, 128 L.Ed.2d 811 (1994). More recently, in light of the maturation of our society's understanding of the very real psychological needs of human beings, courts have recognized the inhumanity of institutionally-imposed psychological pain and suffering. As the Third Circuit stated, "[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).

It goes without question that an incarceration that inflicts daily, permanently damaging, physical injury and pain is unconstitutional. Such a practice would be designated as torture. Given the relatively recent understanding of the primal necessity of psychological well-being, the same standards that protect against physical torture prohibit mental torture as well-including the mental torture of excessive deprivation. In the eloquent words of Judge Thelton Henderson:

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 pain and suffering caused by a cat-o'-nine-tails lashing an inmate's back are cruel and unusual punishment by today's standards of humanity and decency, the pain and suffering caused by extreme levels of psychological deprivation are equally, if not more, cruel and unusual. The wounds and resulting scars, while less tangible, are no less painful and permanent when they are inflicted on the human psyche.

Before the court are levels of psychological deprivation that violate the United States Constitution's prohibition against cruel and unusual punishment. It has been shown that defendants are deliberately indifferent to a systemic pattern of extreme social isolation and reduced environmental stimulation. These deprivations are the cause of cruel and unusual pain and suffering by inmates in administrative segregation, particularly in Levels II and III.

This court does not today wholly condemn Texas's system of administrative segregation. It is recognized that 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. The need for such segregation is certainly within defendants' discretion. And, if the evidence had not demonstrated that conditions are so extreme as to violate basic concepts of humanity and deprive inmates of a minimal level of life's basic necessities, the court would not interfere. Here, however, the constitution has been violated as to a portion of the plaintiff class. As we are reminded by the Fifth Circuit, "[w]hen the remedial powers of a federal court are invoked to protect the constitutional rights of inmates, the court may not take a `hands-off' approach." Ruiz, 679 F.2d at 1126.

Separately from, and independent of, the determination that the conditions of deprivation in administrative segregation violate the constitution, it is found that administrative segregation is being utilized unconstitutionally to house mentally ill inmates-inmates whose illness can only be exacerbated by the depravity of their confinement. Conditions in TDCJ-ID's administrative segregation units clearly violate constitutional standards when imposed on the subgroup of the plaintiffs' class made up of mentally-ill prisoners. In light of the obvious severity of these inmates' needs, it is determined that defendants have been deliberately indifferent to the serious risks posed by subjecting such inmates to confinement in administrative segregation for extended periods of time.

It is deplorable and outrageous that this state's prisons appear to have become a repository for a great number of its mentally ill citizens. Persons who, with psychiatric care, could fit well into society, are instead locked away, to become wards of the state's penal system. Then, in a tragically ironic twist, they may be confined in conditions that nurture, rather than abate, their psychoses. The United States Constitution cannot abide such a perverse and unconscionable system of punishment. As to mentally ill inmates in TDCJ-ID, the severe and psychologically harmful deprivations of its administrative segregation units are, by our evolving and maturing society's standards of humanity and decency, found to be cruel and unusual punishment.

X. INMATES' SAFETY

A. Compendium

Despite the professionalism of TDCJ corrections staff in many areas, institutional resistance to resolving serious safety problems pervades the system. It is this resistance that forms the basis for the violations here described. According to one of the defendants' experts, running a prison is about complete control. Sandahl, Feb. 5 Tr. at 29. The lack of protection for inmates by TDCJ, therefore, evidences a lack of control by prison officials. It is the obligation of prison officials to ensure that inmates are not subjected to any punishment beyond that which is necessary for the orderly control of the prison. Gates v. Collier, 501 F.2d 1291, 1309 (5th Cir. 1974). Evidence has shown that prison officials have failed in this charge.

The evidence before this court revealed a prison underworld in which rapes, beatings, and servitude are the currency of power. Inmates who refuse to join race-based gangs may be physically or sexually assaulted. To preserve their physical safety, some vulnerable inmates simply subject to being bought and sold among groups of prison predators, providing their oppressors with commissary goods, domestic services, or sexual favors. The lucky are those who are allowed to pay money for their protection. Other abused inmates find that violating prison rules, so that they may be locked away in single cells in administrative segregation, is a rational means of self-protection, despite the loss of good time that comes with their "punishment." To expect such a world to rehabilitate wrong-doers is absurd. To allow such a world to exist is unconstitutional.

B. Testimony and Findings of Fact

1. Introduction

In their presentations at trial and in their pleadings, plaintiffs and defendants advanced completely contrary characterizations of the Texas prison system. Indeed, plaintiffs' counsel noted in her closing arguments that, comparing the closing arguments of the defendants and the plaintiffs, "it is as if we didn't sit through the same trial and hear the same evidence." Brorby, Feb. 12 Tr. at 63. While plaintiffs' experts, principally Allen Breed and Chase Riveland, asserted that TDCJ policies and procedures meet acceptable correctional standards, their conclusions were that, in the implementation of such policies and procedures, TDCJ notably failed. Many inmates credibly testified to the existence of violence, rape, and extortion in the prison system and about their own suffering that resulted from such abysmal conditions. Defendants' experts, including Gary DeLand, David Sandahl, and numerous employees of TDCJ, namely, TDCJ-ID Director Gary Johnson and Deputy Director of Security Janie Cockrell, countered with the argument that the evidence presented by the plaintiffs was anecdotal at best and did not reflect any system-wide problems.

2. Physical Assaults on Inmates

Since 1992, a marked increase in inmate-on-inmate assaults in the Texas prison system has been shown. In 1992, there were 366 reported assaults, for a total of 6.4 incidents per 1,000 inmates. At that time, the prison population was 57,069. In 1993, the prison population was 69,186. That year, there were 555 reported assaults, for a total of 8.0 incidents per 1,000 inmates. By 1998, there were a total of 1,510 reported inmate-on-inmate assaults, for a total of 10.5 incidents per 1,000 inmates. The prison population was then 144,036. Def. Ex. 104 at 17167-68.

Defendants stated, "[t]hat Plaintiffs were able to identify a number of inmates who have been physically or sexually assaulted is no great surprise." Def. Reply in Supp. of Mot. for J. at 13. Testimony from inmates, often unrefuted by defendants and largely credible, described conditions of confinement that lead to fear for their personal safety. The stories of inmate violence and sexual assault were vivid and gruesome. Examples are numerous: one inmate described how he was attacked in his cell at the Neal Unit by three inmates, with a resulting injury of a ruptured spleen and nine days of hospitalization. Jan. 22 Tr. at 116-121. Inmate A was attacked by gang members numerous times in an open dormitory and activity areas at Garza East Transfer Facility. Jan. 21 Tr. at 90-91. He was also knocked unconscious outside of his cell at the Allred unit. Jan. Id. at 108. Inmate FF was assaulted in the cell of the prisoner he was paying for protection. Jan. 22 Tr. at 239-240. Inmate F was beaten by his cellmate in his cell at the Beto unit. Jan. 21 Tr. at 205-06.

A prison officer, with the rank of lieutenant, testifying on behalf of the defendants, confirmed that one-on-one inmate fights and group-on-one inmate fights are a regular part of prison life. Feb. 9 Tr. at 226-228. He also testified that it was a regular part of prison experience that inmates report to officers with a problem and seek protection. Id. at 226-228. He identified a term, "a catch out," that is used for an inmate who is weak, cannot stand up for himself, and needs protection. Id. at 228.

According to Deputy Director Cockrell, prisoners victimize other prisoners because they want money, because they are angry at them, or because they are "just perverts." Feb. 10 Tr. at 35. TDCJ-ID Director Gary Johnson stated that, "it's not uncommon for an offender [inmate] to complain that other offenders [inmates] are trying to make him pay commissary to them." Johnson, Jan. 30 Tr. at 28.

3. Sexual Assaults on Inmates

No attempt is made to monitor the total number of reported sexual assaults in TDCJ, and the result is that differing accounts are extant concerning the total number of sexual assaults throughout the prison units. Johnson, Jan. 30 Tr. at 158-67. According to a document prepared by TDCJ, in 1997 there were 87 reported sexual assaults. Pl.Ex. 7. However, in a letter written by TDCJ Executive Services, the number is reported as "approximately 123" for the year of 1997. Pl.Ex. 9. According to the first document, in 1998 there were 81 reported sexual assaults. Pl.Ex. 7. Yet another number reveals that in 1998, there were 107 reported sexual assaults that were examined by the Internal Affairs Division. Johnson, Jan. 30 Tr. at 166. Director Johnson also acknowledged the existence of inmates' fears of sexual assault. Asked if sexual assault was a major problem in the Ferguson Unit during his tenure there are warden, Director Johnson replied "[sexual assault] was always a — a problem which we heard a lot of allegations about, yes. The fear of — fear of sexual assault and some allegations of sexual assault, that's correct." Id. at 153.

Many inmates credibly described their genuine fears of future harm and the effects of their suffering from past assaults. Of thirty-two male inmates who testified before the court, at least nine had suffered from rape or sexual assaults, and some of those were victimized more than once. Inmate UU testified that after complaining to officers about threats to his safety, to no avail, he was attacked in the showers by four inmates, three who beat him while one inserted his fingers into inmate UU's rectum. Inmate UU, Jan. 29 Tr. at 105-107. Later, when inmate UU missed a payment to a gang that was extorting him for money, four gang members held him down and anally raped him. Id. at 112-113. Inmate M testified that he was raped by his cellmate despite repeated requests for protection to correctional officers. Inmate M, Jan. 22 Tr. at 101-02. Inmate R testified that innumerable inmates forced him to pay protection in the form of sexual favors, and two cellmates raped him several times at the Robertson unit. Inmate R, Jan. 22 Tr. at 195-99; Pl.Ex. 1725A. Another inmate testified that he was beaten up by a new cellmate who was "checking him" to see if he would "be a girl." Ferguson, Jan. 22 Tr. at 21-22.

Another inmate testified that he was raped at his Beto unit Mentally Retarded Offender Program, later raped and beaten by his cellmate at psychiatric hospital Jester IV, and then raped in his cell at Robertson unit. Lee, Jan. 21 Tr. at 152-59. Inmate C testified that he was raped by his cellmate. Inmate C, Jan. 21 Tr. at 189-91. Inmate F testified that he was sexually assaulted in his cell at the Ferguson unit. Inmate F, Jan. 21 Tr. at 196-224; Inmate J testified that he was raped in his cell at Beto unit by three prisoners he was paying for protection. Inmate J, Jan. 22 Tr. at 177-79. Inmate GG, who was allegedly orally and anally raped by three gang members, asked for protection from prison officials, and was denied. Inmate GG, Jan. 22 Tr. at 265-266; Pl.Ex. 670. On another occasion, inmate GG's cellmate reportedly pulled a homemade knife on him and raped him. Id. at 274-75. A doctor confirmed that he had been raped. Id. at 276; Pl.Ex. 657.

There are a number of plausible reasons why the number of reported sexual assaults are so low in a prison system that has a population of over 140,000. For example, according to Janie Cockrell, male inmates have difficulty reporting sexual assaults. Feb. 10 Tr. at 70. Cockrell also testified that it is apparently hard for a male inmate to accept the fact that he has been sexually assaulted. Id. at 70.

An additional, perhaps alternative, explanation for the low reporting of sexual assaults lies in the fact that the system does not encourage victims to come forward and complain about the sexual crimes perpetrated against them. First, there appears to be a strong presumption on the part of prison officials that, in the absence of outward physical harm to assaulted inmates, such as cuts, abrasions, and bruises, no sexual assault has occurred. Second, according to the testimony of inmates, those who report that they have been raped are subjected to disdain by officers. Third, with only six disciplinary actions for sexual abuse in 1998 for the entire prison population, Pl.Ex. 10., there appear to have been no serious ramifications for those inmates committing sexual assault against other inmates. Indeed, there is apparently little concern that six confirmed sexual assaults for 1998 might not accurately reflect what is actually occurring in Texas prisons, for no investigation has been ordered to determine whether the number of confirmed sexual assaults corresponds to the number of sexual assaults actually occurring there. Johnson, Jan. 30 Tr. at 170.

4. Plaintiffs' Experts

Allen F. Breed, one of plaintiffs' experts in the field of corrections, examined the Texas prison system by following a methodology that he has used in 32 cases, in many of which he was appointed by state and federal courts to evaluate and monitor prison conditions. Jan. 27 Tr. at 85; Feb. 10 Tr. at 110. He visited 18 TDCJ prisons, each involving two to three, and in some cases, four-day visits. At no time were his visits to a unit less than 12 to 14 hours a day. The institutions that he visited represented about 32 percent of the total population of inmates in the TDCJ system. Breed walked through the units he visited and reviewed between 500 and 600 grievances, over 500 discipline reports, and approximately 800 institution files. Jan. 27 Tr. at 91.

Breed examined a universe of between 500 and 600 grievances filed over a month's time. Feb. 10 Tr. at 112. He then asked the grievance staff to deliver to him every tenth grievance. Breed found from examining those grievances that about half concerned unprofessional conduct, and almost all concerned endangerment in which the inmates felt that the staff had not been responsive to their problems. Id. He then examined the life endangerment/emergency housing logs. At each unit, either Breed or plaintiffs' expert Chase A. Riveland talked to the chief classification officer about the process relating to life endangerment and what alternatives were available. Id. at 113. If Breed found no documentary verification of life endangerment, he eliminated those cases and interviewed the inmates in the remaining cases. Feb. 10 Tr. at 115.

Breed, initially, was not allowed to speak with staff members without there being a lawyer for defendants present. Later, he was not allowed to speak with staff members without the presence of a high ranking unit administrator. Jan. 27 Tr. at 92. He found that, as a result of the presence of either a lawyer or a supervising officer, staff members found great difficulty in communicating with him; consequently, he did not continue to interview staff as a part of his methodology. Id. at 92-93.

Breed found that, generally, the policies and procedures governing use of force, protection of prisoners from harm, and conditions in administrative segregation were acceptable. Jan. 27 Tr. at 94-95. But, Breed asserted, unless the policies and procedures are carried out, they have no meaning. Id. at 95. Breed found a prevalence of inmates who feared harm from other inmates, and many of them were repeatedly suffering harm. Feb. 10 Tr. at 160.

Chase A. Riveland, another of plaintiffs' experts in the field of corrections, toured and inspected sixteen prison units between March 1998, and October 1998. These prisons comprise approximately 25 percent of the inmate population in Texas. Riveland based his opinions on his knowledge and experience in the field of corrections, review of TDCJ policies and procedures, and information gathered by him during visits to Texas prisons, including information obtained from interviews with inmates, review of inmate files, and informal interviews with TDCJ employees. In assessing the conditions of confinement in TDCJ prisons, Riveland concentrated on conditions in the areas of protection of inmates, administrative segregation, use of force, and general crowding and staffing issues. Pl.Ex. 1522 at 2-6.

Riveland interviewed inmates chosen randomly while "walking the run" and by plaintiffs' counsel, reviewed medical and unit files of inmates, attended Unit Classification Hearings, and talked to selected TDCJ staff members. Riveland testified that, while existing policies and procedures of TDCJ concerning the issue of protection of inmates were adequate, TDCJ's compliance with such policies and procedures was deficient. Jan. 26 Tr. at 21.

First, Riveland testified that often no formal record is made of inmate-on-inmate violence, thus allowing routine dismissal of inmates' complaints by the line staff. Jan. 26 Tr. at 21-23. Second, Riveland asserted that a number of inmates in obvious need of protection, because of their physical or mental status, were not considered by the system. Id. at 24. Third, Riveland testified that there was a lack of meaningful investigation with respect to the need for safety of individual inmates. Id. at 32. Fourth, Riveland observed requests for safekeeping or protection being denied to inmates where the circumstances seemed clearly to require such protection. Id. at 22. He also testified that there were many instances in which a thorough investigation of an individual inmate's request for protection was warranted prior to a determination of the request, but such an investigation were not conducted. Id. at 39.

5. Defendants' Response

Gary DeLand testified as defendants' expert witness and offered testimony in opposition to plaintiffs' experts, Chase Riveland and Allen Breed. Feb. 2 Tr. at 207-08. DeLand conducted interviews with prison staff, including, among others, Gary Johnson, Janie Cockrell, Lepher Jenkins, and Debbie Lyle. He recently toured the Fergusen, Estelle, and Eastham units, where he interviewed the wardens, as well as interviewing staff involved in administrative segregation. He also reviewed use of force reports and grievance processing. Id. at 203-06. At each of the three facilities DeLand visited, he looked at the last three use of force files. Id. at 204. DeLand spent a total of about two and a half hours at each of the three units. Feb. 6 Tr. at 133.

DeLand criticized both Chase Riveland and Allen Breed for making conclusory statements. Feb. 2 Tr. at 208. That is, DeLand expressed concern that the plaintiffs' experts drew conclusions about the prison system based solely on anecdotes. Feb. 6 Tr. at 82. It was the opinion of DeLand that experts in the field of corrections should not rely solely on anecdotal evidence. Id. at 83. DeLand was further of the opinion that, in order to have a valid result of system-wide violations, an expert would need a random sample from all institutions. Id. at 83-84. He also argued that Breed should not have tainted the sample by including inmates recommended by the plaintiffs. Id. at 84. In addition, DeLand criticized the failure of Breed to interview administrative staff and the policymakers within TDCJ-ID. Id. at 102.

DeLand stated that, over the years, state corrections systems, nationwide, were pushed, perhaps "kicking and screaming," to make necessary constitutional changes in their operations. As a result, DeLand has seen vast improvements in states like California and Montana. Feb. 2 Tr. at 212.

While at the same time criticizing plaintiffs' experts for drawing conclusions based on limited exposure to the prison system, DeLand himself arrived at conclusions about Texas prisons with much less examination of the prison system than Breed or Riveland. DeLand criticized Riveland's testimony that TDCJ evidenced a systemic lack of concern for inmates requiring protection. However, more than merely stating that Riveland's testimony was flawed, DeLand stated unequivocally that Riveland's testimony was very inaccurate. Feb. 2 Tr. at 231. In his report, DeLand asserted that "TDCJ-ID's policies, procedures, and practices generally meet or exceed appropriate correctional standards in the area of prisoner safety and protection." Def. Ex. 205 at 25. DeLand concluded that, from his exposure to the Texas system, he felt "very strongly that [he could] support those opinions [in his report]." Feb. 2 Tr. at 206. DeLand testified that Texas is one state that can be held out as an example to other states. Id. at 212. The conditions of prison life, between now and even a few years ago, according to DeLand, is "remarkably different." Id. at 213.

It is evident that a well-designed statistical evaluation of the safety issues facing Texas prisons and the inmates that reside in them would be a very useful tool, as testified by DeLand. Nevertheless, DeLand implicitly advocated a methodology similar to, if not simpler than, that used by Breed and Riveland for the examination of prison policy compliance. Specifically, in conjunction with another case, DeLand examined the Los Angeles County corrections system, where about 22,000 people were incarcerated. After DeLand had spent two days looking at policies, talking to people, and looking at records, he concluded that they were in compliance with use of force policies. Feb. 6 Tr. at 128.

Breed and Riveland have invested much more than two days in their efforts to evaluate the Texas prison system. Neither relied solely on anecdotes to reach his conclusions. Therefore, in light of the thoroughness of the methodology used by plaintiffs' experts, the impressive credentials of plaintiffs' experts (particularly Breed) in the field of corrections, DeLand's own use of a similar methodology in at least one instance, and additional evidence before the court, plaintiffs' methodology is found to be an acceptable tool with which to determine TDCJ compliance with its policies.

6. The Grievance Process

Between January and September of 1998, inmates filed 235,213 grievances. Twenty-eight percent of those were complaints against staff. Pl.Ex. 1659. Breed asserted that, while a function of the grievance procedure is to resolve the problems of inmates, it is also a tool for administrators to determine whether or not their policies are carried out. Feb. 10 Tr. at 160. Despite the high numbers of grievances, Breed found that many inmates did not want to report life endangerment complaints, because they would be forced to "snitch" on others, or because they were embarrassed about the nature of the treatment they had suffered. Id. at 115.

In his experience, DeLand felt that an administrator could not disregard the complaints from inmates, but, rather, should cause investigations or inquiries to be conducted. He stated that he never took at face value what was said in a prison. Feb. 2 Tr. at 153. Riveland testified that, even when an inmate followed the proper procedure outlined in the TDCJ's policy, the system's response was often inadequate. He gave examples of inmates who filed grievances complaining of specific problems, such as the need for protection, but whose grievances were neither investigated nor thoroughly reviewed. Jan. 26 Tr. at 51. The responses were often nothing more than standard stock responses. Id. As a result of inadequate responses and routine denial of requests, Riveland testified that prisoners regarded the grievance system with a "universal distrust." Id.

7. Safekeeping and Protective Custody

Gary Johnson, testifying as an expert for the defendants, described the different steps that can be taken by TDCJ to protect specific inmates from other, violent inmates. The first step is verbal intervention by correctional staff trained to identify the source and seriousness of the problem. A second method of alleviating tension between inmates is to change housing assignments, work assignments, or work-shift hours. Def. Ex. 119 at 8. A third, more formal, option is to change the custody of the aggressor as the result of a major disciplinary offense. This is done by the Unit Classification Committee. Id. at 10.

A fourth method for protecting inmates is a transfer of either an oppressed inmate or his oppressor from one unit to another. Def. Ex. 119 at 10. A fifth option is safekeeping. Both of these procedures, unit transfers and safekeeping, begin with a report to the unit classification department (from either the inmate or a correctional officer) that a life endangering situation may exist. An investigation is conducted, and usually within 36 hours of the report, the investigation is returned to the classification department for further action. The sixth method is protective custody. In protective custody, inmates are confined alone in cells designated specifically for them. A security escort accompanies them whenever they leave their respective cells. These inmates do not participate in work or educational programs, and they do not shower, recreate, or join other inmates in any other status. The seventh and most extreme measure is an interstate corrections transfer. Id. at 11.

As of December 1, 1998, there were 2,592 safekeeping beds and 128 protective custody beds in the Texas system. Pl.Ex. 15. DeLand testified that safekeeping beds available in TDCJ are sufficient to meet the needs of inmates. He based his opinion that those in safekeeping are receiving appropriate consideration in TDCJ on Def. Ex. 102, as well as by interviewing people involved in formulating policies. Feb. 6 Tr. at 172. Some of his opinions about safekeeping were founded on discussions with Johnson and Cockrell, but they did not supply him with any data to support their compliance with good safekeeping practices. Id. at 174-75. DeLand also asserted that, looking at statistics concerning safekeeping and life endangerments requested in ten units in the month of June 1998, TDCJ-ID staff was appropriately handling inmate requests for protection. Id. at 96-97.

On the other hand, Breed found that although safety problems are numerous, not enough protective custody beds or safekeeping beds exist that are truly safe. Feb. 10 Tr. at 160. Riveland expressed a similar concern when he criticized the amount of access that general population inmates have to those in safekeeping. Jan. 26 Tr. at 53-54. This was because, pursuant to TDCJ policy, inmates in safekeeping shower, eat, recreate, and attend educational programs with inmates in the general population. Id. at 54. Riveland also testified that the number of protective custody and safekeeping beds was too low in light of the size of the prison population in TDCJ. Furthermore, he testified that the "predator-specific" basis criterion failed to help vulnerable inmates who were targeted for assault by other inmates. Id. at 55-56.

According to DeLand, it is fairly commonplace for inmates to attempt to use the safekeeping process to "manipulate" the system, for reasons that are not related to any supportable requests for protection. Feb. 6 Tr. at 114. He gave no statistics concerning this, only his experience. Riveland agreed that some requests for protection had to be denied. However, he testified that the requests of inmates for protection needed to be treated as "supportable requests" for the purpose of investigation; that is, until they were proven to be unsupported. Jan. 26 Tr. at 96-98. He argued that TDCJ incorrectly places a great burden on inmates to prove their need for protection, often beyond showing physical injuries and almost always beyond threats of violence. Id.

There was a common pattern in the testimony of inmates seeking safety from prison officials. That is, inmates, seeking refuge from harm or threat of harm, consistently had to subject themselves to punishment and disciplinary measures by prison officials in order to remove themselves from harm's way. Several inmates, for example, testified that they refused to enter their cell so as to be disciplined (and therefore removed from the reach of their predators). The case of Inmate M is descriptive. Fearful of a cellmate who had threatened him, Inmate M was subjected to chemical agents for refusing to return to his cell, in an attempt to be removed to disciplinary confinement. He testified that he was told by corrections officers that his abusers would leave him alone if he fought them. Jan. 22 Tr. at 89-90. Inmate N similarly was punished for "refusing housing"-a punishment he claims to have needed in order to avoid abuse. Jan. 22 Tr. at 154-155. One inmate testified that he conspired with a sympathetic officer to throw water on the officer. Jan. 21 Tr. at 220. The officer agreed to give the inmate an "assault of an officer case," so that the inmate would be sent to administrative segregation. Id. "We didn't tell," the inmate testified. "It was just our secret and I really respected that man for that." Id. at 221. Attempts by inmates to remove themselves from danger by breaking prison rules were not always successful. Inmate R was physically carried back to his cell where he was again attacked by his cellmate. Jan. 22 Tr. at 197-198. See also Jan. 27 Tr. at 25-26.

Defendants argue that when an inmate has his or her safekeeping or transfer request denied, nothing of a negative nature happens to the inmate. Def.'s Supp. Mot. for J. at 23. However, TDCJ does not appear to have accurate system-wide data concerning requests for protection. Without adequate records, it cannot clear what the safety needs of inmates truly are.

One indication that the defendants do not have accurate data on protection and safekeeping requests stems from the fact that many such requests are not documented. Correctional officers do not record requests for protection, but rather they report them to a sergeant (the first line supervisor), who has "a tremendous number of duties." Roesler, Feb. 9 Tr. at 207. Typically, the sergeants will refer such a request to at least a lieutenant, if not a captain. Only when the request comes to the attention of a captain will it be documented. Id.

8. Accreditation

The defendants introduced the testimony of David G. Sandahl, whose testimony concerned policies, procedures, and practices of the TDCJ. His opinions were based on his participation in the American Corrections Association (ACA) accreditation audits of two TDCJ units, Robertson and Eastham units. Sandahl bases his decision for accreditation on his experience with a facility and with the accreditation files. Feb. 5 Tr. at 32. The accreditation files are those assembled by the unit to show examples of compliance with the standards of the ACA. Feb. 5 Tr. at 48-49. Sandahl concluded from his visits to Robertson and Eastham units, as well as from his experience and a review of TDCJ policies, that inmates at those two units did not express concern over staff control, that staffing at the two units was sufficient, and that both units had been recommended for ACA accreditation. Def. Ex. 216 at 5-6.

While TDCJ's participation in the ACA accreditation process is to be commended, accreditation, in itself, is not a clear indication that TDCJ is properly following its policies and procedures. Experts from both parties recognized the limitations of ACA accreditation. According to Sandahl, American Correctional Association accreditation do not win lawsuits; but, he asserted, such accreditations show that prison officials are engaging in a good faith effort to improve the area in which they are working. Feb. 5, Tr. at 24. Similarly, DeLand recognized that ACA accreditation is a tool, but not a constitutional standard. Feb. 2 Tr. at 162. According to Breed, the American Corrections Association accreditation is useful; however, the ACA does not carry out what would be considered a performance-based review in the accreditation process. Feb. 10 Tr. at 100. The ACA auditor's role is to see that the unit has a policy and that there are some indications that the policy is being carried out. Id. at 104. Breed testified to a number of examples where a prison system was accredited by the ACA, but was, nevertheless, held by a court to be operating in an unconstitutional fashion, including prisons in Florida and the San Quentin prison in California. Id. at 108.

9. Classification Staff

In 1995, unit classification staff numbered 328, with a ratio of inmates to unit classification staff at 233 to one. In 1998, there were 61 on the unit classification staff, with a ratio of inmates to unit classification staff at 1,251 to one. Pl.Ex. 11. Unit classification case managers, under the supervision of a chief of unit classification, were responsible for referring inmates to appropriate treatment programs, interviewing inmates, and reviewing their records, as required. Case managers represented inmates at classification hearings by providing the committee with information concerning the inmate's case. Additionally, case managers provided a unit orientation for newly assigned inmates and other support for unit administration. Def. Ex. 119 at 2.

Breed stated that the most important factor in the existence of the problems of undue and excessive use of force and failure to provide protection from harm is the reduction in the number of classification counselors, who were reduced from one counselor to several hundred inmates to one to over a thousand inmates. Jan. 27 Tr. at 102-03; Pl.Ex. 11. Classification counselors, he emphasized, form the bridge between corrections officers and the inmate. Jan. 27 Tr. at 104.

DeLand responded that it was not reasonable for Breed and Riveland to conclude that the reduction in the classification officers resulted in increased levels of violence. Feb. 6 Tr. at 111. DeLand learned of the classification counselors and had their roles explanation to him by Gary Johnson. Feb. 2 Tr. at 215. DeLand criticized the plaintiffs' experts for not undertaking a systematic analysis to determine a causal connection between the reduction in classification staff and the increased level of violence. Feb. 6 Tr. at 113. DeLand did acknowledge that the TDCJ found the classification counselors to be quite important to the system and did not want to lose them. Id. at 114.

It would seem plausible that Breed is correct in his assessment of a nexus between the increase of violence and excessive use of force and the reduction of classification counselors. However, causal connection in this matter has not been proved by a preponderance of the evidence. It is noted, nevertheless, that evidence presented by both parties indicates that the dire circumstances now faced by inmates would most likely be significantly improved by the increased presence of classification counselors.

10. Deliberate Indifference

A TDCJ captain who testified for the defendants stated that good officers are firm but fair and "know how to use their discretion . . . [and] send the information and communication up the chain of command as well as receive information from above." Feb. 9 Tr. at 212. Plaintiffs argue that this is precisely the discretion that is often abused by such officers. Riveland testified that he found evidence in TDCJ's own documents indicating its officers had knowledge of some of the harms occurring to inmates, but still took no action to provide protection for those individuals. Jan. 26 Tr. at 39.

There was uniform testimony from inmates that prison officials are aware of the violence and victimization perpetrated by inmates against other inmates. Furthermore, inmates testified that if they had no injuries or had only minor injuries, their requests for protection from harm are repeatedly denied for "insufficient evidence." Jan. 22 Tr. at 171; Pl.Ex. 270-277; Pl.Ex. 94.

A TDCJ captain who testified for the defendants about protection matters stated that he disbelieved an inmate's threat of harm from sexual assault after the inmate told him that the inmate had been forced to perform sexual acts with two different inmates. Feb. 9 Tr. at 214. The reason the captain refused to acknowledge the threat of harm to the inmate was because the inmate had not physically fought against his attackers after they threatened him with physical harm. Id. at 215-17. Counsel for the plaintiffs questioned him about that belief:

Q: Now, if an inmate attempts to defend himself [from a sexual assault] and he is fighting an inmate who is maybe stronger and a better fighter than he is, the consequence of that is that he'll get a beating; is that right?

A: Probably so.

Q: And then what's going to happen after the beating?

A: Then that's something that helps me make my decision whether or not his allegations are actually legitimate or not.

Q: Do you think that maybe the inmate just has a choice between having a beating and being raped versus just being-you know, he could skip the beating and go right to the rape?

A: That's an individual choice that he can make.

Q: But you disapprove of that choice; is that right?

A: Yes, ma'am, I would.

Q: Because that would be consenting to sexual contact.

A: It could be misconstrued by the inmates that are attempting to rape him. It could be misconstrued by them that he is consenting.

As mentioned above, TDCJ does not have a central system for tracking reports of sexual assault. However, even for those reported instances of sexual assault, very few inmates are disciplined for sexual assault. Pl.Ex. 10. In 1998, there were only six disciplinary actions for sexual abuse. Id. Therefore, according to Janie Cockrell, in 1998 there were only six confirmed sexual assaults in TDCJ. Feb. 10 Tr. at 32.

C. Legal Analysis and Conclusions

Prison officials must provide humane conditions of confinement, and must "take reasonable measures to guarantee the safety of the inmates. . . ." Hudson v. Palmer, 468 U.S. 517, 526-27, 104 S.Ct. 3194, 82 L.Ed.2d 393 (1984); see DeShaney v. Winnebago County Dept. of Social Services, 489 U.S. 189, 200, 109 S.Ct. 998, 103 L.Ed.2d 249 (1989). It is "cruel and unusual punishment to hold convicted criminals in unsafe conditions." Youngberg v. Romeo, 457 U.S. 307, 315-16, 102 S.Ct. 2452, 73 L.Ed.2d 28 (1982). "[P]rison officials have a duty . . . to protect prisoners from violence at the hands of other prisoners." Farmer v. Brennan, 511 U.S. 825, 833, 114 S.Ct. 1970, 128 L.Ed.2d 811 (1994) ( quoting Cortes-Quinones v. Jimenez-Nettleship, 842 F.2d 556, 558 (1st Cir. 1988)); Horton v. Cockrell, 70 F.3d 397, 400 (5th Cir. 1995). "Gratuitously allowing the beating or rape of one prisoner by another serves no `legitimate penological objectiv[e].'" Farmer, 511 U.S. at 833, 114 S.Ct. 1970 ( quoting Hudson, 468 U.S. at 548, 104 S.Ct. 3194 (Stevens, J., concurring in part and dissenting in part)).

The objective component of a failure to provide reasonable protection from harm claim is "incarcerat[ion] under conditions posing a substantial risk of serious harm." Farmer, 511 U.S. at 834, 114 S.Ct. 1970. The Fifth Circuit has affirmed that "[s]ome conditions of confinement may establish an Eighth Amendment violation `in combination' when each would not do so alone, but only when they have a mutually enforcing effect that produces the deprivation of a single, identifiable human need. . . ." Moore v. Mabus, 976 F.2d 268, 271 (5th Cir. 1992).

"[A] remedy for unsafe conditions need not await a tragic event." Helling v. McKinney, 509 U.S. 25, 33, 113 S.Ct. 2475, 125 L.Ed.2d 22 (1993). Inmates must be protected against both present and future dangers. Horton v. Cockrell, 70 F.3d 397, 401 (5th Cir. 1995) ( citing Helling v. McKinney, 509 U.S. 25, 113 S.Ct. 2475, 125 L.Ed.2d 22 (1993)). "Prison authorities must protect not only against current threats, but also must guard against `sufficiently imminent dangers' that are likely to cause harm in the `next week or month or year.'" Horton v. Cockrell, 70 F.3d 397, 401 (5th Cir. 1995) ( quoting Helling v. McKinney, 509 U.S. at 32-33, 113 S.Ct. 2475).

The subjective component of a failure to provide reasonable protection from harm claim is "deliberate indifference," which is "more than mere negligence" in failing to prevent harm, and "less than acts or omissions for the very purpose of causing harm or with knowledge that harm will result. Thus, it is the equivalent of acting recklessly." Farmer, 511 U.S. at 835, 114 S.Ct. 1970. The obviousness of the risk from facts known to the defendant is circumstantial, but rebuttable, proof that defendant inferred the risk. Id. at 840-44, 114 S.Ct. 1970; Bradley v. Puckett, 157 F.3d 1022, 1025 (5th Cir. 1998)

In Alberti v. Klevenhagen, the Fifth Circuit addressed circumstances similar to the plaintiffs' claim of failure to provide reasonable protection from harm. In that case, the court held that "[v]iolence and sexual assault among inmates may rise to a level rendering conditions cruel and unusual." 790 F.2d 1220, 1224 (5th Cir. 1986). In Alberti, the district court concluded that "`[i]nmate beatings and homosexual rapes and attacks are prevalent.'" Alberti 790 F.2d at 1224 ( quoting Alberti v. Heard, 600 F. Supp. 443, 457 (S.D.Tex. 1984)). The Fifth Circuit held that it "need not determine whether any of those incidents individually constituted an Eighth Amendment violation, for the evidence established that the totality of the circumstances in the jails were condemnable." Alberti, 790 F.2d at 1225.

The district court in Alberti did not accept reported incidents of violence as the sole measure of conditions in the jail. Alberti, 790 F.2d at 1225. Often violence occurred when guards were beyond reach. Id. at 1225. As the Fifth Circuit noted, "[f]ear of reprisal, desire not to be known as a `snitch' and, in the case of homosexual attacks, reluctance to identify oneself as an `easy mark' deter victims from reporting such attacks." Id. at 1226 ( quoting Doe v. District of Columbia, 701 F.2d 948, 966 (D.C. Cir. 1983) (separate statement of Edwards, J.)).

Similar to the instant action, the Alberti court found that "[q]uantitative measures are significant, for Eighth Amendment determinations must be informed by objective factors to the extent possible." Alberti, 790 F.2d at 1226. The Fifth Circuit further found that statistics were unreliable, because they did not take account of the possibility of vast number of unreported acts; therefore, they were subject to a wide margin of error. Id. 1226. It was also found that "even an accurate statistic would be of little use in making comparisons to other institutions and communities unless reporting methods and other variables were controlled. . . ." Id. It is the logic and judgment of the trial court that determines the sufficiency of the evidence presented. See Ruiz, 679 F.2d 1115, 1133.

The Alberti court also concluded, on the basis of evidence that the jail, with 60,000 inmates passing through in a year, had 1,200 acts of violence, as well as considering the testimony of inmates and experts, that many more incidents went unreported, and together this constituted proof of a violation of the Eighth Amendment, for failure to protect the inmates. Alberti, 790 F.2d at 1224-26. It was also held that "the pattern of violence . . ., coupled with such inadequate supervision, creates a constant threat to the inmates' safety." Id. at 1226. Finally, the court held that, `"[i]t is not necessary that every inmate be assaulted every day before a federal court may intervene." Id.

It is concluded from the credible evidence adduced at trial, in the form of testimony by experts, inmates, and other evidence offered by the parties, that beatings, rape, and other attacks of inmates are prevalent throughout TDCJ, much more prevalent than reported in data offered to the court. It is difficult to believe that, in a population of more than 140,000, there were only six confirmed cases of sexual assault in 1998. That only six inmates were eligible for punishment for sexual assault further defies credulity. It is also disturbing to learn that there are apparently no reliable system-wide statistics concerning requests for protection by inmates. The inability of TDCJ to compile and report data on such assaults evidences disregard by prison officials for a crime that rips at the dignity and humanity of its victims. And, when prisoners do seek protection and are granted safekeeping-not protective custody beds, of which TDCJ has very few-they are faced with a custody that is not truly protective, and places them frequently in the same danger they have sought to avoid. Whatever protective measures in TDCJ-ID policy manuals may be, they do not result in the protection of vulnerable inmates.

While there may be well-designed and intentioned policies and procedures to respond to the safety issues faced by inmates, these policies are not well implemented. Evidence has shown that, in fact, prison officials deliberately resist providing reasonable safety to inmates. The result is that individual prisoners who seek protection from their attackers are either not believed, disregarded, or told that there is a lack of evidence to support action by the prison system. Inmates who find no other way to protect themselves than to rebel against the prison officials themselves find themselves in a free fall of disciplinary measures that strips them of their privileges and rights. Prison officials at all levels play a game of willing disbelief, one that appears adequate on paper and fails dismally in practice.

While many of the incidents presented before the court certainly rise to the level of an Eighth Amendment constitutional violation, it is not necessary to determine whether any of the incidents individually constitutes such a violation, for the evidence has established that the totality of the circumstances in the TDCJ are condemnable. Alberti v. Klevenhagen, 790 F.2d 1220, 1225 (5th Cir. 1986). The plaintiffs have shown by a preponderance of the evidence both the objective and subjective elements of an Eighth Amendment claim for failure to reasonably protect them from harm. The conditions of confinement of the inmates, as described above, in combination, have a mutually enforcing effect that deprives inmates of a single, identifiable human need: safety from their fellow inmates. More specifically, the combination of inmates who are routinely subjected to violence, extortion, and rape, of officers who are aware of inmate-on-inmate victimization but fail to respond to the victims, of high barriers preventing inmates from seeking safekeeping or protective custody, and of a system that fails accurately to report, among other data, instances of requests for safekeeping and sexual assaults, and, as well, the obviousness of the risk to prison officials, when all are considered together, have the mutually enforcing effect of rendering prison conditions cruel and unusual by denying inmates safety from their fellow inmates.

XI. Excessive Force

A. Compendium

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). The job of a corrections officer is, by any measure, a thankless and demanding one. At times, the use of force is not only a justified, but also a necessary, tool in the quest to maintain an institution's order, or a guard or inmate's safety. To serve these purposes, the Texas prison system is delegated wide-ranging power over its inmates. It is not hyperbole to describe prison inmates as at the mercy of prison guards.

With such power, wielded by corrections officials as young as eighteen-years old, comes a tremendous potential for abuse. It is such abuse in TDCJ, evidenced by the testimony of expert witnesses, of numerous inmates, and extensive documentation, that is condemned by the Constitution, and therefore, by this court. Plaintiffs presented this court with a collage of inmates' injuries-broken jaws, bruised faces, and broken bones, among others. These injuries were shown to be the results of wholly unnecessary physical aggression by prison employees. Unfortunately, the pattern of uncalled for "slamming," hitting, and kicking by corrections officers in the cellblocks of TDCJ-ID is so prevalent as to implicate the Constitution. Simply stated, the culture of sadistic and malicious violence that continues to pervade the Texas prison system violates contemporary standards of decency.

B. Testimony and Findings of Fact

1. Introduction

Plaintiffs argue that class members throughout TDCJ live under a substantial risk of subjection to unnecessary, malicious, and sadistic force by TDCJ corrections officers. They assert that TDCJ officials know that these problems exist, yet fail to take reasonable steps to mitigate them. Plaintiffs condemn correction officers' quick resort to the use of force without ever attempting to abate a confrontation through communication or tactical force-avoidance techniques. The inmate class alleges a tendency of investigators to accept at face value, without investigation, staff member accounts of use of force incidents, and a failure of unit wardens and supervisors to manage and supervise use of force on their own units. Plaintiffs further claim that TDCJ systematically charges offending correctional officers with technical violations rather than unnecessary or excessive force. Pl. Post-Tr. Br. at 6. Plaintiffs rely on the testimony of numerous inmates and prison expert Allen Breed, among others, to support these allegations.

Defendants respond that TDCJ, as reflected in easily ascertainable statistical data, has a well-functioning use of force plan that both limits use of force and adequately monitors its incidence. Defendants assert that plaintiffs' claims of injuries are unsubstantiated and not supported by the evidence. Defendants also attack the opinion of Allen Breed, plaintiffs' use of force expert, as being "of no constitutional consequence," because he did not know the proper constitutional standard and his "so-called analysis contained no review of culpability. . . ." Def. Reply in Supp. of Mot. for J. at 20-22. In addition to exhibits and the testimony of TDCJ officials, including Gary Johnson and Janie Cockrell, defendants relied primarily on the expert testimony of Gary DeLand.

2. An Overview of Expert Methodology

Since both parties criticized the others' experts' methods of evaluating the uses of force in TDCJ-ID, it would be prudent to clarify the bases for the experts' conclusions. As in his examination of TDCJ policies and practices concerning inmate protection, prison expert Allen Breed performed an in-depth investigation into the use of force in Texas prisons.

In his examination of the TDCJ system, Breed used a methodology similar to the one he has used in many of the approximately 32 cases in which he has been appointed by the courts as an expert or Special Master to evaluate and monitor prison conditions. For this phase of this litigation, he visited 18 prisons, representing about 32% of the total inmate population of TDCJ, and each unit visit involved two to four days lasting between 12 and 14 hours each. Jan. 21 Tr. at 91.

Before he visited the prison units, Breed reviewed all of the use of force reports for the previous month. Feb. 10 Tr. at 111. Following his review of the records at the unit, Breed then interviewed all of the inmates whose files contained a use of force report, and who were still at the unit. Id. at 111. This amounted to 319 cases in which he reviewed individual medical files and interviewed each inmate. Id. at 118. Following those interviews, he then reviewed medical records in the central file, as well as any other documents that might give him an indication of the extent of force used. Id. at 112. Breed also made contact with inmates by reviewing grievances, looking at administrative segregation rosters, and walking through the units. He interviewed a number of inmates and/or reviewed the records of inmates at the request of plaintiffs' attorneys. Jan. 27 Tr. at 191-93.

Defendants' argue that Breed's opinions should not be considered because he "could not even come close to reciting the proper constitutional standard." Reply in Supp. of Mot. for J. at 20. While Breed's opinions as an expert concerning desirable prison conditions may help establish contemporary standards of decency, see Rhodes v. Chapman, 452 U.S. 337, 348 n. 13, 101 S.Ct. 2392, 69 L.Ed.2d 59 (1981), the court does not rely on his opinion to establish the constitutional minima. Id. ( quoting Bell v. Wolfish, 441 U.S. at 543-44 n. 27, 99 S.Ct. 1861, 60 L.Ed.2d 447 (1979)). As mentioned above, whether prison conditions are sufficiently harmful to establish an Eighth Amendment violation is a purely legal determination for the court to make. See Hickey v. Reeder, 12 F.3d 754, 756 (8th Cir. 1993) ( citing Hudson, 503 U.S. at 6, 112 S.Ct. 995); Madrid v. Gomez, 889 F. Supp. 1146, 1246 (N.D.Cal. 1995). In fact, experts are entitled to little weight in determining whether a condition is "cruel and unusual punishment" under the Eighth Amendment. See Toussaint v. McCarthy, 801 F.2d 1080, 1107 n. 28 (9th Cir. 1986); Madrid, 889 F. Supp. at 1159.

Chase Riveland also made conclusions for the plaintiffs about the prevalence of excessive force in the Texas prison system. As bases for those conclusions, Riveland toured and inspected 16 prisons, accounting for approximately 25 percent of the inmate population of TDCJ. Riveland reviewed the medical files and interviewed inmates in 155 use of force cases. Breed, Feb. 10 Tr. at 122.

Defendants responded with testimony from prison expert Gary DeLand. To arrive at his conclusions, DeLand toured three units and conducted interviews with prison staff. Feb. 2 Tr. at 203-06. DeLand spent about two and a half hours total at each of the three units. Feb. 6 Tr. at 133. DeLand also evaluated eight use of force reports. Feb. 6 Tr. at 139.

One document inspected by DeLand that contained information on the kinds of force and the injuries that resulted from use of force incidents was Def. Ex. 103, "Offender Use of Force Injuries from June 1998 to November 1998." Feb. 6 Tr. at 151. DeLand apparently was led to believe that the document covered the entire TDCJ system. This was the document that he used to arrive at his conclusions concerning the incidence of the use of force in TDCJ and whether or not it is excessive. Id. at 152. However, DeLand acknowledged, in cross examination, that, in fact, the document reflected only 25 units and was incomplete in its reporting of even those units. Id. at 152-56.

Defendants also offered the testimony of expert witness Dr. Daniel H. Freeman, Ph.D., who reviewed reports by Breed and evaluated Breed's methodology. Dr. Freeman also reviewed the testimony of defendants' witness Edward Ellis McElyea, chief of the administrative support bureau in TDCJ Internal Affairs Division (IAD), as well as McElyea's audit of IAD files, later introduced as Def. Ex. 215A, 221, and 227.

Dr. Freeman criticized Breed for his methodology, concluding that Breed's audits of use of force reports and other evidence of use of force did not meet minimum standards for scientific research and are flawed from a statistical point of view. Def. Ex. 226. While the court has addressed this criticism in its discussion concerning proving systemic constitutional violations, it should also be noted that the court has assessed Breed, not as a statistician, but rather as a preeminent expert in the field of the corrections. His opinion is regarded by the court as one that informs contemporary standards of decency, and his appointment by approximately 30 state and federal courts to monitor and evaluate prison conditions is a testament to his ability to assess the existence of problems in a prison system.

On the other hand, Dr. Freeman's assurances that McElyea's audits of use of force reports, from which the court could draw conclusions about use of force patterns and IAD investigations, meet the standard for scientific research from a statistical point of view, Def. Ex. 227, may, nevertheless, demonstrate both the potential limitations of Dr. Freeman's assessment and the apparent inability of TDCJ to compile proper data concerning problems with the system such as use of force.

3. TDCJ-ID Policies and Procedures

Experts for both parties were generally in agreement that TDCJ's written policies adequately address use of force concerns. Plaintiffs' expert Breed found that the policies and procedures concerning use of force in TDCJ are generally acceptable. Jan. 27 Tr. at 94-95. DeLand also testified that TDCJ's use of force plan is well written and well-developed. Feb. 6 Tr. at 123. In his expert report, DeLand concluded that "TDCJ-ID's policies, procedures, and practices generally meet or exceed . . . appropriate correctional standards in the area of use of force against prisoners." Def. Ex. 205 at 22.

The apparent point of contention in this battle of experts is the use of the term "practices." In his discussion of TDCJ-ID policies, Breed, who has in his career assisted in the development of innumerable prison policies and procedures, emphasized, as have a number of defendants' experts in this litigation (including DeLand), that unless policies and procedures are carried out, they have no meaning. Jan. 27 Tr. at 95.

4. A Culture of Force

The practice of pervasive physical contact by correctional officers with inmates represents a clear violation of basic penological principles, Breed asserted. "One of the first things you do in corrections is teach people not to touch unless it's absolutely necessary," he testified, "and not even then unless you're sure that you have sufficient manpower, people power, to be able to handle the situation." Id. at 97-98. Defendants' expert David G. Sandahl concurred, testifying that a correctional officer is trained that force is to be used only "as an absolute last resort to resolve any situation." Feb. 5 Tr. at 53.

Based on his evaluation of the use of force by TDCJ-ID corrections officers, however, Breed opined that there is a "culture of force" in the Texas prison system. Jan. 27 Tr. at 96. He stated that "one of the biggest problems in the Texas system that [he] saw is that the officers cannot keep their hands off prisoners." Id. at 97. Breed testified that Texas prison guards are habitually pushing, shoving, and using other physical contact with inmates. Id. Breed also found that the provocations for officers' uses of force were often verbal insults. Id. at 99. In Breed's expert opinion, physical force is never to be used, including "slamming," in response to such a verbal assault, although he saw evidence of this practice at TDCJ. Id.

Riveland also testified that prison officials "relie[d] dominantly on force or threat of force for the control of people." Jan. 26 Tr. at 72. Riveland testified that he found many examples of use of force simply to punish or hurt a inmate, rather than for any legitimate corrections purposes. Id. at 73-74.

Although DeLand agreed with Breed's assertion that avoiding physical contact with inmates should be a basic tent of corrections officer training, DeLand criticized Breed for stating that there was a deeply entrenched culture of force as TDCJ's primary mechanism of inmate control. Feb. 2 Tr. at 211. Asserting that plaintiffs' experts drew impermissible conclusions about the prison system based solely on anecdotes, Feb. 6 Tr. at 82, DeLand reiterated TDCJ's well-regarded policies on the use of force by prison employees.

5. The Prevalence of Excessive Force

It is notable that in almost every prison and jail civil action to which Breed was appointed by state and federal courts, he was responsible for monitoring excessive used of force. This makes all the more alarming the fact that Breed found in Texas more excessive force, in quantity and degree, than in any other state system he has seen. Jan. 27 Tr. at 117. Breed testified that, in forming his opinions about use of force in TDCJ, he found a large proportion of excessive or unnecessary force among the hundreds of use of force instances he reviewed. Id. at 101; see also Def. Ex. 215, 251A, 221. While Breed acknowledged the occasional need for force to control an aggressive inmate, Breed found that correctional officers are extremely quick to put inmates to the ground, an act known as "slamming." Jan. 27 Tr. at 98. "[I]nstead of stepping back, instead of calling for help, instead of getting a supervisor, [officers] slam [inmates] to the ground," he testified.

In 27 of the 82 Major Use of Force (MUOF) reports, offered by the defendants as evidence in Def. Ex. 221, inmates were subjected to force while in restraints; 13 of those 27 inmates were taken to the ground, or "slammed." Sandahl testified that an inmate whose arms are handcuffed behind his back can be aggressive, but that it does not happen very often. Feb. 5 Tr. at 54. One inmate credibly testified that, after he demanded to see a supervisor (because he wanted soap with which to shower) and while in shower slippers and with his arms handcuffed behind his back, an officer grabbed his arm, jerked him off his feet, and slammed him head-first to the ground. The inmate lost consciousness briefly and awoke with a broken jaw and lacerations to his face. Jan. 25 Tr. at 13-16, 23-24.

Another inmate testified, without contradiction, that after he sought help from prison officials because of threats of physical and sexual assault from his cellmate, he was denied protection. Fearful for his safety, the inmate "refused housing" (refused to return to his cell) and was placed in a cage. The inmate testified that he still refused to return to his cell, so he was grabbed by officers, thrown to the floor, and punched in the ribs and back. The inmate sustained abrasions to his neck, abdomen, and hip, and he received a disciplinary case for refusing housing. Inmate M, Jan. 22 Tr. at 93-98.

Sandahl testified that incidents where inmates are punched in the face by officers should raise concern. Feb. 5 Tr. at 41. Breed testified that he knew of "no correctional administrator that would advocate striking anybody in the face with their closed fist." Jan. 28 Tr. at 37. DeLand also testified that it is inappropriate for an officer to punch an inmate with a closed fist, and that such a practice does not serve any legitimate corrections purpose and is solely retaliatory. Feb. 2 Tr. at 150. In fact, DeLand fired a corrections officer for attempting to punch a restrained inmate after the inmate had spit in the face of the officer. Id.

Nevertheless, in one IAD investigation, it was determined that a corrections officer hit an inmate on the head with his closed fist after the inmate turned toward the corrections officer and spit in the face of the officer. The inmate, who was handcuffed behind his back, fell to the floor face down. The officer was not disciplined for this incident, and it was determined that the file was "not sustained;" ( i.e., by a preponderance of the evidence, it was found that the employee had not violated a TDCJ rule or policy). Def. Ex. 215/215A, # 623970. In another case, a corrections officer was disciplined with only one day suspension and nine months probation for kicking an inmate in the head while the inmate was lying on the ground in restraints. Def. Ex. 191.

Other evidence before the court demonstrates a similar pattern of such abuses. See Def. Ex. 103 and Def. Ex. 104. For example, during September 1998, 14 of the 36 reported major use of force incidents in the Connally unit resulted in injuries, and five of those 14 injury cases resulted when inmates were "struck by officer with fist." Def. Ex. 103 at 3, Def. Ex. 104.

Injury from use of force is also prevalent. For example, in the MUOF reports provided to the IAD for the month of February 1998, approximately 35 percent involved injuries. Cockrell, Feb. 10 Tr. at 19. Gary Johnson concurred that injuries contained in the use of force reports may not show the actual severity of injuries suffered by inmates. Johnson, Jan. 30 Tr. at 14. The court heard testimony from one inmate who suffered a broken ankle, as a result of use of force, Batiste, Jan. 27 Tr. at 2, but the MUOF report indicated little or no injury. Def. Ex. 215/215A, # 531735. See also Def. Ex. 194 (inmate suffered contusions to the face and abrasions to the neck, but this was not reflected in medical notes for the MUOF Injury Report).

6. The Ambiguity of the Numbers

Defendants' primary response to Breed and Riveland's reports of grand-scale unjustified violence by correctional officers is reference to self-reported data on uses of force that compare favorably with other states. Defendants, through DeLand, showed that, per capita, the number of reported use of force incidents have decreased. Feb. 6 Tr. at 88-90. The incidence of reported use of force have gone from 6,378 incidents in 1992, to 10,107 in 1995, to 8,051 in 1998. When compared against the total population, there were 122.2 reported incidents per 1,000 inmates in 1992, 93.9 reported incidents per 1,000 in 1995, and 56.3 reported incidents per 1,000 in 1998. Def. Ex. 104 at 17170. These figures, according to DeLand, contradicted the plaintiffs' experts' claims that use of force against inmates by staff is on the rise. Feb. 6 Tr. at 89.

However, on cross-examination, DeLand acknowledged that, within units of TDCJ with administrative segregation populations greater than 100 inmates, the reported incidents of use of force actually increased from 93.9 per 1000 in 1995, to 102.2 incidents per 1000 inmates in 1996, and then decreased to 90.4 per 1000 in 1997, only to rise again to 93 per 1000 in 1998. Feb. 6 Tr. at 141-43; Pl.Ex. 1615. DeLand also testified that this was the type of data that should be compiled to properly manage a system the size of TDCJ. Feb. 6 Tr. at 145.

The plaintiffs also presented the court with numbers, purportedly supporting their use of force allegations. Experts for both plaintiffs and defendants acknowledged California as one of the systems most comparable to Texas. Breed reported California had 3,991 use of force incidents last year in 1998, with a prison population of 158,000, which is 253 incidents per 10,000. Feb. 10 Tr. at 131. In Texas, the rate is 560 incidents per 10,000, although Breed believes that there is significant under-reporting. Id.

The persuasive value of these statistics, however, was in question throughout the hearing. The variations in definitions of "use of force" in different states make inter-state comparisons potentially misleading. Defendants contend that Texas has an extremely broad, all-encompassing definition of use of force. According to DeLand, TDCJ officials have not only broadened the definition of what is a serious use of force or assault on an officer, but they also increased reporting. This, according to TDCJ-ID officials, explained the rise in assaults. Feb. 6 Tr. at 85. DeLand did not explain when the definition was refined, nor did he explain how and when TDCJ expanded its efforts to increase reporting.

At the same time, plaintiffs assert that the use of force data must be viewed in light of the fact that Texas has the highest per capita rate of incarceration of any state in the United States, Pl.Ex. at 10, and relatively more minimum custody inmates than other systems. Johnson, Jan. 29. Tr. at 146; Def. Ex. at 120. There was evidence presented that the relatively high incarceration rate of individuals who commit minor crimes in Texas skews the state's numbers relative to other states, where such lesser criminals would be in county facilities or on some form of supervised probation. Therefore, plaintiffs assert that Texas's higher incarceration rate and greater proportion of minimum custody offenders than other systems indicates that its use of force rate is actually much higher than it appears. Pl. Sec. Part. Br. at 34.

7. Non-physical Force

Not all uses of force reported to the court were functions of active physical aggression. These practices, however, further illustrate the strong emphasis on physical control in TDCJ. For example, the use of "tear gas" was the subject of several witness's testimony. TDCJ has no standards regarding the amount of pepper spray (or OC, for olio capsicum, gas) that should be used in a single administration, or the time that should elapse before a second dosage can be used. McElyea, Feb. 11 Tr. at 187. There can be medical contraindications to the use of OC gas. Johnson, Jan. 20 Tr. at 233. Breed asserted that four administrations of OC gas in two second bursts, each five minutes apart, was more than was normally used in other prison systems in the United States. Feb. 10 Tr. at 193.

Breed found an excessive use of the administration of gas in TDCJ. For example, Breed testified to an instance in which there was no penological justification for a forced cell extraction and found the use of OC, three bursts (six ounces) of OC gas in quick succession, to be excessive. ( Jan. 28 Tr. at 97. See also Def. Ex. 221, # M-45-38-07-98, M-4771-07-98). Furthermore, MUOF reports in Def. Ex. 221 demonstrate that TDCJ regularly uses OC gas to force inmates to remove an arm or pillow from a food slot (through which food is delivered to inmates).

Riveland testified that an example of the prevalence of the culture of force in TDCJ involved a decision at the Estelle unit to release gas into an entire "pod" of administrative segregation, which contained 23 inmates, in response to a disturbance in which some of the inmates were banging on cell doors and plugging up toilets. Jan. 26 Tr. at 77-79. Apparently those inmates not involved were not moved out of the pod prior to the gassing. Id. at 78. One inmate testified before the court that he was denied protection, even though he had genuine fears for his safety, so he refused housing (refused to return to this cell). In response, the inmate was sprayed with pepper gas. Inmate M, Jan. 22 Tr. at 91-92.

In another manifestation of the culture of excessive physical control, Riveland testified to one particularly disturbing incident concerning an inmate who was placed in a "cage" for almost five days as retribution for swallowing a handcuff key. Jan. 26 Tr. at 75-77 (Felder). For most of the five day period, the inmate was not released to urinate or defecate and received very little food or water. Id. at 75. During this time, the inmate was visible to several ranking officers, none of whom took any action. Id. at 76. The only employee disciplinary action was a demotion, by one rank, of the major who initially ordered the confinement in the cage. Id.; see also Feb. 6 Tr. at 178-80.

Plaintiffs allege that defendants' indifference to matters of excessive force is further indicated by the use of detention trailers or cages as punishment. Inmate Lewis Hall testified that in June 1998, he sat in such a cage on a trailer, without a bench, water, or toilet facilities, for three hours, all the while with his hands cuffed behind his back. Jan. 22 Tr. at 39. Jennifer Grey related to the court a similar experience. She allegedly spent four hours in a hot cage with, at times, so many inmates that she could not sit down. Id. at 52. Defendants witness Warden Sharp defended the use of these cages. They are necessary, he testified, because corrections officers cannot immediately deal, by themselves, with intransigent inmates in the fields. Feb. 9 Tr. at 276-278. Defendants' expert DeLand, however, acknowledged that the only penological justification for handcuffs was to control the inmate while he or she is being taken into or out of the cage. Feb. 6 Tr.

8. Monitoring, Supervision, Grievances, and Investigations

Plaintiffs allege that these punitive and unnecessary uses of force are afforded implicit approval by a knowingly indifferent system of monitoring and supervision in TDCJ-ID. While any major use of force in the prison system must, by policy, be reported to the administration, plaintiffs' witness contend that these report have little or no real effect on the pervasive use of excessive force in TDCJ-ID.

Edward Ellis McElyea is employed by the Texas Department of Criminal Justice Internal Affairs Division (IAD) as chief of the administrative support bureau. In this capacity, he is the custodian of all records, and he supervises all administrative and technical personnel, such as the computer technicians who evaluate the division's data and statistics. Feb. 4 Tr. at 195. The Internal Affairs Division has a computerized system that tracks and records information on every case file that is generated by the Internal Affairs Division (IAD). Id. at 196. All grievances that contain an allegation of excessive use of force are then forwarded to the Internal Affairs Division. Id. at 201. On allegations of excessive force, IAD opens up an information file and subsequently makes a determination concerning whether to conduct an investigation, and how to dispose of the file. The major use of force bureau of the IAD reviews every use of force reported in TDCJ. Id. at 203-04. All investigations that find violations of policy are forwarded to Cockrell for disciplinary action against the employee.

McElyea testified that in 1996, 2,835 IAD cases were opened for excessive use of force allegations. Feb. 4 Tr. at 206. Of that number, 2,366 cases were determined not to be a major use of force. Id. at 209. Of the 2,835 cases, 174 were held to be violations of the rules or policies of TDCJ. Id. at 206.

Plaintiffs' witnesses criticized the fact that these files do not include inmates' medical reports. Feb. 6 Tr. at 159. And, a number of the cases apparently have notations that state that the inmate was struck by the officer, but do not state how. Id. at 162. Defendants, maintaining that Texas has a broad definition of use of force, defend the reliability and appropriateness of these records. McElyea also testified that, if an officer pushed an inmate into his cell hard enough to make him fall down and hit something, a violation of the use of force plan was likely, whether the inmate had injuries or not. Feb. 4 Tr. at 223.

Plaintiffs point to what they believe to be a knowing and institutionalized resistance to investigating effectively uses of force by corrections officers against inmates. Based on his thirty-day and sixty-day analyses of various institutions' use of force records, Breed testified to having found more inmate grievances about uses of force than there were reports of such force by correctional officers. Jan. 27 Tr. at 160-165. According to the plaintiffs, this indicates an under-reporting of uses of force. Id. Equally troubling to Breed was the fact that a very small portion of the prison-created uses of force documents included victim statements. Id. at 119.

Breed's assertions may be supported by the fact that in the first seven months of 1998, the Internal Affairs Division opened 1,185 use of force cases that had not been brought to the administration's attention by corrections officers. That is, such reports were only opened after inmates, or someone else (i.e., through letters from persons in the free world) brought the situation to the administration's attention.

While the reason for these alleged victims' silence is unknown, Breed hypothesized that when forms were being filled out, inmates were not actually being offered an opportunity to detail the incident. Id. at 232-233. Breed testified that he substantiated this suspicion by talking to several inmates who were never asked to give statements for their use of force reports. Jan. 27 Tr. at 116. One inmate, who suffered head injuries after allegedly being beaten by guards while he was on the ground, had a use of force report on file that indicated that twenty-four inmates refused to give statement regarding the incident. Breed, after investigation, reported that when he talked to six of those twenty-four inmates, they maintained that they were never asked to give a statement. Id.

Breed also criticized the method of producing the written use of force reports. The officer who used the force initially writes a report. Then, according to Breed's evaluation of the documents, other "witness" officers provide statements that are practically identical to the involved officer's statement. Jan. 27 Tr. at 118. Defense witness DeLand testified that officers may help each other minimize the extent to which they used force in a use of force incident. Feb. 6 Tr. at 165. According to DeLand, a policy cannot be written that prevents under-reporting. Id. at 165-66.

Breed also found that there was nothing in the files to indicate that any investigation was made by the supervisor or middle management level to confirm the facts of the case. Jan. 27 Tr. at 120. In his experience, first-line supervisors must be required to investigate thoroughly use of force incidents to curtail excessive and unnecessary force. Id.

Defense witness DeLand categorically rejected plaintiffs' assertions that claims of excessive use of force were rarely, if ever, investigated. DeLand based this opinion on an interview with Ed McElyea, Chief of the Internal Affairs Division of TDCJ-ID. Feb. 2 Tr. at 223. DeLand also looked at statistics provided to him by McElyea. Id. at 225. DeLand concluded from this information that TDCJ-ID takes the investigative process seriously. Id. at 225.

However, McElyea testified that the IAD investigators have complete discretion as to what investigation will be undertaken and when it will be closed. Feb. 10 Tr. at 173. He also indicated that cases that involve no more than an officer's word against an inmate's word are generally not investigated. He resisted plaintiffs' attempts to consider hypothetical situations and determine when an investigation is necessary. Id. at 169-173.

DeLand countered plaintiffs' contentions that TDCJ is deliberately indifferent to sadistic and malicious uses of force by guards. According to DeLand, if units are not collecting information on a systematic basis, "it may not be a systemic disregard; it simply may be that they haven't thought of collecting that information. . . ." Feb. 6 Tr. at 149.

In 1997, TDCJ imposed disciplinary action a total of 228 times as a result of an Internal Affairs Division Investigation. As a result of a major use of force, 276 employees were disciplined, but not as a result of an investigation by the Internal Affairs Division. Def. Ex. 104 at 17190. However, there is no indication in the record how the employees were disciplined.

C. Legal Analysis and Conclusions

Prison officials may not use excessive physical force against inmates. Hudson v. McMillian, 503 U.S. 1, 4-7, 112 S.Ct. 995, 117 L.Ed.2d 156(1992). Correctional officers must "balance the need to `maintain or restore discipline' through force against the risk of injury to inmates." Id. at 6, 112 S.Ct. 995. The core inquiry in a claim for excessive use of force is "whether force was applied in a good-faith effort to maintain or restore discipline, or maliciously and sadistically to cause harm." Id. at 7, 112 S.Ct. 995. Such maliciousness can be shown by inference. 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); Hill v. Shelander, 992 F.2d 714, 717-18 (7th Cir. 1993) (holding that evidence that officer assaulted inmate without provocation and that the inmate did not resist would allow the finder of fact to "infer that [the officer] acted with malicious intent").

There is no serious injury requirement for a use of force violation of the Eighth Amendment. Hudson, 503 U.S. at 9, 112 S.Ct. 995; Tijerina v. Plentl, 984 F.2d 148, 151 (5th Cir. 1993). This is because "[w]hen prison officials maliciously and sadistically use force to cause harm, contemporary standards of decency are always violated." Hudson, 503 U.S. at 9, 112 S.Ct. 995. Eighth Amendment violations for excessive use of force do not include " de minimis uses of physical force, provided that the use of force is not of a sort `repugnant to the conscience of mankind.'" Id. at 9-10, 112 S.Ct. 995 ( citation omitted). Nevertheless, the Fifth Circuit has held that "to support an Eighth Amendment excessive force claim a prisoner must have suffered from the excessive force of a more than de minimis physical injury, but there is no categorical requirement that the physical injury be significant, serious, or more than minor." Gomez v. Chandler, 163 F.3d 921, 924 (5th Cir. 1999) (holding that, as a matter of law, the court could not say that cuts, scrapes, contusions to the face, head, and body of an inmate were not more than de minimis). Blows that cause "bruises, loosened teeth, and a cracked dental plate are not de minimis for Eighth Amendment purposes." Hudson, 503 U.S. at 10, 112 S.Ct. 995.

It is found that in numerous cases brought before the court, there was no justification for use of force by TDCJ officers. In some situations, use of force was appropriate, but the force used was entirely disproportionate to the circumstances in which it was imposed. No attempt has been made to address in this memorandum opinion every incident that was raised at trial in plaintiffs' experts' declarations, by inmates, or in the reams of documentary evidence. Nevertheless, the totality of this evidence has been taken into account. It is, accordingly, found that plaintiffs have presented sufficient evidence to show, by a preponderance of the evidence, that, in TDCJ penal institutions, the prevalence of excessive use of force constitutes cruel and unusual punishment.

The enormous size of the Texas prison system does not mandate that plaintiffs prove a particular number of excessive force incidents to demonstrate a system-wide pattern of excessive use of force. The logic and judgment of the trial court determines the sufficiency of the evidence. Indeed, the preponderance of the evidence shows that excessive force incidents are under-reported, so it would appear to be impossible for the defendants or plaintiffs to provide an exact number of excessive use of force cases in TDCJ.

Inferences have been drawn that a malicious purpose existed by examining the extent of the injury suffered, the need for the application of force, the relationship between that need and the force used, the threat reasonably perceived by the responsible officials, and any efforts made to temper the severity of a forceful response. See Hudson 962 F.2d at 523 ( citing Hudson, 503 U.S. at 7, 112 S.Ct. 995). The record reflects many instances, in a variety of circumstances, in which excessive force was used, not for the good faith purpose of restoring or maintaining order, but maliciously-for the purpose of causing harm, that is, causing significant pain or injury. Severe force was used where there was little or no need. Plaintiffs have also demonstrated that the pattern of excessive force caused injuries that were more than de minimis, and serious enough to violate contemporary standards of decency and, therefore, the Constitution.

Evidence shows that prison officials abdicate their responsibility in the area of supervision of use of force. Prison officials permit officers to submit overly general incident reports, which do not always accurately reflect the gravity of the injury to inmates. Officials turn a blind eye to serious injuries that are unexplained or suspiciously explained. It is notable that defendants never acknowledged that there was a genuine problem to address. The extent to which excessive force is used in TDCJ, combined with the inability or failure of the prison system to control use of force incidents, reflects what can only be described as an affirmative management strategy to permit the use of excessive force for both punishment and deterrence. It is clear that, while IAD goes through the motions of filing paperwork on cases, it seldom finds officer misconduct. The result is to send a clear message to line staff that excessive force will be tolerated.

XII. CONCLUSION

It has been over three decades since the matter of Texas prisons' constitutionality first came before this court. In light of the egregiousness of the violations of the Constitution found in 1980, the Texas Department of Criminal Justice, through the sometimes strained partnership with the representatives of the inmate plaintiffs in this civil action, has dramatically overhauled its prison system. The imposition of extensive policies and the formation of a bureaucracy do not, however, immunize the system from constitutional challenge. The measure of a prison system's constitutionality, as always, is not its production of policies, but its treatment of inmates.

Texas prison inmates continue to live in fear-a fear that is incomprehensible to most of the state's free world citizens. More vulnerable inmates are raped, beaten, owned, and sold by more powerful ones. Despite their pleas to prison officials, they are often refused protection. Instead, they pay for protection, in money, services, or sex. Correctional officers continue to rely on the physical control of excessive force to enforce order. Those inmates locked away in administrative segregation, especially those with mental illnesses, are subjected to extreme deprivations and daily psychological harm. Such practices and conditions cannot stand in our society, under our Constitution.