Full opinion text
MEMORANDUM AND ORDER KOPF, District Judge. This matter is before me upon remand from the United States Court of Appeals for the Eighth Circuit. Jensen v. Clarke, 73 F.3d 808 (8th Cir.1996) (.Jensen IV). I have been directed to certify findings to the court of appeals. Id. at 811. Because I presume both an understanding of the long history of this matter and an appreciation of the need for its expeditious resolution, I shall respond to the remand order as simply and rapidly as possible. I. DISCUSSION Imagine you committed a crime and are entering the Nebraska State Penitentiary for the first time as a convicted felon. You are locked in a cell no bigger than a closet for 10 to 14 hours each and every day. In the cell you find a monster in the form of a man. The lights go down. It’s hard to see in. If you scream, you probably will not be heard. Imagine further that this creature has a well-documented history of taking his recreation by sodomizing any available prey. If the prey resists, the monster may use a razor to slice the victim from the “shoulder down to the ass.” Imagine also that your keepers, who are by no means evil, know they can save you from brutalization by glancing at a few bits of paper. But your keepers have consciously decided that efficiently packing the available cells is more important than looking at the papers that might reasonably provide for your safety. Space is valuable, and you, as a prisoner, are not. Imagine this scene (or some no less troubling variation) being repeated over and over again every year. As you digest the dry discourse that follows, it is useful to keep the monster in mind. Relying upon Farmer v. Brennan, — U.S. -, 114 S.Ct. 1970, 128 L.Ed.2d 811 (1994), the court of appeals has directed me to specifically “find whether the defendants actually knew of and disregarded a substantial risk to the safety of the plaintiffs.” Jensen IV, 78 F.3d at 811. The court was specifically concerned with “the subjective-state-of-mind component of deliberate indifference described in Farmer.” Id. The Jensen TV court expressly reserved judgment on whether the plaintiffs were exposed to a substantial risk of physical harm. Id. at 810. Thus, the Jensen IV remand order was limited to the question of whether the defendants were “subjectively aware of a substantial risk of harm to the plaintiffs.” Id. at 811. In order to put the remand order in proper context, one must understand this court’s prior ruling in favor of the plaintiffs. Among other things, the court previously ruled that: (1) a pervasive risk of harm from violence existed at the Nebraska penitentiary that carried over into the double cells of the housing units, and (2) the random placement of newly arrived inmates into double cells (without consideration of information regarding compatibility gathered and compiled during the initial intake of inmates) was not a reasonable response to the pervasive risk of harm. Jensen I, 807 F.Supp. at 1488. Reading the general language of Jensen IV together with the specific holding of Jensen I, I understand the specific question on remand to be this: Given that the “level of violence in the four main housing units exceeds what can be characterized as an unavoidable level ... [and that such] violence has carried over into the double cells,” Jensen I, 807 F.Supp. at 1483, whether the “defendants actually knew of and disregarded [the] substantial risk [from violence] to the safety of the plaintiffs,” Jensen IV, 73 F.3d at 811, by “failing to use the classification information available to them in placing newly arriving inmates in double cells.” Jensen I, 807 F.Supp. at 1483. In order to answer this question, it is helpful to first answer the following seven questions: 1. Are inmates, including newly arrived inmates, frequently double-celled? 2. Are newly arrived inmates randomly double-celled; that is, are newly arrived inmates assigned cells without consideration of classification information that could easily be used to predict inmate compatibility? 3. Did Clarke and Hopkins actually know that newly arrived inmates were frequently randomly double-celled? 4. Did the level of violence at the penitentiary, including violence in the double cells, pose a substantial risk of serious harm to the plaintiffs? 5. Did Clarke and Hopkins actually know the level of violence at the penitentiary, including violence in the double cells, posed a substantial risk of serious harm to the plaintiffs? 6. Was the failure to use the classification information and the concomitant failure to determine inmate compatibility an unreasonable response to the substantial risk of serious harm to the plaintiffs?. 7. Are Clarke and Hopkins personally responsible for the failure to use classification information and the concomitant failure to determine inmate compatibility during the placement of newly arrived inmates in double cells? I find as fact and conclude as law that the answer to each of these questions is “yes.” Accordingly, both as a matter of fact and as a matter of law, I find and conclude that the plaintiffs have proved that Clarke and Hopkins “actually knew of and disregarded [the] substantial risk to the safety of the plaintiffs,” Jensen IV, 73 F.3d at 811, by “failing to use the classification information available to them in placing newly arriving inmates in double cells.” Jensen I, 807 F.Supp. at 1483. The reasons for my findings and conclusions are set forth hereafter. A. MOST INMATES ARE DOUBLE-CELLED. Placing two inmates in a cell designed for one person was the norm and had been since at least 1981. Jensen I, 807 F.Supp. at 1467-1469 (especially Ex. 321). This court previously found that in the housing units, which normally contain 80 single cells, 60 of those single cells housed two inmates. Id. This means that 75 percent of the cells designed for one man held two instead. Indeed, by 1991 all the housing units at the penitentiary were, on average, 64 percent over design capacity. Id. at 1468 n. 4. Á long waiting list for single cells required inmates to wait years before they were placed in single cells because such cells were assigned on the basis of “seniority.” Id. at 1469 (Vol. I, at 48:1-4; Vol. II, at 176:14-21; Vol. Ill, at 458:24-459:1). Clarke summarized the situation well when he described the housing units as “bursting at their seams.” Id. at 1468 (Vol. VIII, at 1406:1-3). Teresa Mulkey, a housing unit manager who also chaired the “initial classification” meeting, testified that virtually all newly arrived inmates are double-celled. Q. And are all the people coming [from] the Diagnostic and Evaluation Center placed into a two-man cell? A. I would say most of them are. Q. Do any of them ever come in from D & E and go right to a single-man cell? A. The only possibility for that might be a handicapped inmate ... [or] someone going immediately to death row- Otherwise, no, they would go to a two-man cell. (Vol. VII, at 1202:10-21.) B. NEW INMATES ARE OFTEN RANDOMLY ASSIGNED TO A DOUBLE CELL; THAT IS, THEY ARE ASSIGNED TO A CELL WITHOUT CONSIDERATION OF CLASSIFICATION INFORMATION THAT COULD EASILY BE USED TO PREDICT INMATE COMPATIBILITY. I shall first reexamine the findings of Judges Piester and Cambridge regarding the question of random cell assignments. I shall then examine the response of Clarke and Hopkins. 1. Judge Piester and Judge Cambridge found that the procedure for assigning newly arrived inmates to double cells was “random.” Id. at 1477. While there was some divergence of testimony on this issue, I agree with Judges Piester and Cambridge that newly arrived inmates, who are unknown to the unit managers, are normally “subject to a random placement unless the incoming inmate can identify a particular inmate with whom he should not be celled.” Id. (Vol. XII, at 2335, 2337, 2342). It is important to emphasize what is meant by “random” placement. The word “random” was accepted by Donn Beaver, a prison employee who managed one of the housing units, as a fair description of the cell-assignment procedure for newly arrived inmates at the penitentiary. (Vol. VI, at 989:5-19 (“Q. Is that more or less a random choice as to where he goes? A. I would say yes.”).) The decision regarding which cell the new arrival will be placed in is normally made before any consideration can be given to important background information gathered about the new inmate prior to his entering the prison. Id. at 1477 (Vol. IX, at 1751-53). A bit of detail is required to understand the significance of this point. Adult males sentenced to prison in Nebraska undergo an examination called a “classification study.” Id. at 1476 (Vol. IX, at 1703:9-1704:16). This study, compiled by a wide range of specialists at an institution separate from the penitentiary (the Diagnostic and Evaluation Center (D & E)), attempts to summarize and evaluate: (1) the legal aspects of the crime of which the inmate stands convicted; (2) the criminal history of the inmate; (3) the medical history of the inmate; (4) the psychological status of the inmate; and (5) the needs and problems of the inmate. Id. Associated with the “classification study” is a “scoring instrument” that amounts to a shorthand compilation of the factors examined by the “classification study.” Id. at 1476-77 (Ex. 328, at 55-56; Vol. IX, at 1710). Both the “classification study” and the “scoring instrument” are sent to the institution where the inmate is to be housed, but the “classification study” does not always arrive with the inmate upon his entry into the prison. Id. at 1477 (Vol. IX, at 1718). However, the “scoring instrument” normally arrives with the inmate. Id. (Vol. IX, at 1719). Both the “classification study’ and the “scoring instrument” could easily be used to predict inmate compatibility for cell-assignment purposes if those documents are consulted before a housing decision is made. Id. at 1477. No one disputes this point. Indeed, Hopkins testified that “I think all the information contained there will give the people involved ... an opportunity to make appropriate cell assignments.” (Prelim.Inj.Hr’g Tr. at 47:8-19.) As Judges Piester and Cambridge found, the difficulty is that the “classification system is used only to determine the institution to which an inmate will be sent. Vol. IX, at 1751-53. Beyond that, neither the study nor the instrument is used to predict compatibility of inmates.” Id. Thus, unless the placement decision makers — the unit managers who run the housing units and assign the cells — know the newly arrived inmate, as from a previous incarceration, the cell assignments are made blindly as far as inmate compatibility is concerned. Id. at 1477. (“When no prison official is familiar with the incoming inmate, that inmate will be subject to random placement unless the incoming inmate can identify a particular inmate with whom he should not be celled.”) Id. (Vol. XII, at 2335, 2337, 2342). In other words, when the newly arrived inmate is unknown to the unit managers, the only factor in cell assignment becomes “space availability” and not whether one inmate will harm the other while both occupy the “available space.” Id. (Vol. VII, at 1199:25-1200:13). 2. Clarke and Hopkins attempt to relitigate these previous factual findings in their primary remand brief. (Defs.’ Br. at 21-29.) They advance two arguments, neither of which are persuasive. a. Initially, the defendants argue that the unit managers make a considered decision about inmate compatibility before making cell assignments. This assertion is untrue. Jeffrey Uttecht, who coordinates the cell-assignment process, explained at trial what actually happens. Uttecht is normally told on Tuesday how many new inmates are coming to the penitentiary by Fred Britten, who has administrative responsibility for all the housing units. (Vol. XII, at 2297:24r-2298:9.) On average, the penitentiary receives 8 to 10 inmates each week as part of this process. (Prelim.Inj.Hr’g Tr. at 83:4-15.) Thus, the penitentiary receives between 400 and 500 newly arrived inmates every year. Once he has the numbers, Uttecht contacts the unit managers, the individuals who run the housing units. The purpose of this contact is to “basically find out exactly where they want to assign that particular inmate.” (Vol. XII, at 2298:22-23.) In response to a question put to him by defense counsel as to “what factors are considered by yourself and the other unit managers in determining how many inmates they’re going to be receiving into their unit,” Uttecht answered, “Available bunks.” (Vol. XII, at 2299:16-20.) It is during this discussion about “available bunks” that a decision is made as to specific cell assignments. The information the decision makers have about a newly arrived inmate is very limited. Unless the unit managers know the new inmate, the decision makers have only the “names, numbers, racial breakdown, [and] custody [level].” (Vol. XII, at 2299:21-25.) “[0]n occasion” a “problematic inmate” will have been identified in the information given to Uttecht by Britten. (Vol. XII, at 2300:1-3.) Uttecht freely admitted the paucity of information. For example, Uttecht testified that when cell assignments are made, the housing unit managers normally do not know the age of the inmate, whether he is large or small, or the length of his sentence. (Vol. XII, at 2337:18-2338:18.) A representative sample of Uttecht’s direct examination testimony concerning these points, adduced by defense counsel, is set forth below: Q. Now, your cell assignment has been made though prior to reviewing [the scoring] instrument; is that right? A. That’s true. Q. So that type of information is done after your initial assignment but when you’re meeting with them at the initial class? A. Right. Q. Do you know anything about — what do you know about the inmate who’s coming to the system when you’re making a cell assignment? A. Their name, their custody, their [race], and that’s about it. Q. So what factors do you consider, I guess, is it your testimony that you look at what space is available and what you know about existing inmates in your population as to how you can place an inmate with them? A. Space available and to try to incorporate an appropriate racial breakdown in the unit. (Vol. XII, at 2307:22-2308:15.) In summary, the housing managers know very little about an incoming inmate when they make a cell assignment. They do not review either the “classification study” or the “scoring instrument” prior to making cell assignments. Two factors influence the cell-assignment decision, and those factors are space and race. Accordingly, it is simply not factually accurate to suggest that the cell-assignment decision is a carefully considered decision based on the issue of inmate compatibility. b. The defendants next argue that the process described by Uttecht is not really a cell-assignment decision at all. The defendants claim that because every inmate attends a meeting with Uttecht (called an “initial classification” meeting) shortly after arriving at the penitentiary, any problems with cell assignments can be ironed out at that time because Uttecht will have the “scoring instrument” (but not necessarily the “classification study”) for each new inmate. (Vol. XII, at 2301:2-2302:22.) There are two factual problems with this assertion that render it utterly disingenuous. The first problem is that the meeting relied upon by the defendants takes place after cell assignments have already been made. The second problem is that this meeting is not intended to, and normally does not, change the cell assignments previously made. As Uttecht repeatedly told defense counsel, the “initial classification” meeting takes place after the cell-assignment decision has already been made. (Vol. XII, at 2305:19-25; 2307:22-24.) Indeed, although the meeting is frequently held on Wednesday, when new inmates typically arrive, according to penitentiary regulations this “initial classification” meeting can be held up to seven days after an inmate enters the penitentiary. (Ex. 251, at 1, ¶ 11(F).) Moreover, the cell assignment decision is not normally changed at the “initial classification” meeting unless the inmate can convince Uttecht there is a problem with his cell assignment. In fact, inmate compatibility for cell-assignment purposes is not even one of the reasons for holding the meeting. Clarke admitted that penitentiary regulations setting forth the purposes of the “initial classification” meeting provide “no reference to making any determination as to compatibility of cell mates from that initial classification.” (Vol. VII, at 1144:16-1146:1; Ex. 251.) The inmates confirmed that no questions were asked about cellmate compatibility at the “initial classification” meeting. (Vol. I, at 44:17-45:9.) Even the assistant director of DCS, who was responsible for “classification” for all Nebraska prisons, agreed that neither the “classification study” nor the “scoring instrument” was used “in any attempt to try and predict the compatibility of cell mates.” (Vol. IX, at 1753:2-6.) Teresa Mulkey, a housing unit manager who also chaired the “initial classification” meetings when Uttecht was absent, forthrightly confirmed both that the housing decision was made before the “initial classification” meeting and that the “initial classification” meeting did not deal with cell-assignment issues: Q. Do you receive any custody study or classification study, then, from Diagnostic and Evaluation Center? A. None until after the inmate arrives at the penitentiary. It comes over at the same time the inmate arrives. Q. So by the time you receive that study, have you already decided what cell this inmate’s going to go to? A. Yes. Q. Now, do you interview new inmates before they come into your housing unit? A. Not unless I’m doing initial classification for Mr. Uttecht. Otherwise, no. Q. And sometimes you fill in for him? A. Yes. Q. Is that interview, is that essentially an orientation interview? A. Yes. We give information to the inmates about — we tell them who their counselor’s name is, we let them know where they’re living, we tell them about the orientation to the penitentiary as a whole. We talk to them about the orientation they’ll receive as soon as they get to the housing units. We explain how the doors are run, how to sign up for phones, how to go to meals. Q. Job assignments? A. Right. Just the real basics. Q. They’re — I assume this is an orientation; you’re not actually interviewing them and asking preferences or anything on their part, then, are you? A. Preferences in relation to what? Q. As to who they would cell with or— A. No, we don’t ask them preferences. Q. So there really isn’t any input at that time from the inmates as to who or what type of person they would want to cell with? A. No. Q. Going back to what you said earlier, you have to put the incoming inmates essentially where you’ve got space? A. Yes. (Vol. VII, at 1200:4-13; 1201:3-6, 10-25; 1202:1-9, 22-25.) However, even if Uttecht infrequently asked about cellmate problems at this meeting, since a newly arrived inmate often would not know whether he had a “problem” when he conferred with Uttecht, the initial random cell-assignment decision normally remained unchanged. Cross-examination of Uttecht convincingly emphasized this point: Q. Okay. And at the time that you and these other people from the mental health people [sic] talk with him, he’s already preliminarily been given a cell assignment in a housing unit; hasn’t he? A. Right. Q. And a job? A. Right. Q. And the only time that that cell assignment’s going to change is if he comes in and has a problem with a person that he knows to be in that housing unit? A. Yes, but it doesn’t become official until we go through the hearing. Q. But the only change that’s going to occur is if he has a problem with an inmate in a housing unit and he might change his cell? A. Or his particular job. Q. And the — but as far as this incoming inmate, you still have to place him in a cell probably within the housing unit that’s been selected for him? A. Not always so. Q. Well, unless it’s protective custody or something else he’s going to have to go into one of the four main housing units? A. Right. Q. And the only reason to change from the one that’s been previously selected for him is if there’s a known problem with one of the inmates already in that housing unit? A. That’s correct. Q. Is there any way that this incoming inmate is going to know who’s in that housing unit to alert you that he may have a problem with him? A. Sometimes they know, sometimes they don’t. (Vol. XII, at 2335:13-25; 2337:1-4; 2342:6-23 (emphasis added).) Donn Beaver, a housing unit manager, gave similar testimony. (Vol. VI, at 977:6-978:15.) That the cell-assignment decision was not normally changed no matter what the inmate said at his “initial classification” meeting was also pointedly illustrated during Judge Poster’s questioning of Uttecht and the subsequent examination of Uttecht by counsel for the plaintiffs. Uttecht testified that even when he learned at the “initial classification” meeting that a newly arrived inmate had a gang affiliation, such information did not change the cell assignment: BY THE COURT: Q. You mentioned at the time of the initial classification inquiry that some questioning is made with respect to gang affiliation. What is done with that information if there’s a positive response? A. There’s a staff member at the institution that maintains gang intelligence for monitoring and supervision, keep an eye on cliques, et cetera. Q. What does that do, if anything, to your initial cell placement? A. I can’t say that it’s had any effect at this time. THE COURT: That’s all my questions. Ms. Weeks [defense counsel], anything on my questions? MS. WEEKS: No, Your Honor. THE COURT: Mr. Hemmerling [plaintiffs’ counsel]? BY MR. HEMMERLING: Q. Mr. Uttecht, about gang affiliation, have you ever had any inmates that have said we’re affiliated with a gang? A. Yes. Q. But that hasn’t changed how you cell them at all? A. Say again, please. Q. That hasn’t changed who you celled them with? A. It hasn’t at this time. (Vol. XII, at 2362:12-2363:12.) In summary, it is factually misleading to suggest that the “initial classification” meeting with Uttecht amounts to the “real” cell-assignment decision. It is equally misleading to suggest that the “initial classification” meeting serves as a meaningful check on the cell-assignment decision insofar as the issue of inmate compatibility is concerned. The fact is that cell assignments for newly arrived inmates are blindly made on a random basis without regard to classification information and cellmate compatibility. C. CLARKE AND HOPKINS ACTUALLY KNEW THAT NEWLY ARRIVED INMATES WERE FREQUENTLY RANDOMLY DOUBLE-CELLED. Clarke and Hopkins argue that the “initial classification” meeting at the penitentiary takes into account classification information, at least informally; therefore, the cell-assignment decisions are not random in terms of inmate compatibility. However, as pointed out earlier, the facts simply do not support this assertion. Moreover, I am convinced that Clarke and Hopkins actually knew of the truly random nature of the cell assignments regarding newly arrived inmates; that is, they actually knew that double-cell assignments were frequently made regarding newly arrived inmates without consideration of classification information and without regard to inmate compatibility. I arrived at this conclusion for the following five reasons. First, both Clarke and Hopkins were intimately familiar with the workings of the penitentiary. They were not ignorant of the random nature of the cell assignments regarding newly arrived inmates because they knew how the prison operated. Clarke began work at the penitentiary in approximately 1982, serving as associate warden, deputy warden, and warden (from June, 1987, through August, 1990). He became director of all Nebraska prisons in August, 1990. (Vol. VII, at 1127:23-1130:5.) Hopkins became deputy warden at the penitentiary in July, 1987, later served as interim warden, and became warden of the penitentiary in January, 1991. (Vol. XV, at 2794:20-25.) According to their own testimony, as wardens these men were “responsible for the overall operation of the Nebraska State Penitentiary.” (Vol. XV, at 2796:7-12.) More specifically, as deputy wardens (second in command), both men had direct ongoing contact with housing unit procedures. The unit administrator, who oversaw all the housing units and supervised the penitentiary’s classification efforts, (Vol. XII, at 2211:23-2212:3; 2237:21-2238:19; 2293:2-3), testified that he reported to the deputy warden. (Vol. XII, at 2285:4 — 16.) Clarke had also been an “associate warden of custody,” and as such was specifically responsible for “safety [and] security” at the penitentiary. (Vol. VII, at 1129:7-8; 22-23.) Immediately before coming to the penitentiary, Hopkins served as security coordinator for all the adult prisons in Nebraska. (Vol. XV, at 2794:20-21.) In this capacity it was his responsibility to conduct “special investigations” and “security audits” of the various adult prisons. (Vol. XV, at 2795:4-9.) As a part of his job, Hopkins was specifically required to review the “procedures at the facilities” regarding, among other things, “safety” and “security.” (Vol. XV, at 2795:10-13.) Second, it was a well-known fact within DCS and the penitentiary that classification information was not used to determine inmate compatibility with regard to newly arrived inmates and that the procedure was “random.” Thus, Clarke and Hopkins cannot credibly feign ignorance. Indeed, Clarke’s deputy, the assistant director of DCS, who was responsible for “classification” for all Nebraska prisons, testified that neither the “classification study” nor the “scoring instrument” was used “in any attempt to try and predict the compatibility of cell mates.” (Vol. IX, at 1753:2-6.) Donn Beaver, a penitentiary housing unit manager, readily accepted the word “random” as a fair description of the cell-assignment procedure. (Vol. VI, at 989:5-19.) Furthermore, Teresa Mulkey, a housing unit manager who also chaired the “initial classification” meetings, testified without equivocation that the housing decision was made before the “initial classification” meeting, that the decision was made before classification information became available, and that the “initial classification” meeting did not deal with cell-assignment issues. (Vol. VII, at 1200:4^-13; 1201:3-6,10-25; 1202:1-9, 22-25.) Third, the statements of Clarke and Hopkins confirm that they understood how the process really worked. Two examples illustrate this point. As Clarke admitted at trial, he knew that prior to the newly arrived inmate’s participation in the “initial classification” meeting, “[usjsudlly, they will already have identified prior to that person coming before that committee where the person’s going to live. Unless something occurs during that process ... that causes them to change their mind, most likely that person will get that cell.” (Vol. VII, at 1154:22-1155:3 (emphasis added).) Clarke actually knew the truth, and in this instance he spoke it. Keeping in mind Utteeht’s repeated direct-examination admission, upon inquiry by defense counsel, that the factor driving the cell-assignment decision was “[available bunks,” (Vol. XII, at 2299:16-20), or “[s]paee,” (Vol. XII, at 2307:22-2308:15), Hopkins’s response to an inmate’s double-cell grievance is instructive regarding his knowledge of how the process really worked. In December, 1992, after much of the testimony in this case had been taken during the summer of 1991, Hopkins responded to an inmate’s cell complaint by stating that “the penitentiary administration will continue to make living assignments based on overcrowded situations.” (Prelim.Inj.Hr’g Tr. at 55:7-56:25 (emphasis added).) Like Clarke, Hopkins’s own words confirm that he actually knew the truth. Fourth, both men participated in the decision to increase the number of double cells used at the penitentiary. They would have known of the intense pressure on penitentiary staff to accommodate approximately 500 newly arrived inmates each year, regardless of compatibility, in a prison whieh was “bursting at the seams.” Or as Mulkey agreed, “you have to put the incoming inmates essentially where you’ve got space.” (Vol. VII, at 1202:22-25.) Clarke testified that he personally participated in the decision-making process regarding increased utilization of “double cells” at the penitentiary and that “during my tenure we increased the number of double cells at the state penitentiary by five cells per unit.” (Vol. VIII, at 1428:8-13.) Clarke was asked “as a warden whether or not we could house more individuals at the state penitentiary, and my response was affirmative.” (Vol. VIII, at 1423:8-11.) In arriving at this decision, Clarke specifically studied “the level of dangerousness of the institution. Things such as assaults, serious assaults.” (Vol. VIII, at 1424:16-18.) Yet, as this court previously observed, Clarke testified that he believed the housing units were “bursting at their seams.” Id. at 1468 (Vol. VIII, at 1406:1-3). Hopkins testified that he participated, along with the former warden (Clarke), in the administrative process which set the “capacity level” of the housing units. (Vol. XV, at 2896:13-20.) Hopkins knew that the “capacity level,” set with his input as deputy warden, was substantially over the design capacity of the housing units at the penitentiary. (Vol. XV, at 2840:24-2842:22.) Moreover, as warden, Hopkins authorized the increased use of double cells. (Prelim.Inj.Hr’g Tr. at 55:7-11.) Yet Hopkins had publicly stated that the “increasing inmate population is the biggest challenge the prison faces,” and he reaffirmed that belief at trial. (Vol. XV, at 2843:2-13.) Fifth, Clarke and Hopkins rely upon the “initial classification” meeting for their assertion that inmate compatibility was in fact considered. Yet the penitentiary regulation that sets out the goals and procedures of the “initial classification” meeting contains nothing whatever to do with cell assignments and inmate compatibility. (Ex. 251.) Clarke and Hopkins knew the content of their own regulation. Such knowledge renders any claim of ignorance of what actually happened at the “initial classification” meeting absurd. Clarke admitted the prison regulation setting forth the goals of the penitentiary’s initial classification meeting did not even mention inmate compatibility. (Vol. VTI, at 1144:16-1146:1; Ex. 251.) He could hardly do otherwise since he personally signed the “initial classification” regulation involved here on July 13, 1987, as warden. (Ex. 251, at 4—5.) Neither could Hopkins deny knowledge of the regulation since he was deputy warden when Clarke signed the “initial classification” regulation for the penitentiary. (Compare Vol. XV, at 2794:22 with Ex. 251.) There is no evidence and no claim that Hopkins amended or revoked the regulation when he became warden. In summary, I have not simply found that Clarke and Hopkins should have known of the real cell-assignment process or that a reasonable person would have known of the real process. Rather, I have found that Clarke and Hopkins actually knew that the cell-assignment procedure for newly arrived inmates was random because it was frequently made without consideration of classification information and without any attempt to determine inmate compatibility. Farmer, — U.S. at -, 114 S.Ct. at 1981-82. D. THE LEVEL OF VIOLENCE AT THE PENITENTIARY, INCLUDING VIOLENCE IN THE DOUBLE CELLS, POSED A SUBSTANTIAL RISK OF SERIOUS HARM TO THE PLAINTIFFS. For the sake of clarity, it is appropriate to reiterate what Judges Piester and Cambridge found with regard to the so-called first component of any Eighth Amendment “failure-to-protect” claim asserted by inmates. Jensen IV, 73 F.3d at 810 (Inmates must first “demonstrate that they are ‘incarcerated under conditions posing a substantial risk of serious harm.’ ”) (quoting Farmer, — U.S. at -, 114 S.Ct. at 1977). It is then appropriate to deal with the defendants’ effort to expand the scope of the remand. 1. The essence of this court’s holding regarding the first component of the Eighth Amendment claim was that the “level of violence in the four main housing units exceeds what can be characterized as an unavoidable level ... [and that such] violence has carried over into the double cells....” Jensen I, 807 F.Supp. at 1483. This court specifically held: The level of violence in the four main housing units exceeds what can be characterized as an unavoidable level. That violence has carried over into the double cells where tensions are increased by the limited cell size, noise, lack of privacy, restricted surveillance of the inmates confined therein, deterrents to the reporting of assaults and other violations of prison rules by a cellmate, suspicions about the presence of contraband, and the large amount of time spent confined on a lockdown status with a cellmate. Id. This decision was based upon compelling evidence presented to the court during 18 days of trial: 1. The housing units have been 40 to 65 percent over capacity for 10 years, with the “overcrowding” trend consistently increasing over time. Id. at 1468. 2. Two inmates are housed together in a 74-square-foot cell designed for one person. Id. 3. Tensions in the double cells are increased by: (a) excessive noise, id. at 1469; (b) the adverse impact on ventilation of welding shut windows in the double cells, id. at 1470; (c) the nearly complete lack of visual surveillance for the purpose of detecting violence due to the construction of the double cells and the cell block, id. at 1470; Exs. 6B & 6C; (d) the limited value of the intercom system for the purpose of detecting violence in the double cells, id. at 1471; (e) a reluctance on the part of inmates to report the misconduct of other inmates, id. at 1472; (f) a prison rule that “is a substantial deterrent” to the reporting of violent altercations because it requires that the victim receive a misconduct report and go through disciplinary proceedings, id.; (g) a prison rule that is a deterrent to the reporting of violent altercations because it does not permit prison officials to issue a misconduct report without corroboration, id.; (h) a contraband rule that allows one cellmate to be punished for the contraband of the other cellmate, id. at 1475; (i) the lack of privacy occasioned by placing two men in a cell designed for one man, id. at 1476; (j) the fact that most inmates spend 10 to 14 hours a day in the double cells, id. at 1478. 4. Statistical information maintained by the defendants indicates “a grim picture of a growing problem with violence at the penitentiary” generally, including information that the number of inmates convicted of violent offenses (assaults, aggravated assaults, fights, and threats of bodily injury) more than tripled from 1983 (77) to 1990 (289) as the average number of inmates over design capacity rose from 41 percent in 1981 to 60 percent in 1990. Id. at 1468 n. 4, 1472-73. (Compare Exs. 291-299 with Ex. 321.) Moreover, the number of inmates cheeking into protective custody increased dramatically (by as much as 100 percent). Id. at 1472. 5. There was reason to believe the statistical information substantially underrepresented the level of violence in the penitentiary because of the inmates’ reluctance to inform on each other and because of two prison rules that discouraged the reporting of incidents of violence. Id. at 1472. 6. “Not only is the penitentiary a violent place institution wide, there was also evidence that a problem with violence exists within the double cells.” Id. at 1473. In particular, various inmates testified that violent confrontations between cellmates were routine, and the inmates frequently gave specific examples. Id. at 1475 (inmate Benzel, Vol. I, at 53:8-23; inmate Lyman, Vol. I, at 111:14-24, 116:13-117:9, 119:11-16; inmate Bradley, Vol. II, at 240:22-241:25; inmate El-Tabech, Vol. II, at 316:11-13; inmate Luna, Vol. Ill, at 415:21-23; and inmate Crisp, Vol. Ill, at 461:5-8). 7. The violence in the double cells was poignantly illustrated by the testimony of two inmates, one a perpetrator (inmate Hart) and the other a victim (inmate Jensen). (a) Inmate Hart, “an aggressive, loathsome individual,” testified about his predatory behavior with “complete indifference.” The defendants “provided no factual basis to conclude that Hart has exaggerated or fabricated his exploits.” Id. at 1473-75. Hart has been continually double-celled despite the following: (i) While in prison in Oregon for attempted murder, Hart was celled with another man. Hart assaulted his cellmate and “slit his throat,” “sliding] [him] from the right shoulder ... and up from his ass.” After using a hammer to beat the man’s head, Hart walked away when he thought the man was dead. Hart was then transferred to Nebraska. Id. at 1473-74 (Vol. V, at 835:20-837:21; 867:5-7). (ii) When Hart arrived at D & E, he reported the incident in Oregon. After he assaulted an inmate who was celled with him at D & E, Hart was given a single cell. Upon being transferred to the penitentiary, however, Hart was immediately double-celled. He had a “confrontation” with his cellmate, and the cellmate moved out. Hart requested a single cell, but his requests were denied. Id. at 1474 (Vol. V, at 839:22-840:4; 840:16-841:16; 842:3-7). (iii)Hart subsequently engaged in various types of assaultive behavior against his cellmates. These incidents ranged from sexual assaults (only “four out of seventeen [cellmates was I] unable to encourage or eoerce into a homosexual act”) to other types of assaults (placing Clorox in his cellmate’s eye solution and toilet bowl cleaner in an inhaler) to extortion (of an older inmate who had a heart problem). Five of Hart’s cellmates checked into protective custody. Id. at 1474-75 (Vol. V, at 851:22-853:19; 855:2-857:5; 859:21-860:24; Vol. VI, at 878:11-24; 891:1-6; 893:19-23; 895:15-16). (b) Inmate Jensen was transferred from a minimum security facility to a double cell in one of the housing units. Jensen was a relatively small (5 feet 10 inches, 145 pounds), 21-year-old man incarcerated for theft. Id. at 1473 (Vol. IV, at 598:4-25; 612:3-10). This is what happened to him three days after he was first placed in a double cell: (i) Jensen was housed with inmate Svitak, a tall, 250-pound, muscle-bound man about 45 years of age who was serving a life sentence. Svitak had previously assaulted a staff member, and he made it clear to the staff that he did not want a cellmate. Id. (Vol. IV, at 600:10-23; 605:6-24; Vol. XIII, at 2661:17-23; Ex. 22, at 39). (ii) On September 27, 1987, Jensen and Svitak met. Their meeting took place during the lockdown preceding the evening meal. Svitak told Jensen he did not want a cellmate and Jensen should move out. Jensen told Svitak he had already asked to be placed in a cell with his brother, who was also at the penitentiary. The next day Svitak reminded Jensen “in a serious manner” that he needed to move out. Jensen then sent a written request to Donn Beaver asking to be moved. Id. (Vol. IV, at 604:15-605:14; 606:9-20; 607:7-17). (iii) On September 30, Jensen was awakened by Svitak’s punching him in the face. The cell was locked, as it always was in the morning. Jensen screamed for help. The incident continued for 10 to 15 minutes. When the door was unlocked, Jensen left the cell and Svitak went to breakfast. Jensen was moved to a another cell. Although Jensen reported the assault, no action was taken against Svitak because there was no corroborating evidence. Id. (Vol. IV, at 610:6-22; 611:4-14; 612:14-23; 613:1-614:9; 619:10-15; 620:12-13; 621:6-11; 622:10-12; Vol. XIII, at 2481:14-19). 2. Much of Clarke and Hopkins’s brief on remand is an attempt to reargue these findings. However, as the plaintiffs point out, the court of appeals’ remand order is clear and limited. As observed earlier, the court of appeals directed me to specifically “find whether the defendants actually knew of and disregarded a substantial risk to the safety of the plaintiffs.” Jensen IV, 73 F.3d at 811. The court of appeals was specifically concerned with “the subjeetive-state-of-mind component of deliberate indifference described in Farmer.'’ Id. 73 F.3d at 811. The Jensen IV court reserved judgment on whether the plaintiffs were exposed to a substantial risk of physical harm. Jensen IV, 73 F.3d at 810 (“We do not reach the first issue, because we must remand this case for further findings on the second.”). Thus, the Jensen IV remand order is limited to the “second” question of whether the defendants were “subjectively aware of a substantial risk of harm to the plaintiffs.” Id. at 811. I therefore make no effort to address the arguments of Clarke and Hopkins that the plaintiffs failed to prove the first component of their Eighth Amendment claim. E. CLARKE AND HOPKINS ACTUALLY KNEW THE LEVEL OF VIOLENCE AT THE PENITENTIARY, INCLUDING VIOLENCE IN THE DOUBLE CELLS, POSED A SUBSTANTIAL RISK OF SERIOUS HARM TO THE PLAINTIFFS. Once again, I reiterate that the pertinent question is not whether Clarke and Hopkins should have known of the substantial risk of harm or whether a reasonable person would have known of the substantial risk of harm. Rather, the pertinent question is whether Clarke and Hopkins actually knew the level of violence at the penitentiary, including violence in the double cells, posed a substantial risk of serious harm to the plaintiffs. Farmer, — U.S. at -, 114 S.Ct. at 1981-82. As the Farmer Court recognized, this is not an unusually difficult question to answer. Essentially, the question of “[wjhether a prison official had the requisite knowledge of a substantial risk [is] a question of fact subject to demonstration in the usual ways, including inference from circumstantial evidence, ... and a factfinder may conclude that a prison official knew of a substantial risk from the very fact that the risk was obvious.” Id. at -, 114 S.Ct. at 1981 (citations omitted). Simply put, Farmer does not require the plaintiffs to produce a “smoking gun” in order to prevail. 1. I find and conclude that both Clarke and Hopkins actually knew the level of violence at the penitentiary, including violence in the double cells, posed a substantial risk of serious harm to the plaintiffs. I arrived at this conclusion for the following five reasons. First, both defendants were specifically provided with credible evidence of a substantial level of violence on a monthly basis. From this evidence alone, Clarke and Hopkins would have known the level of violence in the penitentiary, including violence in the double cells, posed a substantial risk of serious harm to the plaintiffs. For example, Clarke was actually informed of virtually every verified incident of violence that had occurred at the penitentiary since at least 1983. Likewise, Hopkins was actually informed of virtually every verified incident of violence that had occurred at the penitentiary since he became deputy warden in 1987. This conclusion is inescapable because the monthly disciplinary reports from 1983 to 1991, (Exs. 291-299), which showed, among other things, disciplinary convictions for assaults or threats of violence, were distributed to the warden and deputy warden. Jensen I, 807 F.Supp. at 1472. Indeed, many of the reports received in evidence show a distribution legend indicating they were distributed to “Warden Clarke” and “Deputy Warden Hopkins.” (Exs. 295(f)-296(i).) Moreover, Clarke admitted that he received reports of sexual assaults at the penitentiary even after he became director of DCS. (Vol. VIII, at 1388:22-1389:5.) Clarke also admitted reading the reports when he stated that “[i]f they were reported, I would know about it.” (Vol. VIII, at 1388:24-25.) Hopkins admitted that since becoming deputy warden he also received reports showing not only the convictions, but also the number of charges (misconduct reports). (Vol. XIV, at 2784:10-25; Vol. XV, at 2870:22-2871:20.) Hopkins also admitted reviewing the reports. (Vol. XV, at 2864:14r-2865:3.) The reports that Clarke and Hopkins received and read demonstrated there was a substantial risk of assault virtually anywhere within the penitentiary. From this evidence alone, it would have been clear to anyone who cared to see that the level of violence within the cells at the penitentiary was very significant. For example, even Clarke and Hopkins admit that the reports delivered to the warden and the deputy warden during this period of time showed there were 1,774 convictions for assaults and threats of assault. (Defs.’ Br. at 10.) Thus, as Clarke and Hopkins also admit, during the time frame captured by the evidence, there was a verified assault or threat of assault at the penitentiary “every other day.” (Id. at 13 (emphasis added).) Second, both Clarke and Hopkins were well aware that the statistical information underestimated the substantial risk of violence, particularly violence in the double cells. For example, Clarke testified: I know that offenders do not always report incidents, happenings within the institution, either may they be assaults or other observations that they may have made. I have reason to believe that offenders do not always report problems for a number of reasons. One of those reasons might be for fear of retaliation from other offenders within the institution, or from the person who is the aggressor. (Vol. VIII, at 1420:19-1421:4.) When asked whether “those fears of retaliation would be increased if the assailant were a cell mate,” Clarke, in a guarded understatement, answered: “One could so assume.” (Vol. VIII, at 1421:5-8.) Clarke quickly added, however, that if “they can give staff reasons” then “that can be corrected.” (Vol. VIII, at 1421:8-13.) However, this qualification seems perverse given Clarke’s knowledge of the lack of surveillance in the cells and the two prison rules that deterred reporting of assaults or threats of assault. As noted earlier, their construction makes it very difficult, if not impossible, to visually monitor the cells during the 10 to 14 hours each day that two cellmates are locked in. Jensen I, 807 F.Supp. at 1470 (Exs. 6B & 6C). Moreover, it was obvious the intercom was not a viable substitute for visual surveillance. Jensen I, 807 F.Supp. at 1471 (Vol. IV, at 613:1-5) (Jensen unavailingly yelled for help just three feet from his door while being attacked); (Vol. VI, at 1060-74) (sound of Harpster’s head being pummelled into table by cellmate undetected by intercom system). Both Clarke and Hopkins clearly knew how the cells were constructed because they toured the housing units as frequently as once a month. (Vol. XVIII, at 3623:9-11.) Indeed, Clarke admitted that “we are not in a position to observe those individuals from a central location.” (Vol. VIII, at 1425:20-25.) Clarke and Hopkins were thus aware of the obvious difficulty in monitoring the cells. Moreover, both men knew the prison rules. (Vol. XVIII, at 3641:19-3642:5.) They knew of one prison rule that was “a substantial deterrent” to the reporting of violent altercations because it required that the victim receive a misconduct report and go through disciplinary proceedings. Id. at 1472. They also knew of another prison rule that deterred the reporting of violent altercations by not allowing prison officials to issue a misconduct report without corroboration. Id. Accordingly, Clarke and Hopkins knew that an inmate who spends 10 to 14 hours in a double cell is not likely to report a beating or threat of a beating by his cellmate if that inmate understands the prison rules and cannot provide his guards with corroboration. This painfully obvious point was illustrated by the testimony of inmate Jensen, only recently arrived at the penitentiary, regarding the beating he suffered at the hands of inmate Svitak after being placed in a double cell. Id. at 1473. Within three days of his arrival, Jensen was assaulted in his cell. The beating lasted 10 to 15 minutes, and Jensen screamed as the assault took place. There was no intervention by the guards. “Jensen reported the assault, but prison officials took no action. There was no corroborating evidence.” Id. Third, as participants in the decision to increase utilization of double cells, both Clarke and Hopkins knew that verified reports of violence increased as the number of double cells increased. Therefore, they knew that as the prison became more crowded, violence increased throughout the penitentiary. As a factual matter, it - is admitted that both Clarke and Hopkins participated in the decision to increase utilization of double cells at the penitentiary. For example, Clarke testified that “during my tenure we increased the number of double cells at the state penitentiary by five cells per unit.” (Vol. VUI, at 1423:8-13.) Clarke was asked “as a warden whether or not we could house more individuals at the state penitentiary, and my response was affirmative.” (Vol. VIII, at 1423:8-11.) In arriving at this decision, Clarke specifically studied “the level of dangerousness of the institution. Things such as assaults, serious assaults.” (Vol. VIII, at 1424:16-18.) The evidence regarding Hopkins was similar. Hopkins testified that as deputy warden he participated in the administrative process that set the “capacity level” of the housing units. (Vol. XV, at 2896:13-20.) Hopkins knew that the “capacity level,” which was set with his input as deputy warden, was substantially over the design capacity of the housing units at the penitentiary. (Vol. XV, at 2840:24-2842:4.) Moreover, after becoming warden, Hopkins authorized the increased use of double cells. (Prelim.Inj.Hr’g Tr. at 55:7-11.) As Hopkins stated, he knew that the “increasing inmate population is the biggest challenge the prison faces.” (Vol. XV, at 2843:2-13.) As a factual matter, it is also undisputed that as the number of double cells increased, so did the convictions for assaults and threats of assault. In fact, the increase in violence was more geometric than arithmetic; that is, the number of inmates convicted of violent offenses more than tripled from 1983 (77) to 1990 (289) as the average over design capacity rose from 41 percent in 1981 to 60 percent in 1990. Id. at 1468 n. 4, 1472-73. (Compare Exs. 291-299 with Ex. 321.) As documented earlier, it is also undisputed that Clarke and Hopkins received monthly reports showing the increasing number of convictions. Accordingly, because Clarke and Hopkins were among the decision makers regarding the double-cell policy and were given specific statistical information on a monthly basis showing that convictions for violence rose as double-celling increased, it is obvious they knew of the “positive correlation between the amount of double celling and the increase of violent incidents” at the penitentiary. Jensen I, 807 F.Supp. at 1473. It strains credur lity for Clarke and Hopkins to argue they were blind to this patently evident reality. Fourth, as Frank Gunter, a defendant in this ease, has admitted, it is a “widely accepted principle of prison management” that “a prison cell should not be used for more than one prisoner” because the ability to manage the prison population erodes and the potential for violence increases when prisoners are double-celled. This being true, Clarke and Hopkins knew it was a risky and dangerous business to put two men into a cell built for one man. In September, 1989, while Clarke was warden and Hopkins deputy warden, Gunter, who was then director of all Nebraska prisons, told the Nebraska legislature’s committee on prison overcrowding: A widely accepted principle of prison management dictates that a prison cell should not be used for more than one prisoner. The rationale for this is obvious. People who are in prison have demonstrated a difficulty in obeying societal laws, ... rules and regulations, and in many cases have demonstrated a difficulty in getting along with others. A significant number of these prisoners either are serving time for a current violent crime or have a history of violent offenses. If a prison with 100 cells must accommodate 120 prisoners, then 40 of its prisoners, or one-third, will be housed two to a cell. Prison administrators generally agree that when a prison population exceeds capacity, their ability to manage the inmate population begins to erode. As the numbers of prisoners increase, the following scenarios develop: .... This lessening of control increases the probability of inmate problems and potential violence. (Ex. 23 Gunter Test. Before the Comm, on Prison Overcrowding of the Neb. Leg. 1, 21-22 (Sept. 22,1989) (emphasis added).) In January, 1990, referring to all Nebraska prisons generally and the penitentiary specifically, Gunter told the Nebraska legislature: The Nebraska State Penitentiary is 145 percent of capacity.... The overcrowding problem and the impact of that problem is now critical and requires immediate at-tention_ The department’s budget deficit request submitted in November, 1989 addresses how present prison overcrowding coupled with projected growth has significantly impacted Nebraska’s inmate system in three important ways. Number one, it has critically reduced the department’s ability to operate safe and secure institutions. (Ex. 24 Gunter Test. Before the Comm, on Appropriations of the Neb.Leg. 1, 2 (Jan. 25, 1990) (emphasis added).) Given the level of violence in the penitentiary generally, Clarke and Hopkins were well aware as experienced administrators that prisoners would be subjected to a substantial risk of violence in the double cells. Clarke and Hopkins cannot credibly deny knowledge of this “widely accepted principle of prison management.” After all, it was their boss, Frank Gunter, who told the Nebraska legislature about what “[pjriso'n administrators generally agree” upon. Moreover, it was their boss who told the Nebraska legislature that the Nebraska prison system, including the penitentiary, suffered from a “critical” reduction in the “ability to operate safe and secure institutions.” Fifth, the statements of Clarke and Hopkins establish their actual knowledge of the substantial risk of harm. Although numerous examples could be used, three of Clarke’s and Hopkins’s statements make the point: 1. Based upon his monthly review of the reports of violence since 1987, Hopkins con-eluded that misconduct reports had increased “quite a bit.” (Vol. XV, at 2836:16-22; 2869:13-2871:15.) Hopkins described his knowledge this way: “The numbers of misconduct reports have gone up, and my understanding is quite a bit.... I suppose it could be — you could say substantial. I know that we’ve had a definite increase in misconduct reports that are lodged against inmates.” (Vol. XV, at 2870:16-21 (emphasis added).) These reports were very significant because, according to Clarke, a misconduct report was not written by prison employees “unless there are — unless acts have been committed which the reporter believes constitutes a violation of the code of offenses.” (Vol. XVIII, at 3640:20-23.) 2. After becoming director of DCS, Clarke told the Nebraska legislature, referring to the penitentiary specifically and Nebraska prisons generally, that “predatory behavior is increasing and weaker inmates are forced to seek protective custody, which has increased by 70 percent in less than two years.” (Vol. VII, at 1268:5-10 (emphasis added).) At trial, Clarke stated that the number of inmates seeking protective custody at the penitentiary had at least doubled from 20 to 40 between 1989 and the time of trial in 1991 (a 100-percent increase in two years). (Vol. VIII, at 1390:18-1391:25.) In fact, Clarke testified that “we have had a waiting list at the penitentiary ” to get into protective custody for “quite some time.” (Vol. VIII, at 1391:10-11 (emphasis added).) 3. Clarke further testified at trial that “I believe that, yes, we are seeing signs of dysfunction,” the “signs of dysfunction comes in the form of it becoming more difficult to find individuals who are compatible,” and “this is all due to overcrowding.” (Vol. VII, at 1250:15-1251:6.) Clarke was “referring to inmates who are [not] compatible.” (Vol. VII, at 1251:7-11) (emphasis added).) He understood “compatibility” to mean the ability “to place offenders together who are not going to have conflicts with each other.” (Vol. VII, at 1251:12-17) (emphasis added).) According to Clarke’s own admission, he knew that “[w]e have had a crisis in Nebraska for quite some time and we have been saying so.” (Vol. VII, at 1255:4^5) (emphasis added).) 2. Clarke and Hopkins assert a variety of arguments in an effort to defeat the finding and conclusion that they actually knew the level of violence at the penitentiary, including violence in the double cells, posed a substantial risk of serious harm to the plaintiffs. I am not persuaded by any of these arguments and will address only those that warrant a response. First, Clarke and Hopkins have misunderstood the significance of the penitentiary-wide statistics insofar as their knowledge is concerned. They want to ignore the penitentiary-wide statistics on the theory that unless the statistics are limited to the cells, the numbers are meaningless insofar as their knowledge is concerned. Kindly put, this argument is not worthy of belief. Without ambiguity the statistics showed (and this court repeatedly found) that a pervasive risk of harm to inmates from violence existed throughout the penitentiary. Quite apart from the testimony of the inmates that assaults in the cells were routine, it was obvious to anyone that the violence did not somehow magically stop at the cell doors. In fact, the “positive correlation between the amount of double celling and the increase of violent incidents” was plain for all to see. Jensen I, 807 F.Supp. at 1473. There is no basis to conclude that Clarke and Hopkins actually believed the cells were a safe haven from violence, given the statistical evidence that conclusively proved a violent assault or threat of assault took place “every other day” throughout the penitentiary. Second, Clarke and Hopkins have misapplied their own statistics in an effort to show they lacked knowledge that cells are violent places. In essence, they suggest the numbers do not reveal a serious problem in the cells. The manner in which they reach this conclusion is, to say the least, unconvincing. For example, Clarke and Hopkins purportedly endeavor to examine the statistics regarding assaultive behavior in the cells. They use the conviction data, (Exs. 291-299), for assaults and threats of assault documenting 1,774 convictions during the relevant time frame. They concentrate on 130 incidents involving assaults by inmates upon inmates, but not threats of assault. By thus limiting their inquiry, Hopkins and Clarke maintain they are able to identify 44 incidents where there is a short narrative statement of what happened. These narratives sometimes, but not always, identify a general location. Using the 44 incidents where there are narratives, Clarke and Hopkins suggest that only 5 of the assaults took place in the cells, although they admit they do not know where 10 of the 44 assaults took place. (Defs.’ Br. at 10-12 & App. A at 72.) Clarke and Hopkins thus argue that they had no reason to believe the cells were dangerous places. However, even a cursory examination of the methodology employed by the defendants makes clear that their statistical sleight of hand proves nothing which helps them. Four illustrations make the point: 1. Clarke and Hopkins ignore 86 of