William Langton, Plaintiff, Appellee, v. Philip Johnston, et al., Defendants, Appellees,john Bruder, et al., Plaintiffs, Appellants.john Bruder, et al., Plaintiffs, Appellants, v. Philip Johnston, et al., Defendants, Appellees.john Bruder, et al., Plaintiffs, Appellees, v. Philip Johnston, et al., Defendants, Appellants, 928 F.2d 1206 (1st Cir. 1991)

Plaintiffs, representatives of a class consisting of patients at a state institution, the Bridgewater, Massachusetts Treatment Center for Sexually Dangerous Persons (the Treatment Center), assail the conditions of their confinement. The defendants are all state officials, including the Secretary of Human Services, the Commissioner and Assistant Commissioner of the Department of Mental Health (DMH), the Commissioner of the Department of Corrections (DOC), and the superintendent of the Bridgewater State Prison. Three Treatment Center officials, Ian Tink, Dennis McNamara, and Michael Stevens, are also named defendants. The plaintiffs appeal from the district court's refusal to hold defendants in contempt of certain preexisting decrees and from the court's denial of equitable redress. Additionally, both sides challenge the court's order regarding attorneys' fees.¹ 

Having followed the threads of the parties' variegated arguments through the cumbersome record, we affirm the district court's judgment on the merits but vacate the fee award.

In 1987, appellants filed an amended complaint in the United States District Court for the District of Massachusetts claiming that defendants had violated two prior consent decrees and the federal Constitution. Charging an institutional failure to provide adequately for the plaintiff class, appellants sought the appointment of a receiver or a wholesale release of inmates (some to be discharged into society and the rest transferred to state prison, depending on their individual circumstances). The district court consolidated plaintiffs' case with an earlier complex of cases challenging the Treatment Center's sequestration procedure and practices. See generally Pearson v. Fair, 808 F.2d 163 (1st Cir. 1986) (per curiam) (describing Pearson litigation); King v. Greenblatt, 489 F. Supp. 105 (D. Mass. 1980). The cases were tried sequentially before Judge Mazzone in the spring of 1989 and were resolved in a single opinion. In a later opinion, Judge Mazzone awarded partial attorneys' fees to the plaintiffs. Neither rescript has been published.

Appeals were taken in all cases. We consolidated the matters for oral argument. Unlike the court below, however, we prefer to write separately about each set of appeals. This opinion will focus on the so-called Bruder appeals, the titles of which are delineated in the caption. The Bruder appeals primarily involve adequacy of treatment and matters ancillary thereto. We will address the so-called Pearson appeals, which deal principally with sequestration issues, in a later opinion.II. THE ANATOMY OF THE TREATMENT CENTER

In an effort to bring perspective to the mass of issues, we believe it helpful to sketch an overview of the Treatment Center and its purposes.

A. The Statutory Framework for Commitment.

Under Massachusetts law, a sexually dangerous person (SDP) is

any person whose misconduct in sexual matters indicates a general lack of power to control his sexual impulses, as evidenced by repetitive or compulsive sexual misconduct by either violence against any victim, or aggression against any victim under the age of sixteen years, and who, as a result, is likely to attack or otherwise inflict injury on such victims because of his uncontrolled or uncontrollable desires.

Mass.Gen.L. ch. 123A, Sec. 1 (1986). The DMH is directed to establish and maintain a facility for the "care, custody, treatment and rehabilitation" of such persons. Id. at Sec. 2. A proceeding to determine whether an individual should be committed as sexually dangerous can be initiated in any of three ways: upon motion of the state trial court or commonwealth (following conviction for some specified sexual offense), id. at Sec. 4; at the request of the head of a correctional facility (following sexually assaultive behavior on an inmate's part while in custody), id. at Sec. 6; or upon voluntary application, id. at Sec. 7.² 

Although the preliminary steps differ with respect to different classes of involuntary initiates,³  the adjudicative processes for determining SDP status are largely the same. First, the individual is committed to the Treatment Center for a period of up to sixty days and examined by two qualified specialists. Id. at Secs. 4, 6. During this period, the examiners submit a report and recommend a disposition. Id. If the report does not "clearly indicate" sexual dangerousness, no further proceedings are conducted and the individual is not sent to the Treatment Center. See id. at Secs. 5, 6. If, however, the report "clearly indicates" that the individual is an SDP, the matter goes before the state superior court. Id. at Sec. 5. If the court concludes, after a hearing, that the commonwealth has carried the burden of proving sexual dangerousness beyond a reasonable doubt, see Commonwealth v. Walsh, 376 Mass. 53, 55, 378 N.E.2d 1378 (1978) (discussing burden of proof), it can then "commit such person to the center ... for an indeterminate period of a minimum of one day and a maximum of such person's natural life." Mass.Gen.L. ch. 123A, Sec. 5. Whichever route is followed, the Treatment Center has no authority to reject placements. It is similarly without authority to expel patients who obstruct, ignore, or resist available treatment, or to remit such patients to any other facility.

Before 1986, chapter 123A did not allow the superior court to sentence a defendant and commit him to the Treatment Center. Rather, the court had to impose the commitment in lieu of sentencing. Hence, if the referral stemmed from a criminal conviction, the commitment would become, in effect, a part of the individual's sentence. If the referral originated in the course of an individual's imprisonment, or after 1986, the criminal sentence would run concurrently with the individual's confinement at the Treatment Center. See Mass.Gen.L. ch. 123A, Sec. 5.⁴ 

Once committed, a patient remains at the Treatment Center until the superior court determines that he is no longer an SDP. Mass.Gen.L. ch. 123A, Sec. 9 (Supp.1990). A patient may file a petition for release once a year, although the DMH may file such a petition at any time. Id. The commonwealth receives, and often uses at judicial hearings, reports issued by the Restrictive Integration Review Board (RIRB). The RIRB is composed of three Treatment Center staffers and three outside clinicians. Its primary duty is to evaluate, no less than annually, each patient committed to the Treatment Center in order to determine the progress of therapy and the advisability of permitting him to reenter the outside community on a limited basis. See id. at Sec. 8.

At a release hearing, the commonwealth must prove beyond a reasonable doubt that the patient is still sexually dangerous. Walsh, 376 Mass. at 55, 378 N.E.2d 1378. If the commonwealth does not carry the devoir of persuasion, the petitioner, depending on the status of his original conviction, is either released unconditionally or transferred to DOC custody. See Mass.Gen.L. ch. 123A, Sec. 9.

B. The Patient Population.

The Treatment Center population consists exclusively of involuntarily committed males. Every inmate has been convicted of a sexual offense identified by Massachusetts law as indicative of compulsive sexual behavior. The typical accrued duration of stay is around ten years. Approximately 50% of the patients were committed following convictions for child molestation and the remainder were committed after being convicted of rape. In March 1989, there were 286 patients in custody. Their ages ranged from 20 to 78. The class in this case comprises 264 indefinitely committed patients (a figure which excludes the 22 persons who were being held for temporary observation).

The class is divided into two subclasses: those who do, and do not, labor under the added weight of unexpired criminal sentences. The first subclass, numbering 191 men, can be further divided: 75 of the members were parole-eligible at time of trial while the remaining 116 were not. As to those persons who were sentenced before arriving at the Treatment Center, 30 were serving life sentences. The average length of sentence referable to the non-lifers is about 15 years. The second subclass, 73 in number, can also be divided along a more sharply defined axis: 27 men were never sentenced because their convictions and commitments occurred prior to 1986, whereas the other 46 were originally given criminal sentences which expired before March of 1989.

For purposes of our preliminary discussion, we examine the record from two perspectives. The gravamen of the plaintiffs' claims is the defendants' alleged noncompliance with the dictates of the decrees previously entered in the so-called Williams cases. We limn these agreements and succinctly summarize the litigation that produced them. We then turn to the present litigation.

A. The Earlier Litigation.

In November 1972, patients at the Treatment Center filed a federal court complaint alleging that living conditions and programs were "substantially inferior to those existing at correctional institutions of the Commonwealth." The complaint charged that the dearth of psychiatric treatment, work opportunities, job training programs, and educational facilities violated the fifth, eighth, and fourteenth amendments to the federal Constitution.

The case was assigned to Judge Wyzanski, now deceased. Two consent decrees resulted. We have set out the relevant portions of each decree in an appendix. The first, entered in June 1974, provided in substance that stewardship of the Treatment Center would be vested in DMH rather than DOC; that a graduated system would be developed to ensure that patients were subjected to the "least restrictive conditions necessary" to achieve full recovery; that vocational and other programs would be implemented; and that rules and regulations governing the daily operation of the Treatment Center would be promulgated. In due course, regulations were prepared. See 104 C.M.R. Sec. 8.00 et seq. (1988). The provision governing administration of treatment capsulates the imperatives of the 1974 decree in the following way:

Every patient shall be offered treatment designed to effect his early return to public society. Such treatment shall consist of medical, psychiatric, educational, vocational, rehabilitative, and social services commensurate with the patient's age, background, abilities, and physical and mental condition. Such treatment shall be administered skillfully, safely and humanely, with respect for the patient's dignity and personal integrity, and in the least restrictive conditions which are consistent with his security needs.

Id. at Sec. 8.03(3).

The second consent decree, entered in January 1975, ordered the defendants to submit a plan describing the therapeutic, educational, vocational, and avocational programs to be offered at the Treatment Center. It also ordered the commonwealth to develop a short-term release program whereby patients would be allowed to partake in outside activities. Shortly after this decree was entered, the case was transferred to Judge Garrity, who oversaw the litigation with skill and diligence for the next five years. During this period, he entered a decree which provided that the Treatment Center would move into a new facility to be constructed at the Bridgewater correctional complex. Some two years later, in 1978, Judge Garrity ordered the defendants to promulgate regulations for a furlough program.⁵ 

In a routine reshuffling of the docket, the case was transferred to Judge McNaught in 1979. It remained under his careful supervision until its completion. None of the orders issued by Judge McNaught have any real bearing on the interpretation to be accorded the consent decrees here at issue. We say this despite our awareness that the parties dispute when the original case was formally closed: appellants argue that it ended in September 1986 while appellees place the time of termination in late 1989. We decline to enter this Serbonian bog, believing the date of closure to be immaterial for our purposes.

B. The Present Litigation.

In April 1986, six patients filed a pro se complaint which outlined a golconda of alleged constitutional violations. The plaintiffs' core theses were that (1) no adequate treatment was being provided, and (2) the Treatment Center was being run, impermissibly, as a correctional facility rather than as a mental health institution. The case was drawn by Judge Young. After counsel were appointed to represent them, the plaintiffs moved for a preliminary injunction and designation of a special master. They sought, among other things, to restrain defendants from conditioning maximum privileges on a patient's consent to "double-bunking," i.e., sharing a cell with another patient. The plaintiffs also asked that the court order the commonwealth to fill all clinical staff vacancies, develop an individual treatment plan for each patient, abolish the movement control pass procedure, and scrap the package restriction policy.

In June 1987, Judge Young denied this motion for the most part. As to one plaintiff, John Bruder, unique among the five plaintiffs in that his criminal sentence had expired, the court concluded that some interim relief was merited; since Bruder would be free but for his commitment to the Treatment Center, he was entitled to all available treatment whether or not he consented to double-bunking. Accordingly, the court enjoined defendants from denying Bruder "any treatment available to individuals at the Bridgewater Treatment Center who are double bunked."

In due season, the plaintiffs moved to amend their complaint and certify a class. Both motions were granted. The amended complaint alleged that defendants had utterly failed to (1) provide treatment sufficient to bring about the patients' recovery; (2) administer the treatment actually furnished in the least restrictive manner consistent with security needs; and (3) operate the Treatment Center as a mental health facility rather than as a penal institution. The amended complaint averred that, because patients were not receiving proper treatment, they could not be rehabilitated; thus, a person's commitment to the Treatment Center had become tantamount to a life sentence without prospect of parole. Proceeding on this theory, plaintiffs asserted that, from 1986 onward, the commonwealth had run roughshod over both the Constitution and the Williams decrees. The defendants denied the allegations.

Before too long, Judge Young was asked to recuse himself pursuant to 28 U.S.C. § 455(a). He did so. The case, along with the Pearson litigation, was reassigned to Judge Mazzone. Trial ensued. See supra Part I.

The lower court concluded that plaintiffs failed to maintain their burden of showing that defendants' conduct was either constitutionally deficient or contemptuous. The court based its conclusions on an extensive, well-elaborated factual panoply. We set forth below a compressed account of the trial court's integral findings.

A. In General.

Prior to negotiation of the consent decrees, conditions at the Treatment Center were medieval--worse than those obtaining in the prison system. Patients resided in cramped, poorly furnished cells. They used buckets as toilets (being allowed to dump the buckets periodically into a central depository). The water supply came directly from the Taunton River and was untreated. The shower facility was located opposite the housing unit, across an outdoor courtyard; patients had to walk outside to reach the showers, even during the dead of winter.

The Center's rehabilitative effort was in a similar state of disrepair. There was no library. There were no educational programs. The vocational facilities were obsolete. There was no furlough or work-release program of any kind. The Center had neither a gymnasium nor any outdoor recreational areas. DMH's role in running the facility was limited to providing diagnostic procedures, testifying at section 123A proceedings, and conducting a few psychotherapy programs. There were no nurses on staff. Correctional officers dispensed medication.

The consent decrees heralded a fresh start. In due course, a totally new facility was constructed. It opened in 1986. The district court found enormous improvements in the general conditions of confinement from that time forward. After touring the Treatment Center, Judge Mazzone wrote:

Patients now live in humanly scaled housing units of twenty-four or thirty rooms, with a large multi-leveled common area. Depending on the unit's privilege status, washing machines and kitchen facilities may be located in the common area. All individual rooms are well-lit and nicely furnished. Unlike their predecessors, they are equipped with modern toilets and sinks, and ... a shower room is located off the common area within each housing unit.

The court found these renovations to be of "inestimable value" in ameliorating the attitude and demeanor of both patients and staff. The modern setting offered a safe and comfortable place to provide care and custody, thereby facilitating treatment. The court acknowledged, of course, that these physical improvements, while important, were not sufficient in themselves to satisfy the dictates of the consent decrees. Therefore, Judge Mazzone proceeded to ask, and answer, two discrete questions: (1) Was the treatment provided in the new setting "adequate" in light of clinical needs and the indefiniteness of the patients' commitment? (2) Did a correctional philosophy so dominate operations as to impede development of an effective therapeutic regime? We describe the court's responses to these questions in the sections that follow.

B. Adequacy of Treatment.

The district court organized its adequacy-of-treatment findings along five separate strata.⁶  In synthesizing the findings, we follow the same mode.

1. Education; Special Needs. The reading proficiency of 75% of the patient population is below the eighth grade level; only 67 patients (less than 25% of the population) have high school diplomas or the equivalent. Beginning in August 1986, the Treatment Center budgeted for an instructor to lead a special education curriculum. Despite the obvious need for such a program, the position went unfilled for at least a year. In August 1987, Tink, the administrator of the Treatment Center, felt compelled to write to DMH's Assistant Commissioner seeking additional funding to enable the Center to fill staff vacancies in educational, and other, programs.

Recognizing this slow start, the court nonetheless found that defendants were on their way toward implementing a comprehensive educational system. At time of trial, a learning center was up and running and a specialized treatment program was in place for patients with modest intellectual skills. As of mid-March 1989, 102 patients were participating in various special needs/educational programs. Over half of them were enrolled in two or more courses; 34 were receiving classroom instruction for five periods a day, five days a week.

Educational opportunities were also being provided on a more generalized plane. Above and beyond those involved in special needs programming, 105 patients were partaking of other educational endeavors. Thirty-one were attending college-level courses taught by visiting faculty from Massasoit Community College and Bridgewater State College. Twelve were enrolled in a special computer course. Six were taking high school equivalency (G.E.D.) courses (an enrollment which, at trial, defendants acknowledged could, and should, be expanded). Classes were also offered in bibliognostic skills, graphic arts, crafts, and other disciplines.

The Treatment Center library, headed by a certified librarian, was fully operational, containing over 12,000 permanent volumes. The inventory included a current and comprehensive law library. Additional materials were available on revolving loan through a standing agreement with the Bridgewater Public Library. In the month prior to trial, a daily average of 89 patients visited the library; during that month, 233 different patients borrowed a total of 546 books.

In the last analysis, the court, although concerned about staffing shortages and the tiny enrollment in the G.E.D./pre-G.E.D. courses, found the overall improvement in the quantity, quality, and availability of educational programs to be "impressive."

2. Vocational and Avocational Programs. The Treatment Center's vocational and avocational workshops also moved with near glacial slowness after construction was completed. While an avocational/occupational therapy (AVO/OT) area was built as part of the new complex, it remained largely unused for three years due to insufficient staffing. Tink, in his 1987 letter to DMH, wrote that "my vocational training component really includes only two [full-time equivalent staff members].... Particularly because of inadequate staffing, I continue to be unable to open the AVO/OT program areas...."

The glacier did not really begin to thaw for over two years. By March 1989, 195 patients had enlisted in vocational or avocational programs. Various levels of permanent work status were available. A large majority of the participating patients were usefully involved in the construction of greenhouses at the Bridgewater State Hospital. The Treatment Center itself was a further source of employment for kitchen and canteen workers, clerical help, library aides, tutors, and graphic arts assistants. The Brockton Area Private Industry Council was providing career counselling. Nine patients were participating in a work-release program, leaving each day to work at local jobs and returning each evening. The Center also started an avocational workshop, targeted to serve 50 patients. Staffing shortages impeded full implementation of the program; at time of trial, only eight persons were enrolled.

Notwithstanding the three year delay in opening the AVO/OT area, the court found recent progress "encouraging." It viewed staffing difficulties as the largest obstacle to complete implementation of the occupational and avocational programs and instructed the defendants to continue to "employ their best efforts to fund and fill positions which have been allocated for these programs."

3. Recreation. A large gymnasium was built as part of the new complex. It enabled the recreation program to be overhauled and rejuvenated. The gymnasium is equipped with basketball courts, a weightlifting area, pool tables, and other accoutrements. As late as 1988, due to security and staffing problems, the facility was only open on a limited basis. By 1989, however, the gym was accessible seven days a week and offered a wide range of activities. During February 1989, 146 patients per day, on average, used the facility.

The Treatment Center also sports an excellent outdoor recreation area, including a baseball diamond and track. Sedentary forms of recreation, such as board games, are available in the common areas adjacent to each housing unit. A Recreation Advisory Board, composed of patients, has been established to offer suggestions regarding development and administration of additional recreational programs.

It seems clear that, by 1989, the patients' recreational needs were being more than adequately met. The district court, in effect, so found.

4. Therapeutic Programs. Under the new regime, sophisticated intake forms were devised and used. Shortly after a patient's admission, a review board conducts a formal assessment of his background and needs. This assessment ranges from general educational and behavioral evaluations to diagnoses of specific problem areas. The assessment is performed by the Treatment Center's research department, which has a national reputation for excellence. Following completion of the assessment, a treatment protocol is implemented. Monthly progress reports are generated as to all patients and a "patient services information summary" is compiled, showing, inter alia, patient utilization of various therapy and other services within the institution as a whole.

Most often, treatment is administered in the form of group therapy. Nevertheless, at time of trial, 45 patients were receiving some kind of individualized treatment. Therapeutic modalities ranged from housing unit meetings (HUMs) to specialized programs focusing on particular deficiencies, e.g., substance abuse, anxiety management, impulse control.

Once the new complex was completed in 1986, management made a reasoned professional decision that many of the one-on-one programs, i.e., those involving individual audiences between patient and clinician, were ineffective. It was decided to transform most of these programs into group sessions. According to the district court, this alteration in format, together with the concomitant shift from a psychodynamic treatment model to a behavioral/rehabilitative model, caused "some upheaval." A significant number of unhappy clinical staffers departed between 1986 and 1989.

By the time of trial, most of the turmoil had subsided. Under the aegis of Dr. Jurgela, the new Clinical Director, and Mr. Cardoza, the new Director of Rehabilitation Services, the defendants were on the way toward establishing a comprehensive treatment program, emphasizing a cognitive behavioral approach. Each patient's specialized needs and treatment protocol were systematically outlined in an individualized treatment plan (ITP) issued by the RIRB. In one form or another, group therapy was being offered to every patient. HUMs were universally available and 213 patients were participating in them. These sessions allowed patients to ventilate and resolve their complaints against their fellow residents. Dr. Jurgela believed that the patients' recalcitrance skewed the statistics somewhat. He testified without contradiction that many of the patients were not desirous of receiving any treatment.⁷ 

Withal, the trial court found that the defendants are often unable to provide all of the care specified by a patient's ITP:

Unfortunately, the treatment program outlined in a patient's ITP often bears little relation to the treatment the patient can actually expect to receive. This failure occurs in part because treatments are sometimes recommended even though they are not offered at the Treatment Center, and [sometimes] because the limited clinical staff at the Treatment Center is overburdened and unable to provide all of the recommended treatment. In fact, the RIRB itself has been unable to fulfill completely its mandate. As many as forty-five patients have not had their ITP's reviewed during the past twelve months....

The defendants argued below that, notwithstanding this shortfall, significant progress had been made, particularly during the year preceding trial, toward implementing a comprehensive treatment system. Programs such as anger management, anxiety management, self control therapy, and relaxation therapy had been instituted.⁸  Despite the recent increase in program availability, however, the very programs which related most directly to patients' sexual dangerousness, e.g., impulse control and systematic desensitization, were in the shortest supply.⁹ 

The court observed that staff shortages posed the largest obstacle to full implementation of therapeutic services.¹⁰  Moreover, staff shortages seemed to be worsening; in 1987, the Treatment Center was authorized to employ a crew of 72 and fielded 53, whereas two years later, it had 78 authorized positions, only 40 of which were filled. The court found these shortages, by and large, to be beyond the defendants' reasonable control; the location of the facility, anxiety over its future, and the violent nature of the population served to dissuade many potential applicants from accepting positions there.

On balance, the court concluded that the range of treatments presently offered satisfied the requirements of the consent decrees, reasoning that the defendants were not obligated to provide every conceivable kind of treatment that might have a beneficial effect. This conclusion stands in bold relief against the backdrop of a record which contains no colorable evidence that the Treatment Center was failing to treat any patients who desired to participate in treatment programs.

5. The Authorized Absence Program. The Authorized Absence Program (AAP), conceived by Judge Garrity in 1978, allows patients to live in the outside community on a limited basis. AAP participants are accorded varying degrees of freedom and supervision based on their clinical needs and the administration's assessment of security risks. Soon after taking office in 1985, Tink began working to eliminate a huge backlog in the processing of AAP applications. He also secured additional AAP staff positions, at higher salaries, in order to improve the quantity and quality of available staffing. By early 1989, 49 patients were enrolled.

The district court found that the defendants devoted great effort to ensuring that the AAP ran efficiently and effectively; that Tink was personally committed to the program's success; and that satisfactory progress had been made in this area. The court explicitly rejected plaintiffs' contention that Tink was not qualified to make final decisions regarding program participation.

C. Nature of the Facility.

The psychiatric profile of the Treatment Center population can be plotted generally along a continuum.¹¹  At the poles are two relatively small inmate groups: under 15% of the population do not suffer from any known mental disorders whereas, at the opposite pole, perhaps 10% suffer from extreme psychoses (such as schizophrenia and bipolar disorders). Individuals in the large "middle" group are afflicted by an array of moderate mental disabilities, including problems associated with substance abuse, borderline personality disorders, and paraphilia.

Commitment to the Treatment Center is intended not only to provide an opportunity for treatment but also to protect society. Knight v. Mills, 836 F.2d 659, 667 n. 11 (1st Cir. 1987). Because every patient is a convicted sex offender, and a majority of them have some kind of mental deficiency, the defendants face legitimate security concerns. These concerns are exacerbated by the absence of statutory authority to deny commitments or to transfer intractable patients. Thus, the Treatment Center is forced to accommodate patients who are continually disruptive and who show no signs of rehabilitation. According to expert testimony, credited by the lower court, the lack of transfer authority, combined with the involuntary nature of the commitment process, distinguishes the Treatment Center from every other facility for sexually dangerous persons in the United States.

In light of the security risks that the patients pose to the staff and to each other, officials have implemented the movement control pass (MCP) procedure. This procedure enables the administration to keep track of patients' locations by requiring them to obtain authorization before entering or leaving certain sections of the facility. The tier system, under which patients are theoretically treated in the least restrictive environment, is also driven by security concerns: a patient is housed in a particular privilege unit according to the threat he is perceived as posing to others' safety. To illustrate, persons assigned to the pre-release house or a maximum privilege unit are thought to present the fewest security problems, and so forth. In other words, a patient's privilege status is inversely proportional to his status as a security risk.¹²  The evidence was quite clear that none of the criteria under which classification decisions were made countervailed, or were inappropriate in regard to, treatment purposes.

The trial judge found the MCP procedure and the tier system to be reasonable in light of the "violent and aggressive" nature of certain patients. He also found them in conformity with the consent decrees. He concluded that both policies gave acceptable weight to the professional judgment of the Treatment Center's administrators. In this respect, the judge relied specifically upon Tink's testimony that a security-based classification system materially advanced the therapeutic environment.¹³ 

Plaintiffs hawk a bevy of issues on appeal. In view of the district court's exhaustive opinion, we will discuss only a few points at any length.¹⁴  The rest have been fully considered and we reject them as either plainly meritless or conclusively rebutted in the opinion below. In some instances, both rubrics apply.

The issues which warrant further discussion include (1) the plaintiffs' claim that the district court erred by supposedly using a standard of constitutional adequacy to evaluate defendants' compliance with the consent decrees, and (2) the plaintiffs' two-pronged assault on the court's conclusion that defendants were not in contempt. We address these issues seriatim.

A. The Proper Measure of Compliance.

In deciding the matters before it, the court below wisely abjured the constitutional issues which plaintiffs labored to raise. It was unnecessary to consider constitutional standards, the court reasoned, because the existing consent decrees "require[d] the provision of adequate treatment for patients" at a level beyond that required by any applicable constitutional minima. Accord Twelve John Does v. District of Columbia, 855 F.2d 874, 878 n. 30 (D.C. Cir. 1988). Hence, a finding that the commonwealth was in substantial compliance with the decrees would perforce establish the absence of any constitutional abridgement and thus eliminate the need for a separate constitutional inquiry. We think that the trial judge's skirting of the constitutional thicket was appropriate. It has been long settled that if a court can resolve a dispute without confronting an unsettled constitutional issue, it should proceed in that fashion. See, e.g., Califano v. Yamasaki, 442 U.S. 682, 692, 99 S. Ct. 2545, 2553, 61 L. Ed. 2d 176 (1979); Hagans v. Lavine, 415 U.S. 528, 547, 94 S. Ct. 1372, 1384, 39 L. Ed. 2d 577 (1974); Ashwander v. TVA, 297 U.S. 288, 341, 346-48, 56 S. Ct. 466, 482-84, 80 L. Ed. 688 (1936) (Brandeis, J., concurring).

Acknowledging that the consent decrees set a higher standard than the Constitution, plaintiffs do not seriously dispute the correctness of the district court's announced methodology. Rather, they argue that the court, despite its avowed intention of using the decrees as a yardstick, did the opposite, in reality eschewing the more demanding criterion and measuring defendants' performance by reference solely to constitutional minima. Warming to this theme, the plaintiffs tell us that the court's application of a less rigorous standard can be seen in its articulation of the cardinal question as involving whether available treatment was "adequate in light of the populations' clinical needs." Because Judge Mazzone's framing of this inquiry focused on the needs of the patients, not the requirements of the consent decrees, plaintiffs perceive the court's very choice of words to be a telltale for fundamental error. See supra note 6. In the same vein, the plaintiffs pounce upon the district court's use of a statement from Youngberg v. Romeo, 457 U.S. 307, 102 S. Ct. 2452, 73 L. Ed. 2d 28 (1982), a case admittedly involving constitutional rights, in the course of arriving at its determination that defendants' provision of therapeutic programs was sufficient.¹⁵ 

Having read Judge Mazzone's opinion with care, we fail to see any transposition of the applicable standards. The defendants' performance was measured against the imperatives of the consent decrees. While the court at times employed language reminiscent of a constitutional analysis, the burden of its findings left no doubt that its primary concern was with the decrees' mandates and the commonwealth's compliance therewith. Indeed, the court's step-by-step delineation of the decree-required programs and its painstaking survey of defendants' efforts to implement them bears reliable witness to the court's unswerving focus on the decrees as its polestar. Plaintiffs' contrary argument represents at best a triumph of hope over reason.

B. The Supportability of the Court's Conclusions.

We come now to the heart of plaintiffs' main appeal: their asseveration that the decision to deny relief was supported by neither the evidence nor the lower court's findings.

1. Sufficiency of the Evidence. Plaintiffs' argument can be broken down into two components. First, they mount a somewhat desultory campaign to convince us that the trial court's findings were flawed. We are unpersuaded. When, as here, a district court sits as finder of the facts, its determinations are accorded great respect. Constrained by the Civil Rules,¹⁶  we review such findings of fact only for clear error. See Henry v. Connolly, 910 F.2d 1000, 1002 (1st Cir. 1990); Jackson v. Harvard University, 900 F.2d 464, 466 (1st Cir.), cert. denied, --- U.S. ----, 111 S. Ct. 137, 112 L. Ed. 2d 104 (1990); Anderson v. Beatrice Foods Co., 900 F.2d 388, 392 (1st Cir.), cert. denied, --- U.S. ----, 111 S. Ct. 233, 112 L. Ed. 2d 193 (1990). And in so doing, we are chary of permitting "[a]ppellate review of complex, fact-dominated issues ... to descend to the level of Monday-morning quarterbacking." Beatrice Foods, 900 F.2d at 392.

The clear-error test is a rigorous one. "It is not enough that [an appellate court] might give the facts another construction, resolve the ambiguities differently, and find a more sinister cast to actions which the district court apparently deemed innocent." United States v. National Assoc. of Real Estate Boards, 339 U.S. 485, 495, 70 S. Ct. 711, 717, 94 L. Ed. 1007 (1950); see also Keyes v. Secretary of the Navy, 853 F.2d 1016, 1020 (1st Cir. 1988). Before setting aside findings of fact or conclusions fairly inferable therefrom, the court of appeals must, "on the whole of the record, ... form a strong, unyielding belief that a mistake has been made." Cumpiano v. Banco Santander Puerto Rico, 902 F.2d 148, 152 (1st Cir. 1990); accord Reliance Steel Prods. Co. v. National Fire Ins. Co., 880 F.2d 575, 576 (1st Cir. 1989). It follows, then, that "[w]here there are two permissible views of the evidence, the factfinder's choice between them cannot be clearly erroneous." Anderson v. City of Bessemer City, 470 U.S. 564, 574, 105 S. Ct. 1504, 1511, 84 L. Ed. 2d 518 (1985).

To state the standard of review is, we think, to strip plaintiffs' assertions of any veneer of plausibility. The most hotly contested disputes between the parties stem from divergent interpretations of established facts. The litigants disagree, for example, as to matters such as the degree of treatment needed to effect successful rehabilitation; the extent of the disruption to patients' progress caused by the changeover from one therapeutic model to another; the cause of the staff exodus in 1986 and 1987; and whether the turnover-related "upheaval" at the Treatment Center was quelled (or merely papered over). Other notable differences of opinion concern whether the defendants were responsible for failing to hire additional staff between 1987 and 1989 or, conversely, whether factors beyond their control prevented them from filling open slots; and, second, whether the security measures implemented by the defendants were necessary. As a subset of the latter controversy, the parties bicker about the extent, if at all, to which the security measures impeded the patients' rehabilitation. The district court painstakingly addressed and resolved these disputes, by and large in appellees' favor. To be sure, many of these areas pose mixed fact/law questions--but the factfinder's answers to such mixed questions must likewise be reviewed under the clear-error test. See Roland M. v. Concord School Comm., 910 F.2d 983, 990 (1st Cir. 1990), cert. denied, --- U.S. ----, 111 S. Ct. 1122, 113 L. Ed. 2d 230 (1991); New England Anti-Vivisection Soc., Inc. v. United States Surgical Corp., 889 F.2d 1198, 1203 (1st Cir. 1989).

Given the court's thorough treatment of the facts, it would serve no useful purpose to repastinate well-ploughed soil. The plaintiffs have argued forcefully, and well, that their evidence contradicted the commonwealth's evidence at several junctures; and that, were plaintiffs' version to be credited, the court's findings could not stand. That is true--but altogether beside the point. We freely concede that the evidence below was conflicting on a wide variety of points, and susceptible to differing inferences. Yet in the Rule 52(a) context, conflicts of this sort are grist for the nisi prius mill. Having failed to limn any issue on which the lower court's factual findings lacked roots in the record, the plaintiffs' argument is really that the court, confronted with a series of equally supportable choices, made the wrong ones. That argument must succumb to the jurisprudence and method of Rule 52(a). Because the plaintiffs have given us no persuasive reason to think--let alone to form a "strong, unyielding belief," Cumpiano, 902 F.2d at 152--that the trial court's findings of fact were in error, we will not disturb those findings.

2. Supportability of the Conclusions. Appellants' next initiative is somewhat more nettlesome. They contend that the court's findings, even if swallowed whole, do not justify its ultimate conclusions. To the extent that this asseveration has a patina of plausibility, it calls into question whether the defendants' provision of programs surpassed the threshold required by the consent decrees.¹⁷ 

In the court's examination of the various treatment programs, it identified a number of instances in which the defendants were slow in implementing programs (e.g., educational and vocational workshops) or in which they had not yet fully developed existing programs (e.g., certain special needs courses). Pointing to the regulations promulgated pursuant to the consent decrees, and particularly to 104 C.M.R. Sec. 8.03(3) (requiring defendants to provide adequate treatment to "every patient"), plaintiffs contend that the defendants cannot be in compliance with the consent decrees, and must therefore be in contempt, so long as the defendants' performance falls short of perfect compliance. We do not believe, however, that perfection is exigible; progress, after all, should not be reduced to a mere numbers game.

We begin with three general rules. First, a complainant must prove civil contempt by clear and convincing evidence. See, e.g., AMF, Inc. v. Jewett, 711 F.2d 1096, 1100 (1st Cir. 1983). Second, substantial compliance can avert a finding of contempt. That is to say, the decrees here at issue, like most such decrees, were susceptible to satisfaction by diligent, good faith efforts, culminating in substantial compliance. See, e.g., Howard Johnson Co. v. Khimani, 892 F.2d 1512, 1516 (11th Cir. 1990) ("Conduct that evinces substantial, but not complete, compliance with the court order may be excused if it was made as part of a good faith effort at compliance."); Drywall Tapers, Local 1974 v. Local 530, 889 F.2d 389, 394 (2d Cir. 1989) ("A court may hold a party in civil contempt only if ... 'the defendant has not been reasonably diligent and energetic in attempting to accomplish what was ordered.' ") (citation omitted), cert. denied, --- U.S. ----, 110 S. Ct. 1478, 108 L. Ed. 2d 615 (1990); Balla v. Idaho State Bd. of Corrections, 869 F.2d 461, 466 (9th Cir. 1989) ("Substantial compliance with a court order is a defense to an action for civil contempt."); see also Board of Educ. v. Dowell, --- U.S. ----, 111 S. Ct. 630, 638, 112 L. Ed. 2d 715 (1991) (question of whether school board complied with desegregation order depends on whether board acted in good faith and whether vestiges of past discrimination had been eradicated to the extent practicable). Third, an appellate court reviews the denial of a contempt motion only for abuse of discretion, Massachusetts Ass'n of Older Americans v. Commissioner of Pub. Welfare, 803 F.2d 35, 38 (1st Cir. 1986), a standard which must be administered flexibly, with due regard for the circumstances. AMF, 711 F.2d at 1101.

To expand on the last of these tenets, it must be recognized that context is extremely important to appellate oversight in cases like this one. Where a contempt motion involves an interpretation of a consent decree between private parties, the court of appeals has comparatively greater freedom to review the trial court's decision (which is akin to a contractual interpretation, and is thus largely a conclusion of law). See, e.g., id. at 1100; cf. Fashion House, Inc. v. K Mart Corp., 892 F.2d 1076, 1083 (1st Cir. 1989) (discussing standard of review in contract cases). In the context of public law litigation, however, the guidelines are much different. In such cases, the district court's construction of a consent decree should be accorded considerable deference, because broad leeway is often necessary to secure complicated, sometimes conflicting, policy objectives. See United States v. Commonwealth of Massachusetts, 890 F.2d 507, 509-10 (1st Cir. 1989); Massachusetts Ass'n of Older Americans, 803 F.2d at 38; AMF, 711 F.2d at 1101.

This double standard derives from the realities of human experience. Different types of consent decrees are ordinarily conceived and hatched in markedly different ways. In a commercial setting, a consent decree is treated like a contract because the court assumes that private parties understand the economic realities and business consequences of their agreements. See Commonwealth of Massachusetts, 890 F.2d at 509. The configuration bears a family resemblance to the conventional case configuration on the civil side, cf., e.g., AMF, 711 F.2d at 1101 ("A civil finding of contempt in the private context essentially triggers a payment of money damages as in garden variety tort or contract litigation."), so the court acts essentially in its accustomed role as an umpire/adjudicator/assurer of procedural fairness.

In public law litigation, courts typically play a proactive role--a role which can have nearly endless permutations. See, e.g., Massachusetts Ass'n of Older Americans, 803 F.2d at 38. Frequently, the trial court's adjudicative function blends with its service as an instrument for change. The relief requested often involves the restructuring of a state or city program, requiring the court to fashion equitable remedies--sometimes unique and often complicated--in order to secure "complex legal goals." AMF, 711 F.2d at 1101. We agree with Professor Chayes that, in the public law context, the consent decree "provides for a complex, on-going regime of performance rather than a simple, one-shot, one-way transfer.... It prolongs and deepens, rather than terminates, the court's involvement with the dispute." Chayes, The Role of the Judge in Public Law Litigation, 89 Harv. L. Rev. 1281, 1298 (1976). What is more, the due administration of such a decree calls for a certain deftness of touch. After all, "appropriate consideration must be given to principles of federalism in determining the availability and scope of [federal] equitable relief." Rizzo v. Goode, 423 U.S. 362, 379, 96 S. Ct. 598, 608, 46 L. Ed. 2d 561 (1976).

It is with an awareness of the three general rules mentioned above, and with this special understanding of the trial judge's pivotal role in the conduct of public law litigation, that we examine the "no contempt" ruling in this case. If a court were to employ tunnel vision and focus solely on the Treatment Center's existing (and admitted) shortcomings, in isolation, the argument for reversal might have a certain allure. But, a wider view of the litigation is required. At least since the time of Brown v. Board of Education, 349 U.S. 294, 75 S. Ct. 753, 99 L. Ed. 1083 (1955), district courts have exercised broad powers and enjoyed great latitude in regulating the operations of state institutions, ranging from school districts, to hospitals, to prisons, as may be necessary to enforce federally assured rights. See The Supreme Court--Leading Cases, 104 Harv. L. Rev. 129, 296-97 (1990) (citing cases); see also Missouri v. Jenkins, --- U.S. ----, 110 S. Ct. 1651, 1666, 109 L. Ed. 2d 31 (1990) (upholding district court order compelling local government unit to levy taxes in excess of limits set by state statute in order to finance school desegregation plan). And although federal district courts have been allowed--indeed, at times constitutionally required--to intrude in the affairs of state and local governments, such intervention has been conducted with the clear understanding that the autonomy of these governmental entities should be safeguarded to the maximum extent possible. As the Court stated in Jenkins, "one of the most important considerations governing the exercise of [federal] equitable power is a proper respect for the integrity and function of local government institutions." Id. 110 S. Ct. at 1663; see also Rosario-Torres v. Hernandez-Colon, 889 F.2d 314, 323 (1st Cir. 1989) (en banc).

The Court demonstrated this very concern in Spallone v. United States, 493 U.S. 265, 110 S. Ct. 625, 107 L. Ed. 2d 644 (1990), holding that a district court had over-extended its equitable powers. There, the lower court, without first levying fines against the municipality alone, imposed contempt sanctions on individual city council members who had failed to comply with a consent decree between the city and the United States. Id. 110 S. Ct. at 634. In setting aside the sanctions, the Spallone Court emphasized that district judges should exercise the "least possible power" to achieve the remedial end proposed. Id. at 635 (quoting Shillitani v. United States, 384 U.S. 364, 371, 86 S. Ct. 1531, 1536, 16 L. Ed. 2d 622 (1966)).

In the case at bar, this panoply of principles is much in evidence. The district court, albeit in the person of a series of different judges, has been the central figure in monitoring the extent and adequacy of services provided at the Treatment Center over the better part of the last two decades. The court has been assiduous in protecting patients' rights. It has entertained numerous enforcement actions brought under the Williams consent decrees. The result has been a tapestry of marked progress, woven more through careful attention to the overall design of the consent decrees than through an unbending insistence on absolutes. The historical record before us is dotted with examples of judicial restraint cast in precisely the mold configured by the Spallone Court: the achievement of legally required goals by methods which, while effective, are among the least onerous and intrusive methods available.

In the present litigation, the lower court appears to have trod the well-beaten path. It undertook a detailed review of compendious evidence and decided that, in light of the defendants' notable progress, there was substantial compliance with the overall mandate of the consent decrees, and hence, no contempt. The district court effectively credited the defendants with exercising their best efforts to comply with the decrees. Considering the district court's prolonged institutional involvement with the litigation and its demonstrated willingness to impose significant demands on the commonwealth when necessary--ordering, for example, that a new Treatment Center be built--we believe that its decision not to hold the state defendants in contempt is worthy of considerable deference. Cf. Commonwealth of Massachusetts, 890 F.2d at 510 (review of district court's ultimate finding on contempt should be conducted with deference to the court's "intimate understanding of the history and circumstances of th[e] litigation").

Viewing matters through this deferential prism leaves us no principled option but to affirm the decision below. Abuse of discretion can most easily be found when the weight and pigmentation of the evidence places a matter at one end of the spectrum or another. This is not such a case. It is quite plain from both the record as a whole and the district court's findings that there was neither letter-perfect compliance with, nor palpable disregard for, the demands of the consent decrees. In circumstances where black and white were both submerged in scumbled shades of gray, the fact that defendants had not completely implemented the programs specified in the decrees could rationally support two opposite, though equally plausible, inferences: either defendants had not taken reasonable steps to comply, or the nature of the institution, coupled with staffing shortages beyond the defendants' control, had prevented defendants from attaining full compliance notwithstanding good faith efforts. Especially given the tangled nature of the issues, deciding which of these inferences to draw was a judgment call best left to the trier.

Cognizant of the heavy burden which a complainant must carry in a contempt proceeding, we conclude that the district court's subsidiary findings were adequate to sustain the ultimate ruling on contempt.¹⁸  8] Compare, e.g., Commonwealth of Massachusetts, 890 F.2d at 510-11 ("[W]hile the Commonwealth has not complied with the decree's staffing ratios, the district court was within its discretion in determining that such non-compliance does not constitute contempt."). We have in mind, particularly, the court's findings (1) that the defendants have established, or are making substantial progress toward establishing, those programs memorialized in the consent decrees, (2) that the defendants have acted, and are acting, in good faith, (3) that the defendants have made great "efforts to achieve better and more enlightened institutional management practices," and (4) that staffing shortages and security problems beyond defendants' control have been, and are, the most persistent obstacles to complete implementation of treatment and programs. Because each of these findings is eminently supportable, we decline the invitation to second-guess the trier's informed choice between competing inferences in this complex situation.

We think that such a conclusion is reinforced by the Supreme Court's recent decision in Dowell. There, the Justices held that court-imposed desegregation orders may be lifted when a school board has complied in good faith with the decrees and "the vestiges of past discrimination ha[ve] been eliminated to the extent practicable." 111 S. Ct. at 638. Despite significant differences between this case and Dowell,¹⁹  the underlying principle is equally applicable here: officials operating under a public law decree are required to employ good faith efforts to satisfy its demands, and fault should not be found if they have implemented its dictates to the extent practicable. Id.

To sum up, the court below acted within its proper domain in preferring, at this juncture, the velvet glove to the mailed hand. There was no clear and convincing proof that defendants were in contempt of the consent decrees. On this record, it appears entirely plausible that defendants' shortcomings reflected the parlous circumstances under which they were compelled to operate rather than reproachable inadequacies in their efforts. Indeed, the continued and continuing development of the Treatment Center's programs strongly suggests that defendants' labors are yielding as fruitful a harvest of progress as conditions reasonably allow.

The merits behind us, we proceed, finally, to the fee award.

A. The Court's Findings.

After the district court's decision on the merits was announced, the plaintiffs moved for an award of attorneys' fees and costs in the amount of $533,021. The defendants objected to any award, contending that the plaintiffs had not prevailed. The district court found that (1) the improvements in the physical plant and treatment programs "were not due to [plaintiffs'] counsel's efforts in any measurable sense"; (2) counsel's actions did not lead to the issuance of any interim order, injunctive relief, or specific programmatic change; and (3) there was no clear causal link between defendants' progress and plaintiffs' remorseless prosecution of the litigation. The court summed up the situation by observing that, although plaintiffs' lawyers had expended much time and effort, they had concentrated on the trees, thereby losing sight of the forest:[C]ounsel, in my judgment, did not pursue the original objective of the pro se plaintiffs to "interpret, clarify and enforce" the decrees, but pursued a course of long, expensive, contentious litigation which, again, in my judgment, prolonged the dispute, heightened the tension and ignored the basic issue, namely, the proper role for the Treatment Center in the correctional system of Massachusetts in light of what I termed the inherently inconsistent law designed to treat and punish the sexually dangerous person.

Despite these findings, the court concluded that some unspecified benefit was achieved and that "[c]ounsel's ... presence throughout the period, and at trial, in presenting the issues, should be acknowledged." The judge also wrote that "[c]ounsel's efforts ... contributed to a resolution of the double-bunking issue...."²⁰  In a somewhat different vein, the court noted that defendants had been dilatory at times in responding to discovery requests. In light of this recalcitrance, "an award of fees is indicated because the delay and uneven course of this litigation is attributable in part to defendants' counsel." The court, acting under 42 U.S.C. § 1988, proceeded to award fees of $35,400 and costs of $10,000. The ruling pleased no one: plaintiffs argue that the award, comprising less than 10% of the amount requested, was niggardly; defendants argue that no fees were due.

B. Analysis.

The governing statute is the Fees Act, 42 U.S.C. § 1988, which provides in pertinent part that, in actions brought under the aegis of 42 U.S.C. § 1983, "the court, in its discretion, may allow the prevailing party ... a reasonable attorney's fee as part of the costs." In Guglietti v. Secretary of HHS, 900 F.2d 397 (1st Cir. 1990), we mapped two routes to prevailing party status:²¹ 

[I]n order to constitute a litigant as "prevailing," ... the party either must have enjoyed some bottom-line litigatory success or her suit must have had a catalytic effect in bringing about a desired result.

Id. at 398; accord Exeter-West Greenwich Regional School Dist. v. Pontarelli, 788 F.2d 47, 50 (1st Cir. 1986); Nadeau v. Helgemoe, 581 F.2d 275, 278-79 (1st Cir. 1978).

The first of these routes permits a court to award fees to plaintiffs only if they succeed on a "significant issue in litigation which achieves some of the benefit the parties sought in bringing suit." Nadeau, 581 F.2d at 278-79; see also Texas State Teachers Ass'n v. Garland Indep. School Dist., 489 U.S. 782, 789, 109 S. Ct. 1486, 1491-92, 103 L. Ed. 2d 866 (1989); Guglietti, 900 F.2d at 400; Exeter-West Greenwich, 788 F.2d at 50. In effect, the merits test "states the obvious, namely, that a party has prevailed if it wins the litigation." Coalition for Basic Human Needs v. King, 691 F.2d 597, 599 (1st Cir. 1982). Because this standard requires a plaintiff to succeed on at least one substantial claim, it is inapplicable here. The plaintiffs did not win on any significant issue in the current litigation and no judgment was entered in their favor.²² 

We turn, then, to the second route to prevailing party status. The catalyst test "applies to plaintiffs who have succeeded in achieving favorable results because of the filing of their Sec. 1983 claim, but have not had a final judgment on the merits entered in their favor." Exeter-West Greenwich, 788 F.2d at 52. It is invoked in those cases in which plaintiffs do not receive a favorable judgment, yet claim to have succeeded in bringing about a beneficial change in defendants' conduct or in the conditions complained of--a change which would not have occurred but for the institution of the suit. The critical inquiry is whether the suit "prompt[ed] defendants to take action to meet plaintiff's claim...." Nadeau, 581 F.2d at 279. In this inquiry, "the litigation's 'provocative role' in the calculus of relief is a sine qua non...." Guglietti, 900 F.2d at 401 (quoting Nadeau, 581 F.2d at 280). If the defendants acted other than in response to the spur of plaintiffs' suit, the catalyst theory does not apply. See id.; see also Nadeau, 581 F.2d at 280-81.

The catalyst test, like the merits test, involves not only a causality requirement but a materiality requirement. "The touchstone of the prevailing party inquiry must be the material alteration of the legal relationship of the parties...." Texas Teachers, 489 U.S. at 792-93, 109 S. Ct. at 1493. Thus, in Texas Teachers, the Court noted that, had the plaintiffs only succeeded on an issue of "minor significance," they would not be "prevailing parties" within the meaning of section 1988. Id. at 792, 109 S. Ct. at 1493. Excepting for the time being the double-bunking issue, plaintiffs' application for attorneys' fees founders on both the causality and materiality components of the catalyst test.

As to causation, the district court squarely found that plaintiffs did not "show[ ] a clear causal link between changes made by the defendants and this lawsuit." To be sure, plaintiffs argue that this finding was erroneous, contending that the chronological sequence of the various improvements, when viewed in conjunction with the progress of the litigation, proves that defendants were responding to pressures imposed by the lawsuit. Chronology, however, is neither the be-all nor the end-all. See Nadeau, 581 at F.2d 281 (chronological sequence, though important, is "clearly not definitive ... in determining whether or not defendant ... guided his actions in response to plaintiff's lawsuit"). Put another way, "the mere existence of a temporal coincidence ... cannot alone suffice" to engage the gears of the catalyst test. Martinez v. Rhode Island Housing and Mortgage Finance Corp., 628 F. Supp. 996, 1001 (D.R.I. 1986). The trial court was in the best position to sort out coincidence from effect, and hence, appellate appraisal of its findings must be deferential. See Guglietti, 900 F.2d at 399; McDonald v. Secretary of HHS, 884 F.2d 1468, 1474 (1st Cir. 1989); cf. Pierce v. Underwood, 487 U.S. 552, 557-63, 108 S. Ct. 2541, 2546-49, 101 L. Ed. 2d 490 (1988) (abuse of discretion standard used in reviewing district court's determination, in connection with EAJA fee-shifting inquiry, of whether government's position was "substantially justified"). As to materiality, the district court determined that whatever changes the litigation prompted were not substantial; and this assessment, too, seems sustainable.

Given the standard of review, we are unable to disturb the lower court's findings that neither causation nor materiality was proven. And based on those findings, we have little difficulty in concluding that the court did not err in refusing to grant the bulk of plaintiffs' fee request. Indeed, once the court's succinctly phrased findings are accepted, as they must be, any broadscale characterization of the plaintiffs as "prevailing parties" would threaten the very integrity of that term. Not only did plaintiffs lose on all their claims, but the positive changes that occurred at the Treatment Center, as the district court itself acknowledged, were not triggered by the suit "in any measurable sense."

There is, however, one aspect of the court's fee award which is less than pellucidly clear. On the one hand, Judge Mazzone stated: "I conclude that there have been some minor benefit [sic] achieved, and some minor changes in the parties' relationships that have been brought about as a result of the efforts of counsel...." He also stated that, in his view, "[t]he Bruder plaintiffs spawned no interim order ... and were rebuffed on virtually every issue by Judge Young. On an issue which survived, namely, the double-bunking of John Bruder, that achievement was shortly negated by Judge McNaught." These comments seem to denigrate the importance of any success the plaintiffs might have attained on double-bunking. And if, indeed, the double-bunking point was of "minor significance" given the scope and tenor of the litigation as a whole, legal services rendered in respect to it would not be compensable. See Texas Teachers, 489 U.S. at 792, 109 S. Ct. at 1493. Yet, notwithstanding this language, the judge appears to have predicated the modest fee award at least in part on plaintiffs' contribution to resolving the double-bunking issue and to have identified no other cognizable issue on which the plaintiffs could be said to have prevailed.

Because of Judge Mazzone's limited, and perhaps conflicting, characterization of plaintiffs' role in setting double-bunking to rest, we think that prudence suggests the desirability of clarification. For one thing, the record is insufficiently developed on this point, see supra note 20, to allow us adequately to evaluate the propriety of the fee award vis-a-vis double-bunking. For another thing, the lower court expressed no findings regarding the methods it employed, or the calculations it made, in arriving at the fee it set, thereby frustrating review. See, e.g., Peckham v. Continental Casualty Ins. Co., 895 F.2d 830, 842 (1st Cir. 1990) ("Appellate review of fee awards ordinarily requires that concrete findings be made and that the court below supply a clear explanation of the reasons undergirding a particular fee award."). These factors counsel strongly in favor of a remand for further findings. See Metrocorps v. Eastern Mass. Jr. Drum & Bugle Corps, 912 F.2d 1, 3 (1st Cir. 1990); Morgan v. Mass. General Hosp., 901 F.2d 186, 195 (1st Cir. 1990); Brown v. Bullard Independent School Dist., 640 F.2d 651, 654 (5th Cir.), cert. denied, 454 U.S. 828, 102 S. Ct. 120, 70 L. Ed. 2d 103 (1981).

There is also another consideration militating in favor of a remand. It appears from what Judge Mazzone wrote that he might very well have made the $45,400 award principally to sanction defendants' counsel for their "recalcitrance and unwillingness ... to move the matter to an expeditious resolution" and to recompense plaintiffs' counsel for the corresponding expense. Although such an essentially punitive purpose has no place in a prevailing party analysis under section 1988, the aim of which is to reward successful plaintiffs for their efforts, not to punish victorious defendants for some unspecified degree of recalcitrance, it is beyond cavil that a district court, acting sua sponte under a variety of other rules and statutes, e.g., Fed. R. Civ. P. 11, 16(f), 37(b); 28 U.S.C. § 1927, may require a party to assume an opponent's fees and costs to punish prolonged foot-dragging. Because the record is murky on what the judge actually intended, and an award under some rule or statute other than section 1988 might be permissible in the circumstances described by the lower court, remand would allow an opportunity for needed clarification of this point.

For these reasons, we return the case to the district court with instructions that more specific findings be entered in connection with plaintiffs' entitlement vel non to fees on the double-bunking issue. If the court should conclude that plaintiffs' achievement with respect thereto was of only "minor significance," Texas Teachers, 489 U.S. at 792, 109 S. Ct. at 1493, or did not have "a catalytic effect," Guglietti, 900 F.2d at 399, in bringing about abandonment of the practice, no fees may be awarded under 42 U.S.C. § 1988. If, however, the court concludes that plaintiffs are deserving of fees with respect to that issue, then the court should make specific findings as to the number of hours reasonably spent and the hourly rate to be employed. Moreover, wholly apart from the matter of double-bunking, the district judge may, if he believes such a course to be indicated, convene a hearing to consider whether monetary sanctions should be levied under the aegis of an applicable rule or statute, and if so, the extent of any monetary award. Of course, specific findings will also be required if such sanctions are imposed. See generally Figueroa-Ruiz v. Alegria, 905 F.2d 545, 548-49 (1st Cir. 1990) (discussing desirability of findings in Rule 11 setting).

We need go no further. After close perscrutation of a voluminous record spanning nearly two decades, we are satisfied that the district court's thoughtful resolution of the merits must be upheld. The fee award, however, must be vacated and the cause remanded for further proceedings as discussed above.²³ 

The district court's refusal to hold the defendants in contempt and its denial of equitable relief are affirmed. The award of fees and costs under 42 U.S.C. § 1988 is vacated. The case is remanded to the district court for further proceedings not inconsistent herewith. No costs on appeal.

It is so ordered.

APPENDIX A

PARTIAL CONSENT DECREE

Pending a hearing on other issues in this action, the parties have agreed to enter into a partial consent decree in said matter, the provisions of which are as follows:

1. The Treatment Center at MCI Bridgewater shall be treated as a facility of the Department of Mental Health.

2. Primary responsibility and authority for the Treatment Center shall be exercised by the Department of Mental Health.

3. All personnel at the Treatment Center (clinical, custodial, administrative) shall be subject to the control of the Commissioner of Mental Health with respect to the handling of patients.

4. Custodial personnel, but not patients, shall be under the administrative, operational and disciplinary control of the Commissioner of Correction.

5. The Department of Mental health shall exercise the responsibility and authority set forth in subparagraph 2 above so that patients at the Treatment Center should have the least restrictive conditions necessary to achieve the purposes of commitment. To implement this and other provisions of this decree, the Department of Mental Health shall forthwith take steps jointly and in cooperation with the Department of Correction including, as appropriate, seeking legislative funding, to do the following:

a. Improve the physical plant at the Treatment Center....

b. Implement a meaningful work program and increasing avocational activities which may, for those patients so preferring, be a substitute for the work program.

c. Have a system of differing security for different categories of patients and modifying the plant if necessary to permit less restrictive conditions for those patients not requiring maximum security.

....

6. The Department of Mental Health and Department of Correction will also act in concert to draw up and promulgate certain rules and regulations to apply to patients at the Treatment Center. These rules and regulations shall include but shall not be limited to the following: (a) implementation of the tier system of security; (b) application of the rules of the Department of Mental Health governing the civil rights of patients pursuant to M.G.L. c. 123 to the patients at the Treatment Center insofar as consistent with M.G.L. 123A....

APPENDIX B

PARTIAL CONSENT DECREE

The Court orders, adjudges, and decrees that:

....

2. On March 15, 1975, the defendant Commissioner of Correction shall file ... copies of a plan describing in detail the contemplated disposition, on June 30, 1975, of the then existing Treatment Center patients, staff, equipment, physical plant, budget, and other resources. The plan shall include a statement of the precise disposition or contemplated alternative dispositions of each person who, on March 15, 1975, is a patient of the Treatment Center. The plan also shall include a detailed description of the therapeutic, educational, vocational, and avocational programs to be provided by the Department of Mental Health to patients of the Treatment Center, and a provision for the day or other short-term release of Treatment Center patients for approved programs outside the Treatment Center where such release is deemed appropriate by the Department of Mental Health....

3. The terms of this decree do hereby supersede and cancel the terms of section 5.a. of the Partial Consent Decree previously entered in this case on June 14, 1974, but do not supersede or cancel any other terms of said Partial Consent Decree.

4. The Court retains jurisdiction in this case.