These are appeals from an attorney's fee award covering work performed by counsel for the plaintiff class, residents of the Northampton State Hospital, during two and one half years subsequent to entry of a complex consent decree. The decree bound the responsible Massachusetts officials to establish a system for the care and treatment of mentally disabled persons in community residential facilities and nonresidential programs.1

The litigation commenced in 1976; the decree was entered in 1978; and the district court made its first fee award in 1982, covering work performed from 1976 through 1981, in the amount of $386,204.2 The decision, resulting in a two-thirds reduction of the amount claimed, included an extensive analysis in which the court discussed and distinguished four kinds of work done by plaintiffs' counsel in implementing and monitoring the decree. The court established different rates for court work, decree implementation and monitoring, general work, and travel. Id. at 1076. Although the decree called for a court-appointed Monitor, the decree also bestowed various responsibilities on the parties. Id. at 1072.

The present fee application covered the period from January 1, 1982 through June 30, 1984, and included 2537.93 hours of services performed by two lawyers and a paralegal. The 191-page application included some 3500 entries (at roughly eighteen items a page) accounting for time in tenths of an hour, and coded to identify six different kinds of work.3 The entry described the kind of activity (e.g., "spoke w/", "met w/", "drafted letter", "reviewed letter", "deposition", "hearing", etc.), as well as the other person or group involved, and, usually, the subject matter (e.g., "on budget", "on crisis intervention program", "on deposition"). The amount claimed was $239,772.10.

The district court awarded $132,639.55, a reduction of forty-four percent. The court decided three issues in favor of the plaintiffs: (1) it rejected defendants' contention that, to recover post-judgment fees, plaintiffs must show that their efforts produced a better result than otherwise would have occurred, holding that "reasonable monitoring", under Garrity v. Sununu, 752 F.2d 727, 738 (1st Cir.1984) imposes a lesser burden; (2) it rejected defendants' challenge to time spent on three motions to hold defendants in contempt, all settled before hearing, finding that the efforts helped produce favorable results; and (3) it rejected defendants' claim that the fee award should be reduced to the extent that hours included in the fee application were also compensated by the Association of Service Providers for Persons with Handicaps ("Association"), accepting counsel's representation that all time claimed was for service benefiting the plaintiff class.

The court decided five issues in whole or in part against the plaintiffs: (1) it disallowed time spent in the unsuccessful defense on appeal of a legal services program ordered by the district court; (2) it disallowed time spent in a candidate search for the position of court Monitor; (3) it cut in half time claimed in connection with a Supplemental Agreement that never came to fruition; (4) it reduced, among other items, the claimed hourly rates from $125 and $115 for court activities to $95, from $115 and $105 for decree implementation to $85, and from $105 and $95 for general work to $75; and (5) it refused to grant plaintiffs' request for a fifteen percent upward adjustment of the lodestar, while also denying defendants' request for a downward adjustment.

Defendants appeal the court's refusal to adopt a rigorous "but for plaintiffs' efforts" standard for post-judgment monitoring fees and its refusal to reduce the award to reflect payments received from the Association. Plaintiffs cross appeal from the court's elimination of time spent in helping select a Monitor, from its reduction of time spent on the Supplemental Agreement, and from its reduction of hourly rates.

Defendants' position is that in a post-judgment context where a defendant-funded court monitor is created by the decree, the only way to avoid creating a state-funded, open-ended "sinecure for counsel" is "to compensate counsel only where a substantial issue arises as to the defendants' obligations under the decree, and the work of plaintiffs' counsel yields a resolution more favorable to the class than the defendants were prepared to concede." They argue that such an extra burden is required to sift out progress or benefits flowing from the decree itself. They seek to distinguish Garrity v. Sununu, supra, on the ground that the instant case has in place a court Monitor, paid some $70,000 during the period at issue, whose presence presumably should have made unnecessary all time spent in routine monitoring. They argue alternatively that Garrity may need to be limited in light of Webb v. Board of Education of Dyer County, --- U.S. ----, 105 S. Ct. 1923, 85 L. Ed. 2d 233 (1985), which held that there should be no 42 U.S.C. Sec. 1988 fee award for work in optional administrative proceedings unless it is "of a type ordinarily necessary" to advance the litigation. Id. 105 S.Ct. at 1929.

We appreciate the fact that devising workable ways, fair to performer and payor, to compensate legal services during the formative period (following issuance of a complex system-creating decree and before satisfactory implementation becomes largely routine) is a difficult and sensitive task. The services are of lower profile and often of a more routinized nature than services preceding judgments. Missing the refining fire of the basic litigation, plaintiffs' attorney may slip into a mode of spending too much time on too many matters, with the result that the decree institutionalizes the attorney, as well as the system.

Defendants' proferred standard, however, leaves us traumatized at the prospect of a multitude of trials--not necessarily mini-trials--on whether issue A, for which X hours are claimed, was really a "substantial" issue regarding a defendant's obligations, and, if so, did plaintiffs' work produce a more favorable resolution than defendants "were prepared to concede". Must the result be measurably more favorable? How may one prove what defendants were prepared to concede, but did not? Such a standard implies the availability of appellate review of each issue, an addition to our domain that we would welcome with something less than unbridled enthusiasm. Moreover, as Amicus points out, defendants' proposed standard would stimulate posturing and undercut the amicable cooperation that a consent decree is designed to foster; plaintiffs would opt for a combative, litigious route in preference to quiet negotiation. Whether we view the likely results in terms of delay, cost to the parties, inflated counsel fees, acrimony, or the additional burden on both the district court and the court of appeals, we see little to recommend the suggested innovation.

Perhaps the most salient approach is to see if the normal method of determining fees for monitoring has broken down. We cannot say that it has in this case, because, while it was begun, it was never completed. Plaintiffs supplied their compendious fee application, which, while often not facially self-explanatory, contained the necessary keys to testing its reasonableness. Defendants generally knew the dates, subject matter, and people involved. They, at some expense to be sure, could have mounted challenges to specific claims--if not on a comprehensive basis, at least on a random one. Had this happened, the district court would have had the benefit of the adversary process and could have developed a sense of the extent to which the claim for services was reasonable. We recognize that a specific challenge to every item in a 3500-item catalogue of time charges would be impracticable, but it is not too much to expect the Commonwealth, relying on its deep involvement in the litigation, to target significant and vulnerable areas for testing. We would have confidence that, given reasonable assistance by counsel, a court could arrive at a fair decision without a dismaying investment of time, particularly during the later, "tapering off" stages of implementing a decree.

This traditional approach not having been attempted, we are left with the general attack on the standard, together with the "Association issue", infra. Without specific indicia of unreliability, we are left with more general indicators, which do not support the Commonwealth. We note first the consistency of the district court's fee decision-making in this case with the approach taken in its earlier decision on the 1976-1981 applications. Second, we are mindful of the extensive forty-four percent reduction made in this case. Third, we do not think it excessive for attorney Schwartz to have spent about one-fourth of his time for two and one half years on this case (i.e., 1323 hours out of 5000), attorney Fleischner one eighth (i.e., 602 hours out of 5000), and paralegal Costanzo one eighth (i.e., 613 hours out of 5000). Fourth, we observe that the Commonwealth, when it perceives that matters have been sufficiently clarified and stabilized, may and indeed should petition the court to relieve it of the burden of paying for private party monitoring. Fifth, we note that plaintiffs will have waited from two to four years for their attorney's fees.

Finally, we are impressed by the fact that defendants have not been able to muster any authority for their proposition. Against the full array of authority for allowing fees for reasonable post-judgment monitoring,4 the defendants can say only that these decisions are "a series of ad hoc assessments of factual circumstances quite unlike those presented here." As for their invocation of Webb v. Board of Education of Dyer County, --- U.S. ----, 105 S. Ct. 1923, 85 L. Ed. 2d 233 (1985), as Amicus points out, its requirement that services (in an optional administrative proceeding) be "useful" and "ordinarily necessary" is fully consonant with our insistence in Garrity that services be "necessary for reasonable monitoring of a consent decree". 752 F.2d at 738. And finally, defendants' effort to justify departure from the Garrity standard where a court-appointed monitor is in place seems to us to have been adequately answered by the district court in its first fee opinion.5

We therefore hold that the district court properly rejected defendants' plea to accept a different standard of proof for post-judgment monitoring.

Plaintiffs' fee application was filed on August 16, 1984. Oral argument was had on October 19, 1984. Shortly thereafter, a routine audit revealed that the Association, a group of state funded contractors providing services to plaintiffs under the consent decree, had entered into a retainer agreement with a charitable corporation staffed by plaintiffs' counsel, the Center for Public Representation ("Center"). Under this agreement, the Center was obligated to provide fourteen hours of service a week to the Association, in return for which the Center received from each Association member one-third of one percent of its contract funding received from the Department of Mental Health. Between 1982 and 1984, the Association paid the Center $87,468.

Defendants moved to reopen the evidence and sought additional discovery on December 3, 1984 on the issues whether all of the time claimed was in reality for work done for plaintiffs and whether plaintiffs had already been compensated by the Association. Plaintiffs immediately moved to supplement their application and also opposed the motion to reopen, stating that the information as to compensation from the Association was irrelevant and that, in any event, they had now provided full information. Defendants countered with a proposal that the parties try to stipulate facts concerning the Center's representation of the Association, failing which they would seek discovery.

On December 13, 1984, the court allowed the defendants' motion to reopen for purposes of receiving written briefs--within thirty days from defendants and sixty days from plaintiffs. It further stated that evidence on alleged conflicts of interest concerning payments of attorney's fees "shall be brought forward to the Court Monitor and he shall conduct hearings on same and make recommendations to the Court." On December 21, noting that the court had allowed the motion to reopen, defendants sought leave to file interrogatories requesting a list of all persons named in the fee application who were employees of a provider agency, and information concerning the make-up of the "Litigation Committee" mentioned in the retainer agreement, including whether it was the "Litigation Committee" referred to in various entries in the fee application. They also sought to telescope the time for responses so that they could incorporate them in their written brief. Plaintiffs opposed these requests, stating that the court had limited any reopening to written briefs. On January 4, 1985, the court denied leave to propound interrogatories, writing: "The Court did not intend that discovery should be reopened." We have traced this bit of procedural history in detail because it plays a large role in shaping our decision.

Defendants' position is that case authorities such as our Palmigiano v. Garrahy, 616 F.2d 598 (1st Cir.1980), holding that the presence of financial support for a public interest legal services organization is irrelevant to the calculation of a reasonable fee under 42 U.S.C. Sec. 1988, do not apply to the instant case where the Association's retainer payments to the Center were for legal services rendered in the Association's behalf and where the Association's Litigation Committee has the power to determine the scope and type of work and legal positions to be taken by the Center. Defendants further point to specific portions of the record, which indicate to them that certain services were performed solely or dominantly for the Association (for example, meetings with the Litigation Committee and time recorded for conferring with employees of various providers). They also claim that their inability to probe further leaves an "irresolvable ambiguity" as to the object of counsels' work, which justifies a fee reduction of between fifteen and twenty-five percent (i.e., between $20,000 and $33,000).

Plaintiffs, on the other hand, claim that the principles of cases like Palmigiano govern, that the existence of additional sources of funding is irrelevant to the disposition of their fee application, that the overarching purpose of the Association, necessarily controlling its Litigation Committee, is the welfare of the plaintiff class, and that plaintiffs' attorneys have often taken positions contrary to the interests of individual providers and have always accorded the plaintiff class their "exclusive and unconditional loyalty". Most pertinently, they assert that their fee application includes no time spent for the Association on unrelated matters but only hours spent for the Association (50% of their total work for the Association) where the interests of the plaintiff class and the Association were identical.

The district court, noting that the key question was whether the time claimed in the fee application was expended for the benefit of the plaintiff class, proceeded to rely "on the demonstrated integrity of plaintiffs' counsel ... who have shown themselves to be extremely conscientious and honest attorneys". The court stated that "[w]hen they represent to the Court that 50% of their Association hours were also for the benefit of the plaintiff class, quite frankly, the Court believes them. Defendants' position that plaintiff's hours must be reduced for this reason is, therefore, rejected."

Such a credibility judgment and exercise of discretion would normally be unexceptionable, particularly in attorney's fee matters. But, as we earlier have signalled, the procedural history of this issue places the ruling in a different light. We have criticized defendants for not making discrete challenges to specific kinds of services performed or the time spent in performing them. But the issue of dual compensation and possibly divided loyalty did not arise until after the October 1984 hearing. Although plaintiffs' attorneys cannot be criticized for any concealment of the retainer arrangement, it clearly was not known by these defendants or their counsel.

The district court itself recognized the importance of the issue, but, when the issue arose, gave signals which were variously interpreted. Plaintiffs felt that only written briefs could address the issue; defendants felt that they were permitted to gather some facts. The court specifically allowed evidence on possible conflicts of interest to be submitted to the Court Monitor for his recommendations. Four months later, the Monitor did recommend, following the joint agreement of the parties, that the retainer arrangement be terminated. This of course avoids recurrence of any problem but did not dissipate any cloud hovering over the instant fee application.

We conclude that the district court should have allowed at least a limited discovery on this newly realized issue to test whether or not some of the time claimed was solely or principally in pursuit of Association purposes that were not coterminous with those of the plaintiff class. While our own impression of the time sheets is that plaintiffs' attorneys pursued a generally scrupulous approach to conscientious reporting, a cursory scrutiny reveals some twenty-five hours recorded as being spent in meetings with the Litigation Committee. Plaintiffs' attorney Schwartz, in an affidavit, stated: "As a practical matter, these communications [at meetings] were primarily for the purpose of my providing information to the Committee, and through them to the members, concerning the status of implementation of this Decree."

This indicates to us that the time spent in these meetings might well not have been spent for the principal benefit of the plaintiff class. Perhaps in the long run the class benefited, but, in view of the fact that defendants were precluded from any discovery or testing, the entries are sufficiently suspect to warrant exclusion. Rather than remand these already prolonged proceedings for reconsideration by the district court, we think it both fair and appropriate to deduct from the award a sum roughly equal to twice the amount of the entries that appear suspect to us--$5,000.

Plaintiffs, regretfully, have engaged in three frivolous claims in their cross-appeal. Their fee application claimed hourly rates of $125 and $115 for court activities for their two attorneys and $115, $105, and $40 for decree implementation for their two attorneys and paralegal. The court, having allowed $80 an hour for court activities and $70 for decree implementation in 1982, increased the rates to $95 for court activities and $85 for implementation. Plaintiffs rely on their evidence of prevailing market rates in Springfield to urge that, as a matter of law, their evidence should carry the day. The legal work involved in this case was most intimately known by the district court; that the district court chose to increase fees by nineteen percent and no more was clearly within its discretion, especially in this advanced stage of implementation of the decree.

Plaintiffs' other two claims are that the court abused its discretion in excluding 64.7 hours of time spent (and consequently $12,629) in the search for a new court monitor and in excluding one-half of the 366 hours for what proved to be an unsuccessful attempt to negotiate a supplemental agreement. As to the former, the court acknowledged that the parties' attorneys were to consult and advise, but noted that the choosing was the court's responsibility. As to the latter, the court gently suggested that plaintiffs had "perhaps put in too much effort" on a supplemental agreement. Both judgment calls were clearly within the discretion of the court.

Accordingly, the judgment below is reduced from $132,639.55 to $127,639.55 and, as reduced, is affirmed. The plaintiffs not having prevailed on part of the defendants' appeal and having lost on their cross-appeal, it is the judgment of the court that the parties bear their own costs and that no counsel fees be awarded for the appeals.