Dismayed by chronic pollution and by the unauthorized and illegal discharge of raw and inadequately-treated sewage into the waters of Boston Harbor, the neighboring city of Quincy, Massachusetts sued the Massachusetts District Commission (MDC), a state agency which operated the port. Quincy's suit, filed in a Massachusetts state court in January 1983, was soon expanded into a full-fledged effort to compel correction of the substandard conditions. Some six months later, the Conservation Law Foundation of New England (CLF), appellant before us, brought a similar action in United States District Court under the Federal Water Pollution Control Act, 33 U.S.C. Secs. 1251-1376 (Clean Water Act). After some preliminary skirmishing, the federal court remarked the advanced state of the Quincy litigation, the striking parallel between the two suits, and the danger that "intervention by the federal district court ... [might] well hinder the cause of cleaning the harbor." CLF v. MDC, C.A. No. 83-1614, slip op. at 5-6 (D.Mass. Mar. 27, 1984) (MDC I ).1 Hoping both "to avoid duplicative litigation, [and] to allow ... the state court action to continue to yield progress," id. at 6, the district court stayed CLF's suit while retaining jurisdiction. Id. at 6-7.

Appellant's litigation was still in repose when, in early 1985, the EPA filed its own federal court action against the MDC, the Commonwealth of Massachusetts, and the newly-created Massachusetts Water Resources Authority (MWRA).2 On May 22, the district court consolidated EPA's suit with CLF's, permitted Quincy to intervene as a party plaintiff, and vacated the stay. The EPA then moved for partial summary judgment as to liability. On September 5, 1985, the court granted that motion, found MWRA liable as MDC's successor, and also granted CLF's earlier-filed summary judgment motion to the extent of the counts on which EPA had prevailed. United States v. MDC, C.A. No. 85-0489 (D.Mass. Sept. 5, 1985) (MDC II ) . The court noted that CLF's motion appeared "substantially similar" to EPA's, id. at 26; to the extent they differed, appellant's claims were denied without prejudice. Id. at 26-27.

Thereafter, the parties' efforts focused on remediation. The process was, by and large, a cooperative one. On December 23, 1985, a consent order was entered covering short-term remedial measures. After holding an evidentiary hearing in early May of 1986, the court entered its long-term remedial order. See United States v. MDC, C.A. No. 85-0489 (D.Mass. May 8, 1986) (MDC III ). CLF then filed its motion for attorneys' and experts' fees, seeking a total of close to $380,000 from MDC, MWRA, and EPA. The motion was filed under the fee-shifting provision of the Clean Water Act:

The Court, in issuing any final order in any action brought pursuant to this section, may award costs of litigation (including reasonable attorney and expert fees) to any party, whenever the court determines such award is appropriate.

33 U.S.C. Sec. 1365(d). CLF settled its fees claim against EPA but not with the state defendants. The district court adjudicated the latter claims in a lengthy memorandum and order. United States v. MDC, C.A. No. 85-0489 (Apr. 24, 1987) (MDC IV ). The court held that a fee award was appropriate because CLF had achieved some success "in certain discrete areas." Id. at 2. But after a full examination into the particulars of the request, the court slashed it deeply, awarding a total of $105,755.47 in fees and disbursements. Id. at 15. On June 4, a supplementary order was entered establishing the extent to which MDC and MWRA, respectively, would bear the brunt of the award. CLF filed its notice of appeal on July 31, 1987.

Appellees have questioned the timeliness of CLF's appeal. They concede that, because the federal sovereign was a party to the case, the appeal period was sixty days. Fed.R.App.P. 4(a)(1). Yet, they say this proceeding--which was instituted more than sixty days after the April 24 order--was late. Appellees, however, have mistaken the starting point. We regard the allocational order of June 4 as the catalyst for appeal purposes.

The general rule is that a judgment becomes final and appealable when the court enters a decision resolving the contested matter, leaving nothing to be done except execution of the judgment. Coopers & Lybrand v. Livesay, 437 U.S. 463, 467, 98 S. Ct. 2454, 2457, 57 L. Ed. 2d 351 (1978); Catlin v. United States, 324 U.S. 229, 233, 65 S. Ct. 631, 633, 89 L. Ed. 911 (1945). In this case, the April 24 order did not meet that benchmark. Although the district court had announced the amount of the award, it had not yet determined how the burden of payment would fall as between the two respondents (MDC and MWRA), nor had it suggested that the obligation might be joint and several. CLF could not have executed against either defendant at that point. Thus, the April 24 order lacked the requisite definiteness to trigger the running of the appeal period.

This conclusion is buttressed by the fact that the district judge clearly did not intend the earlier determination to be final. On the contrary, the April 24 decision explicitly contemplated the future entry of another order "to allocate between [MDC and MWRA] their respective responsibilities" for satisfaction of the award. MDC IV, supra, at 15. Inasmuch as the judge did not mean his April 24 order to be the last one as to this award, we must effectuate his intention and treat the later (June 4) order as the starting point for appeal purposes. See United States v. Evans, 365 F.2d 95, 97 (10th Cir.1966); see also Martha's Vineyard Scuba Headquarters, Inc. v. Unidentified, Wrecked and Abandoned Steam Vessel, 833 F.2d 1059, 1066-67 (1st Cir.1987) (noting judge's "special role ... in elucidating the meaning and intendment of an order which he authored"). It follows, then, that since CLF's appeal was taken within sixty days of the entry of the allocational order, it was prosecuted in a timely fashion.3

We review a fee award only for mistake of law or abuse of discretion. Wojtkowski v. Cade, 725 F.2d 127, 130 (1st Cir.1984); Maceira v. Pagan, 698 F.2d 38, 39 (1st Cir.1983). In pursuing the inquiry, we regard the judge's discretion in this area to be "broad." E.g., Gabriele v. Southworth, 712 F.2d 1505, 1506 (1st Cir.1983). We address appellant's assignments of error mindful of this standard. Cf. In re Josephson, 218 F.2d 174, 182 (1st Cir.1954) (Magruder, J.) (no abuse of discretion unless appellate tribunal forms "a definite and firm conviction that the court below committed a clear error of judgment in the conclusion it reached upon a weighing of the relevant factors"). In the realm of fee awards, "[t]o a far greater extent than is true of discrete legal issues, the battle is likely to be determined in the trial court." Rogers v. Okin, 821 F.2d 22, 30 (1st Cir.1987), cert. denied, --- U.S. ----, 108 S. Ct. 709, 98 L. Ed. 2d 660 (1988).

In this case, our perscrutation starts from the premise that courts should use "a time-and-rate-based method of calculation in cases involving the fee-shifting provisions of those federal statutes which do not expressly dictate an alternative method." Segal v. Gilbert Color Systems, Inc., 746 F.2d 78, 85-86 (1st Cir.1984). Section 1365(d) is such a provision. See id. at 86 n. 9 (citing Save Our Sound Fisheries Ass'n v. Callaway, 429 F. Supp. 1136, 1147 (D.R.I.1977)). Accordingly, we apply the vast body of jurisprudence which has sprung up in the crowded vineyard where Congress has planted a proliferous array of fee-shifting statutes. See Pennsylvania v. Delaware Valley Citizens' Council for Clean Air, 478 U.S. 546, 106 S. Ct. 3088, 3096-97, 92 L. Ed. 2d 439 (1986), on reargument, --- U.S. ----, 107 S. Ct. 3078, 97 L. Ed. 2d 585 (1987); Segal, 746 F.2d at 86.

The time-and-rate-based approach requires that the trial court fix a base fee or "lodestar." As the Court has instructed, "[t]he most useful starting point for determining the amount of a reasonable fee is the number of hours reasonably expended ... multiplied by a reasonable hourly rate." Hensley v. Eckerhart, 461 U.S. 424, 433, 103 S. Ct. 1933, 1939, 76 L. Ed. 2d 40 (1983). See also Delaware Valley, 106 S. Ct. at 3097-98; Blum v. Stenson, 465 U.S. 886, 888, 104 S. Ct. 1541, 1543, 79 L. Ed. 2d 891 (1984); Grendel's Den, Inc. v. Larkin, 749 F.2d 945, 950 (1st Cir.1984). Typically, a court proceeds by "determin[ing] the number of hours actually spent and then subtract[ing] from that figure hours which were duplicative, unproductive, excessive, or otherwise unnecessary." Grendel's Den, 749 F.2d at 950. This time around, however, the judge acknowledged the "traditional approach," yet found it unsuited to the exigencies of the particular case. MDC IV, supra, at 3. He decided instead "to identify specific tasks, allocate to [each] task a reasonable amount of time necessary to accomplish it, and then, set a compensation rate for a competent lawyer in performing that task." Id. This departure from orthodoxy, CLF tells us, was egregious error. In the singular circumstances of this case we do not agree.

As the Court recently noted, "[t]here are over 100 separate [federal] statutes providing for the award of attorney's fees ... [in a] wide variety of contexts and causes of action...." Delaware Valley, 106 S. Ct. at 3096. Given the farrago of eligible claims, it is no wonder that counsel fee petitions--like the underlying cases themselves--come in a kaleidoscopic assortment of sizes, shapes, and colors. Moreover, in the fee-shifting milieu, appellate courts are peculiarly dependent upon the personal observation, familiarity, insight, and case-specific knowledge of the trial bench. As we have recognized in the past, the judge has some acquired savvy, and ought to put it to good use. See Gabriele, 712 F.2d at 1507. The spoor which mark the trail are more readily evident at first hand. Viewed from a remove, a cold record all too often fails to convey the true character of the hunt.

In our estimation, there is little to be gained by mandating that district judges march in lockstep, following an unyielding, essentially wooden approach in all fee award cases. To say that there is one approach, and only one approach, to evaluating fee applications would be idly to sacrifice substance on the altar of form. The Court's opinions, as we read them, have left reasonably open the question of precisely how the judge ascertains "the number of hours reasonably expended...." Hensley, 461 U.S. at 433, 103 S.Ct. at 1939. For our part, we will not insist that the trier squeeze square pegs into round holes merely for the sake of some theoretical uniformity. Our bottom line has not changed since Judge Coffin wrote:

The difficulty for both fee-setting and fee-reviewing courts, in a field so susceptible to arbitrariness, is the achievement of decision-making that is fair to the parties and understandable to the community at large yet not unnecessarily burdensome to the courts themselves. Thus, we normally prefer to defer to any thoughtful rationale and decision developed by a trial court and to avoid extensive second guessing.

Grendel's Den, 749 F.2d at 950.

Consistent with this flexible paradigm, we perceive no need for the district courts to drown in a rising tide of fee-generated minutiae. We will not demand that a judge "become so deluged with details that he is unable to view the claims for fees in perspective." Gabriele, 712 F.2d at 1507. Although there is "some burden on the court to explain why it makes a substantial adjustment, up or down, of a diary-supported bill," Jacobs v. Mancuso, 825 F.2d 559, 560 (1st Cir.1987), we have "never required that [district] courts set forth hour-by-hour analyses of fee requests." Id. at 562. What we expect the trial court to do is make concrete findings, id. at 560,4 supply a "clear explanation of its reasons for the fee award," Grendel's Den, 749 F.2d at 950 (quoting Hensley, 461 U.S. at 437, 103 S.Ct. at 1941), and most of all, "retain a sense of overall proportion." Gabriele, 712 F.2d at 1507. In certain cases, these objectives may be better met by concentrating on what was necessary to be accomplished rather than on a welter of time sheets. See Hart v. Bourque, 798 F.2d 519, 523 (1st Cir.1986) ("the real test cannot be the number of hours logged, but what was done"); Grendel's Den, 749 F.2d at 954-55 (fixing 200 hours as "maximum time that should reasonably have been spent" for certain tasks); Gabriele, 712 F.2d at 1507 (specific tasks "reasonably required" no more than 64 and 72 hours, respectively).

In this instance, we believe the district court's methodology and rationale, though unusual, to be deserving of deference. The litigation arose in a highly idiosyncratic posture: CLF's suit, when filed, was substantially duplicative of Quincy's earlier action; it was stayed for an appreciable period of time; and, by the time the stay was lifted, the case had become virtually pleonastic (because of EPA's decision to prosecute its own suit). CLF "prevailed" in the technical sense, but its victory was well shy of total; the district court found it had enjoyed "in certain discrete areas ..., 'some success'...." MDC IV, supra, at 2 (citations omitted). Then again, as the court below aptly noted, this was not "truly adversarial litigation." Id. at 5.5 And there were other complications, e.g., amounts properly chargeable against the federal defendant had to be segregated and deducted, the overlaps between the work of CLF's lawyers and of Quincy's counsel had to be considered.

Given this unique set of circumstances, the judge's decision to focus on specific tasks and the time fairly needed to accomplish them, rather than on the hours claimed by each of CLF's several attorneys, made eminently good sense.

Appellant offers several reasons why the district court's approach was impermissible. We find none of these persuasive. First, though we agree that the reasonableness of counsel's commitment must generally be gauged at the time it is made, rather than in hindsight, see Boston and Maine Corp. v. Moore, 776 F.2d 2, 8-10 (1st Cir.1985), we do not believe that the district court transgressed this principle. Furthermore, we see no need for an absolute requirement that a judge shut his eyes to what a retrospective glance may reveal. Next, the asseveration that the district court chopped matters too fine--that it broke the case into too many discrete tasks and thus "confuse[d] the trees for the forest," Appellant's Brief at 21, is little more than rhetoric; the decision as to how to separate wheat from chaff in a fees contest, within broad limits, is a matter for the district court's discretion. Those frontiers were not outstripped in this case.

Perhaps the capstone of appellant's remonstrance is its insistence that the district court abandoned a market-based standard and penalized CLF's trial counsel--one of Boston's largest and most prominent law firms--for providing the same level of effort in this case as it would have mobilized in a major litigation for a private corporate client. We do not question counsel's good faith, but where fee-shifting is involved, the situation is different in at least one very material respect:

[In private practice] the fee usually is discussed with the client, may be negotiated, and it is the client who pays whether he wins or loses. The ... fee determination is made by the court in an entirely different setting: there is no negotiation or even discussion with the prevailing client, as the fee--found to be reasonable by the court--is paid by the losing party.

Blum, 465 U.S. at 895-96 n. 11, 104 S. Ct. at 1547 n. 11. Thus, the private market can at best "afford relevant comparisons." Id. And there is no single "reasonable" fee. The term connotes a range rather than an absolute. As we have said in an analogous context, borrowing Emerson's description of nature, reasonableness "is a mutable cloud, which is always and never the same." Sierra Club v. Secretary of the Army, 820 F.2d 513, 517 (1st Cir.1987).

Once this doctrine is accepted, common sense conduces to an obvious conclusion: the law firm's bill need not be swallowed whole by the client's litigation adversary just because it is the law firm's bill. That the firm is, as here, highly reputable and well regarded, does not change the equation. The loser cannot be left at the mercy of the winner's lawyers, bound to pay not a "reasonable" fee, but a fee on the order of what the victor--for whatever reasons--might be willing to tolerate. It is precisely because there is no external market-based check upon the scope of counsel's efforts in such a case that it falls to the court to act as the guarantor of fairness. See Wojtkowski, 725 F.2d at 130. In the process, the court has a right--indeed, a duty--"to see whether counsel substantially exceeded the bounds of reasonable effort." Pilkington v. Bevilacqua, 632 F.2d 922, 925 (1st Cir.1980). If so, the court must act. See, e.g., King v. Greenblatt, 560 F.2d 1024, 1027 (1st Cir.1977) (when presence of extra lawyers unnecessary, time "may obviously be discounted"), cert. denied, 438 U.S. 916, 98 S. Ct. 3146, 57 L. Ed. 2d 1161 (1978); Ricci v. Key Bancshares of Maine, Inc., 111 F.R.D. 369, 378 (D.Me.1986) (no charges warranted where presences of added lawyers "purely reassuring"); United Nuclear Corp. v. Cannon, 564 F. Supp. 581, 590 (D.R.I.1983) (courts "must zealously guard against any propensity to over-staff litigation"). In performing exactly this function in the instant case, the district court did not depart from established tenets.

We need not linger long over the amount of the award. The district court identified twelve task areas and distributed the claimed time of CLF's lawyers (2612.55 hours) among those categories. The court then deducted the time which it thought fairly allocable to the EPA rather than to the state defendants (which we compute to be 342.55 hours), using for the most part appellant's estimates. The court then painstakingly analyzed each category "to decide how much time was reasonably necessary to accomplish [that] task." MDC IV, supra, at 5. Since the amount of time which an attorney reasonably requires to do a piece of work depends to some extent (often, to a great extent) on the qualifications and work ethic of the attorney, the court gauged reasonably necessary time by "us[ing] the yardstick of a single, competent, industrious lawyer, knowledgeable in environmental litigation, supported by adequate staff" over the relevant time frame. Id. The court then proceeded to apply a range of rates to the time amassed, varying rate according to task complexity. CLF's claimed hours and the district court's findings are set forth in an appendix to this opinion.6

Although the lower court sliced the claimed hours appreciably, we cannot say that this was error. The test, of course, is not whether we, collectively, would have made such deep cuts. Rather, we examine the record to see if the trial judge's determinations seem plausible, given what has transpired in the litigation. The trial bench need not feel handcuffed by counsel's submission of time records, no matter how elaborate. To the contrary, the presiding judge must "draw[ ] on his own experience and wisdom [in] deciding whether the time spent on each phase was in excess of a reasonable amount." Gabriele, 712 F.2d at 1507. We recognize, of course, that "a court may readily make substantial reductions, particularly if based on personal observation...." Jacobs, 825 F.2d at 560. What we require is neither blind allegiance to, nor unexplained rejection of, counsel's diary entries, but findings adequate to show that acceptance, reduction, or rejection, as the case proves to be, was made reasonably and on an informed basis. While such findings may be less than exhaustive, see id. at 564 (memorializing "principle of allowing considerable room for a district court's unexplained discretion"), they must be sufficiently explicit to permit meaningful appellate review. See Hensley, 461 U.S. at 437, 103 S.Ct. at 1941 (demanding "clear explanation" of reasons for fee award); Grendel's Den, 749 F.2d at 950 (same). Where--as here--the trial court opts to make a substantial adjustment in the charges, it has a burden to spell out the whys and wherefores. See Jacobs, 825 F.2d at 560.

In this case, the obligation was fulfilled. The district court presented a rational basis for segregation of the work into the dozen selected categories. Within each classification, the court made specific findings explicating its rationale.7 It paid due obeisance to the obligation to reduce fee claims for overstaffing or "where the hours expended on the litigation are excessive given the nature of the specific task, the experience of the attorney, the number of attorneys assigned to the task, and the results obtained." MDC IV, supra, at 2 (citing Hensley, 461 U.S. at 433-34, 103 S.Ct. at 1939-40). Its assessments of reasonably productive time--though certainly arguable--appear to be solidly grounded in the record. These serial findings seem free from legal taint, within evidentiary bounds, and thus deserving of our deference. In a court's consideration of fee petitions, realignment of time inexorably leads to judgment calls. Reasonable minds may differ. But where they do, it is the district court's findings which should control, unless they are so vague as to be unreviewable or are palpably wrong. Here, they are not.

In addition to shrinking time allotments, of course, the district court also substantially massaged billing rates. This approach, to some extent, was a necessary corollary to the court's finding that the case was overstaffed. Equally as important, the lawyers submitted a battery of proposed charges for attorney time ranging from $69/hr to $185/hr. The rates used by the court (see Appendix) were well within this broad range. Moreover, they reflected the court's belief that the rate charged for substantially similar work by counsel for the city of Quincy was the best barometer of reasonableness in synthesizing the rate-of-remuneration formula.8 Whether or not we would have made an identical finding is not the point. What counts is that the district court did not act arbitrarily.

We acknowledge that the reasonableness of hourly charges depends on prevailing rates in the community for comparably qualified attorneys. Blum, 465 U.S. at 895-96, 104 S.Ct. at 1547-48; Grendel's Den, 749 F.2d at 955. In setting rates, the court should bear in mind that "the standard of services to be rendered and compensated is [not] one of perfection." Id. at 953; cf. United States v. DeCologero, 821 F.2d 39, 42 (1st Cir.1987) (Constitution requires that prison inmate receive "adequate medical care ... [not] the most sophisticated care that money can buy") (emphasis in original). A law firm's customary schedule of charges, though entitled to consideration, is not dispositive of the issue: a private client might well be willing to buy a Stradavarius when a Guadagnini would plainly do, or to pay top dollar for either when the same instrument could be purchased less expensively elsewhere. The court must, therefore, turn a realistic eye on the proffered pricing, endeavoring to fashion rates "adequate to attract competent counsel but which do not produce windfalls to attorneys." Hensley, 461 U.S. at 430 n. 4, 103 S. Ct. at 1938 n. 4 (citation omitted).

In this case, the district court, following an approved practice, see Jacobs, 825 F.2d at 561 n. 3 (listing cases), assigned different rates to different tasks. Compare, e.g., MDC IV, supra, at 7 ($100/hr rate applicable to preparation of complaint) with id. at 12 ($80/hr rate applicable to document review). More menial work--much of which was, or could have been, performed by paralegals--was accorded a rate of $40/hr. Id. The rates used by the judge were accompanied by adequate explanations,9 fell within the range of rates charged by CLF's law firm, and generally exceeded the rate billed by Quincy's lawyer. Given the state of the record, and the judge's entitlement to draw on his own knowledge of the legal marketplace, Wojtkowski, 725 F.2d at 131, we cannot say that reversible error was committed.10

It is often difficult to strike the proper note in fee-setting matters, to balance the need adequately to compensate successful counsel against the need to burden unsuccessful defendants fairly, but no more. The district court--which, as in this case, has frequently lived with the litigation and the lawyers for long periods of time, and which is likely to be more familiar with the marketplace--has the best coign of vantage. When it makes the thankless (but necessary) judgment calls thoughtfully, with no apparent misperception of law and with clearly articulated findings enjoying adequate record support, an appellate court is in a perilously poor position to descend from Olympus (as it were) and meddle.

We need go no further. Having reviewed the papers in this case with care, we cannot conclude that any abuse of discretion occurred below.

Affirmed.

                                                            Time
                                                         Allowed
                                                 Time         As
                                            Attribut-    Against
                Task             Claimed  able To EPA      State    Rate    Fee
                                                          Defendats
             Categories             Time  Involvement             Assigned  Award
 1................. Complaint     145.95  76.15 hours      30.00  $100/-  $3,0-
             (2/11/836/7/83)      hours                   hours    hour     00
 2....... Notice of Citizens'      15.00                    6.00  $100/-   $600
          Suit 3/3/834/9/85)      hours                   hours    hour
 3....... 33 U.S.C. Sec. 1311(h)     103.10                   30.00  $100/-  $3,0-
         Work 6/28/834/9/85)      hours                   hours    hour     00
 4.......... Summary Judgment     245.80       152.20      80.00  $90/h-  $7,2-
             8/25/838/23/85)      hours        hours      hours     our     00
 5.... Reply Brief/Stay Brief     105.45  43.35 hours      20.00  $90/h-  $1,8-
             3/27/845/22/85)      hours                   hours     our     00
 6........ Fed.R.Civ.P. 25(c)     138.60                   40.00  $100/-  $4,0-
                       Motion      hours                   hours    hour     00
              5/30/858/7/85)
 7...... Work on Fee Petition     180.10                 k-70.00  $70/h-  $5,0-
             5/20/869/18/86)      hours                   hours     our     00
           (see n. 6, supra )
 8.......... General Research     259.50  55.25 hours     100.00  $90/h-  $9,0-
             2/11/836/18/86)      hours                   hours     our     00
 9........... Document Review     297.20                  120.00  $80/h-  $9,6-
             2/11/836/18/86)      hours                   hours     our     00
10...... Meetings/Conferences     627.35  15.60 hours     200.00  $80/h-  $16,-
             2/11/836/18/86)      hours                   hours     our    000
11...... Preparation/Hearings     191.00                   80.00  $100/-  $8,0-
             2/11/836/18/86)      hours                   hours    hour     00
12........... Other/including     303.50                  150.00  $40/h-  $6,0-
               Paralegal Work      hours                   hours     our     00
             2/11/836/18/86)
                               ---------  -----------
                                          -----------                     -----
              TOTALS:            2612.55       342.55     926.00           $73-
                                   hours        hours                       ,2-
                                                                             00
       ----------------------