GRENDEL'S DEN, INC., Plaintiff, Appellee, v. John P. LARKIN, et al., Defendants, Appellees, Cambridge License Commission, Defendant, Appellant. GRENDEL'S DEN, INC., Plaintiff, Appellee, v. John P. LARKIN, et al., Defendants, Appellants, 749 F.2d 945

COFFIN, Circuit Judge. Appellants challenge an award of attor-. ney’s fees. Arguing that the district court abused its discretion by granting an award which was not “reasonable” within the meaning of the Fees Act, 42 U.S.C. § 1988, appellants ask us to modify it. After carefully reviewing the district court’s opinion, as well as the evidence submitted to support appellee’s original fees application, we conclude that the district court’s analysis was in some respects erroneous and the resulting award excessive.

In view of the already protracted and expensive proceedings and the adequacy of the existing record for fee decision purposes, we make our own modifications to the award. A. BACKGROUND 1. The Underlying Case Appellee, Grendel’s Den, Inc. (Grendel’s), commenced this action on November 7, 1977, after its application for a liquor license had been denied by appellants, the Cambridge License Commission (CLC) and the Alcoholic Beverage Control Commission of Massachusetts (ABCC).

The primary charge, brought under 42 U.S.C. § 1983, was that Mass.Gen.Laws Ann. ch. 138, § 16C (section 16C), which permitted a church to “veto” the issuance of a liquor license to an establishment located within five hundred feet of the church, violated the Establishment Clause of the First Amendment. In this case Grendel’s license application had been denied because the Holy Cross Armenian Catholic Parish (the Church) of Cambridge, Massachusetts, filed a written objection with the CLC.

Also included in the complaint were charges of violations of the Due Process and Equal Protection Clauses of the Fourteenth Amendment, as well as of violations of the federal antitrust laws and the constitution and laws of Massachusetts. These charges subsequently proved to be of secondary importance in the case, which was ultimately resolved solely on the First Amendment question.

After the CLC and the ABCC filed an initial motion to dismiss, all parties agreed, and the district court ordered, that further proceedings be continued pending resolution of the same issues in a similar action then before the Supreme Judicial Court of Massachusetts. On January 17, 1979, the Supreme Judicial Court upheld the validity of

section 16C under both federal and state law, Amo v. Alcoholic Beverages Control Commission, 377 Mass. 83, 384 N.E.2d 1223 (1979). Subsequently, the CLC and the ABCC renewed their motion to dismiss, which was denied on January 14, 1980. On April 14, 1980, the parties stipulated to all material facts necessary to resolve the federal constitutional claims and to determine whether the state action exemption to the reach of the Sherman Act applied to Grendel’s antitrust claim. Each party then applied for partial

summary judgment on these issues. On August 14, 1980, the district court rejected the Equal Protection claim but entered judgment in favor of Grendel’s on the First Amendment and Due Process claims and denied defendants’ motion to dismiss the antitrust claim. On September 10, 1980, this court granted the CLC and the ABCC leave to appeal pursuant to 28 U.S.C. § 1292(b). After hearing oral argument, the panel reversed the judgment of the district court with respect to the constitutional issues and expressed doubts as to the viability of the antitrust claim. Grendel’s Den, Inc. v.

Goodwin, 662 F.2d 88 (1st Cir.1981). Rehearing en banc was subsequently granted, however, and on July 28, 1981, this court vacated its earlier judgment and affirmed the judgment of the district court. Grendel’s Den, Inc. v. Goodwin, 662 F.2d 102 (1st Cir.1981). Holding that

section 16C, on its face, violated the Establishment Clause, we did not reach the Due Process or antitrust issues in our second decision. Id. at 104. The CLC and the ABCC then appealed to the Supreme Court. The Court noted probable jurisdiction, heard oral argument, and on December 13, 1982, affirmed this court’s en banc decision and that of the district court. Larkin v. Grendel’s Den, Inc., 459 U.S. 116, 103 S.Ct. 505, 74 L.Ed.2d 297 (1982). Shortly thereafter, the CLC issued a license to Grendel’s over the opposition of the Church, and on April 1, 1983, Grendel’s began serving alcoholic beverages. In July of 1983 the Massachusetts legislature amended

Section 16C to bring it into compliance with the Supreme Court’s decision. 2. Attorney's Fees Application Having prevailed on its

section 1983 claim, Grendel’s applied to the district court for attorney’s fees and costs pursuant to the Fees Act, 42 U.S.C. § 1988. In support of the application, affidavits were filed by its counsel, Professors Laurence Tribe and David Rosenberg of Harvard Law School, and Mr. Ira Karasick. Each affidavit documented the individual’s educational and professional background, summarized the role he played in the litigation, and described the legal services that he had performed and the expenses that he had incurred.

Compensation was requested in the amount of $176,137.50 (640.5 hours at $275 per hour) for Professor Tribe, $21,-750.00 (174 hours at $125 per hour) for Professor Rosenberg, and $15,747.50 (399 hours at hourly rates ranging from $25 to $75) for Mr. Karasick. Unfortunately, despite the duration of the litigation and the distinction of counsel, no contemporaneous time records were kept or submitted in support of these requests.

Also included were a request for an upward adjustment of 50 percent to reflect the contingent nature of the fee, the long delay in payment, and the significance of the results achieved, and a request for $7,489.68 in expenses. The CLC and the ABCC responded by challenging the reasonableness of the time spent and the rates requested. On October 3, 1983, an evidentiary hearing was held. Professors Tribe and Rosenberg, Mr. Kara-sick, and an expert for the defendants gave testimony and were cross-examined regarding the accuracy of the time estimates and the reasonableness of the requested hourly rates.

Affidavits, receipts and other documents were also introduced into evidence. After the hearing supplemental briefs were filed, and Grendel’s filed a supplemental application for attorney’s fees and expenses incurred in making the original application for fees. In this second application, Mr. Jonathan Shapiro sought compensation of $17,087.50 (for 136.7 hours at $125 per hour) and reimbursement for $261.23 in expenses. On March 19, 1984, the district court issued an Opinion and Order, awarding all the itemized fees and expenses requested by Professors Tribe and Rosenberg but none to Mr. Karasick.

The court found that the applicants’ reconstruction of hours spent was “honest and reliable” and rejected the defendants’ contention that much of applicant counsels’ time had been spent on duplicative and unnecessary tasks. The court also found the evidence submitted in support of the requested hourly rates to be persuasive. It denied any award for Mr. Karasick’s hours, however, on the ground that he had served primarily as a legal intern and that his costs should be considered as part of the overhead of Professors Tribe and Rosenberg.

It commented that, although an upward adjustment would otherwise have been merited, failure to keep accurate time records precluded such an increase. Finally, the court found that Professors Tribe and Rosenberg had proven their expense claims. In April of 1984, the district court issued a brief Memorandum and Order, finding the second application for fees and expenses to be reasonable and awarding to Shapiro the full amount that he had requested. The court assessed one half of the total award against the CLC and one half against the ABCC.

On appeal, the CLC and ABCC ask this court to examine the evidence submitted at the hearing and to recalculate the award. They contend that the district court abused its discretion by failing to look critically at the alleged time spent on particular tasks, by failing to reduce the number of hours actually spent by a certain percentage to correct for the absence of contemporaneous time records, by setting hourly rates that were grossly excessive, and by reimbursing the attorneys for unnecessary or unwarranted expenses.

Also, the CLC separately challenges the district court’s assessment of one half of the award against it. It claims that it is a local entity, one of many in the state of Massachusetts, and that the ABCC, the statewide authority, should have been assessed the lion’s share of the award, especially since the case presented a challenge to a state statute that the CLC was required to enforce. The CLC argues that the only fair way of apportioning the cost of enforcing this statute is to assess most of the award against the ABCC. We consider each of these contentions. B.

DISCUSSION Although fee litigation as a substantial component of judicial time is a relatively new phenomenon, the general ground rules are well known. District courts have discretion when awarding fees and expenses under 42 U.S.C. § 1988, Hensley v. Eckerhart, 461 U.S. 424, 103 S.Ct. 1933, 1941, 76 L.Ed.2d 40 (1983), and appellate courts accord deference to the exercise of that discretion. It must, of course, be exercised within evidentiary bounds. Stanford Daily v. Zurcher, 64 F.R.D. 680, 681 (N.D.Cal.1974), aff'd, 550 F.2d 464 (9th Cir.1977).

Similarly, when reviewing fees awards for abuse of discretion, appellate courts have a duty to review carefully the basis for the award and to ensure that the amount is reasonable. Johnson v. Georgia Highway Express, Inc., 488 F.2d 714, 717 (5th Cir.1974). To permit meaningful appellate review, the district court must provide a “clear explanation of its reasons for the fee award.” Hensley, 103 S.Ct. at 1941; see also McGinty v. Beranger Volkswagen, Inc., 633 F.2d 226, 232 (1st Cir.1980). Conclusory statements concerning reasonableness are insufficient to withstand appellate review.

See Hensley, 103 S.Ct. at 1942-43 n. 15. The attorney’s account of the value of the legal services and the amount of time spent must be scrutinized with care. King v. Greenblatt, 560 F.2d 1024, 1027 (1st Cir.1977), cert. denied, 438 U.S. 916, 98 S.Ct. 3146, 57 L.Ed.2d 1161 (1978). The ultimate goal is to award fees “adequate to attract competent counsel but which do not produce windfalls.” Hensley, 103 S.Ct. at 1938 n. 4 (quoting S.Rep. No. 94-1011, p. 6 (1976)).

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. Occasionally, however, we feel constrained to examine a fee appeal in greater depth, particularly if it presents novel problems or claims.

The absence of contemporary time records, in conjunction with extraordinarily high hourly rates and claims for time spent in the most punctilious appellate research and preparation, makes this such a case. The district court started off its analysis appropriately enough by establishing that it would use the “lodestar” approach to analyze the fees application, see, e.g., Furtado v. Bishop, 635 F.2d 915, 919-20 (1st Cir.1980).

The “lodestar” — a threshold point of reference which is subject to additions or deductions for specific reasons — is determined by multiplying the total number of hours reasonably spent by a reasonable hourly rate. Hensley, 103 S.Ct. at 1939; Furtado, 635 F.2d at 920. To determine the number of hours reasonably spent, one must first determine the number of hours actually spent and then subtract from that figure hours which were duplicative, unproductive, excessive, or otherwise unnecessary. Hensley, 103 S.Ct. at 1939-40; Wojtkowski v. Cade, 725 F.2d 127, 130 (1st Cir.1984); Furtado, 635 F.2d at 920.

In calculating a reasonable hourly rate, one must consider such factors as the type of work performed, who performed it, the expertise that it required, and when it was undertaken. Furtado, 635 F.2d at 920; King, 560 at 1027-28.

Once the basic lodestar figure is calculated, the court must consider whether an upward or downward adjustment should be made “to reflect the contingent nature of any fee (if such is not reflected in the hourly rate), delay in payment, quality of representation (i.e., an unusually good or poor performance above or below the skill already reflected in the hourly rates), exceptional (and unexpected) results obtained, etc.” Furtado, 635 at 920. Finally, the court must apply a similar test of reasonableness and necessity to determine what expenses should be awarded. Palmigiano v.

Garrahy, 707 F.2d 636, 637 (1st Cir.1983). What concerns us in this case and what we now proceed to review is the district court’s analysis of the following critical points: the failure of counsel to keep contemporaneous time records, the reasonableness of hours claimed, the justifiability of the hourly rates requested, the reasonableness of certain of the claimed expenses, and the proper apportionment of the total fees assessed against the two defendants.

In reviewing these matters, we find that the evidence which causes us to criticize the district court also suggests the upper limits of what we think is reasonable. This is not, therefore, the kind of case where the finding of legal error or abuse of discretion suggests the need for a second effort by the district court. Moreover, in a case which has been years in litigation, it makes no sense to remand and require even further expenditure of time and money. See, e.g., Copeland v. Marshall, 641 F.2d at 880, 901 (D.C.Cir.1980).

As the Supreme Court has noted, “[a] request for attorney’s fees should not result in a second major litigation.” Hensley, 103 S.Ct. at 1941. 1. Failure to Keep Contemporaneous Time Records The initial problem confronting the attorneys, parties, trial court and reviewing court stems from the process by which a fee application covering six years of work must be constructed when no contemporaneous time records exist. Professor Tribe described the process in the following terms: “I surrounded myself with masses of paper, consisting of the briefs, of some notes in connection with the briefs.

I looked at the calendar to figure out when I had filed certain briefs, because in some cases I don’t think we actually had the date on it. When my calender didn’t indicate when a brief was filed and when the brief didn’t indicate, I tried to find out from others — sometimes even calling the Attorney General’s Office— because my own records are just not that comprehensive. Then I tried to work backwards. For example, I know the practices that I go through in preparing for an oral argument. I lock myself into a room and think about it. I know roughly during which periods before an argument I would have done that.

And I also know — I can remember vividly staying up all night on a number of nights before filing a petition for rehearing.

And so I sat down and tried to figure it all out.” Professor Rosenberg told of his similar approach: “Well, my basic method was to look at the documents and either the docket sheet or whatever docket entries we have in the files to indicate court dates, hearings, and the papers that were filed; and based on that and some notations in my calendar and a little notebook I was keeping — which were rather sporadic — I built up my recollection of the time I spent.” The unlikelihood, indeed perhaps the impossibility, of achieving a result under such procedures that would be fair to the applicants, fair to the paying parties, and decently reviewable by a court is obvious.

Indeed, the CLC and the ABCC contend that this failure to keep accurate and contemporaneous time records requires either the disallowal or arbitrary reduction of any fee award. At oral argument, the defendants asked specifically for a 25 percent reduction in the hours each attorney has claimed. The district court, in response to this argument, ruled that the failure to maintain contemporaneous records increased the applicant’s burden of proving the hours spent but did not by itself justify an automatic reduction or rejection of the hours claimed.

Focusing on the affidavits and schedules produced by Professors Tribe and Rosenberg, as well as on their testimony at trial, the district court found that they had presented an honest and reliable reconstruction of the time they spent on this case. Later in its opinion, however, it relied on the absence of contemporaneous records to justify the denial of a substantial upward adjustment in the basic lodestar figure, which it said it otherwise would have awarded.

The reasoning seems to be that a failure to keep records does not justify a reduction in hours claimed, but does justify eliminating a bonus which has otherwise been earned. This ambivalence persuades us that there is a need to clarify what effect the lack of contemporaneous time records should have upon a fees award. As long ago as Souza v. Southworth, 564 F.2d 609, 612 (1st Cir.1977), we warned that a failure to document time “might merit disallowal, or at least drastic reduction, of a fee award”. We recently strengthened that admonition in Wojtkowski v.

Cade, 725 F.2d at 130, by advising attorneys to “maintain detailed, contemporaneous time records”. Other courts have taken the additional step of requiring such contemporaneous time records. See, e.g., Ramos v. Lamm, 713 F.2d 546, 553 (10th Cir.1983) (prospectively requiring the filing of contemporaneous time records); New York Ass’n for Retarded Children v. Carey, 711 F.2d 1136, 1148 (2d Cir.1983) (taking identical action); National Ass’n of Concerned Veterans v. Secretary of Defense, 675 F.2d 1319, 1327 (D.C.Cir.1982).

We now take the same step and serve notice that henceforth, in cases involving fee applications for services rendered after the date of this opinion, the absence of detailed contemporaneous time records, except in extraordinary circumstances, will call for a substantial reduction in any award or, in egregious cases, disallowance. In the instant case we feel it would be unfair to apply this standard, but subject the retrospectively created record of time spent to a more exacting scrutiny than we would bring to contemporaneous and detailed records. 2.

Reasonableness of Time Spent Following exactly the process that the attorneys themselves undertook to reconstruct their time spent, we look to the documents submitted to the courts and to the hours claimed to determine whether a reasonable number of hours was spent, given the nature of the task at hand and the results achieved. In doing this we focus only on those hours which appear, at first glance, to have been unproductive, excessive, or duplicative. The remainder, which have also been carefully scrutinized, we find were reasonably spent and require no discussion.

We find that there are several groups or types of suspect hours listed by Professors Tribe and Rosenberg. Taking the most obvious first, we note that Professor Rosenberg’s individually itemized hours add up to a total of 161 hours, not the 174 hours that he claimed and was awarded. Accordingly, 13 hours must be deducted from his total. Second, we find that each of the two professors lists a total of 19 hours of oral argument before various courts.

Leaving aside the fact that most of the nineteen hours were probably not spent in the courtroom during oral argument, we see no justification for the presence of two top echelon attorneys at each proceeding. Indeed, even their records indicate that Professor Rosenberg was merely in attendance, not that he actually addressed the court. One principal reason for his presence was to help in deciding if post-argument memoranda should be attempted.

Even if we assume that he made some contribution during the 19 hours that he was present, however, we have no choice but to conclude that his presence was largely unnecessary and that compensation for his time should, therefore, be considerably reduced. Giving the benefit of all doubts to the claim of need for dual presence, we allow Professor Rosenberg 8 hours and reduce his claim by 11 hours.

Third, and most significantly, we find that on a number of occasions, Professor Tribe spent inordinately large numbers of hours analyzing the briefs of his opponents, researching and drafting briefs for Grendel’s, and preparing for oral argument. To set the stage for our analysis, we point to the fact that Professors Tribe and Rosenberg together took only 25 hours to prepare a forty-page response to defendants’ initial motion to dismiss.

Not only do we find here an economy of effort but also a legal brief sufficiently persuasive that Grendel’s subsequently decided to file no new brief after the motion to dismiss was denied and cross-motions for partial

summary judgment were filed. Indeed, the district court was so persuaded by the brief that it ruled in Grendel’s favor on all issues except the Equal Protection claim. What this demonstrates is that Professors Tribe and Rosenberg clearly had the ability to draft a compelling brief quickly and efficiently. What followed during the rest of the litigation, however, convinces us that the early economy of effort and careful focus upon only what was necessary was lost in the heat and excitement of litigating an interesting First Amendment case.

Leading up to the initial decision of this court, Professor Tribe appears to have spent a reasonable amount of time reading and analyzing the ABCC’s appeal (10 hours to digest a fifty-eight page brief which included perhaps eighteen pages of new argument) and drafting a response (16 hours to write a thirty-seven page brief which included an equivalent proportion of new argument).

However, the brief having been completed, he then spent from 5 to 7 hours a day for 11 days preparing for oral argument in this court — a total of 66 hours for a case in which the total argument of both sides would occupy less than one hour. This time, the basis for a claim for $18,150, is perhaps the clearest example of a basic assumption underlying Grendel’s fee applications: that the standard of service to be rendered and compensated is one of perfection, the best that illimitable expenditures of time can achieve.

But just as a criminal defendant is entitled to a fair trial and not a perfect one, a litigant is entitled to attorney’s fees under 42 U.S.C. § 1988 for an effective and completely competitive representation but not one of supererogation. It is simply not conceivable to us that the ablest of lawyers, having covered the same ground in arguments in the district court, would have required the equivalent of a full week and a half of billable hours to prepare for oral argument. Similarly, we find excessive the 55 hours that Professor Tribe lists for the preparation of a seventeen-page petition for rehearing.

The petition merely discussed the need to reconsider the appeal in the light of a just-handed down Supreme Court decision, Thomas v. Review Board, of the Indiana Employment Security Division, 450 U.S. 707, 101 S.Ct. 1425, 67 L.Ed.2d 624 (1981), and listed a number of ways in which plaintiff believed the original panel had erred. We accordingly reduce the allowance for pre-argument preparation and post-argument petitioning from a total of 121 hours to 60 hours.

At the Supreme Court level, the reconstructed record of time spent is as follows: 16 hours to analyze this court’s seven-page en banc decision. 55 hours to analyze and respond in 17 pages to defendants’ fifty-seven page jurisdictional statement to the Supreme Court. 25 hours analyzing defendants’ subsequent 147 pages of briefs, two-thirds of which added nothing new to what had previously been argued. 130 hours drafting a thirty-seven page reply that spends only twelve pages directly addressing the issues that were raised by our en banc decision and ultimately discussed by the Supreme Court in its decision. 70 hours analyzing defendants’ reply brief, doing further research, and preparing for oral argument. 12 hours preparing a two-page post-argument motion and supplemental statement.

Spending 308 hours for a claimed $84,700 to produce a seventeen-page motion to affirm, a thirty-seven page response brief, and a two-page supplemental statement, and to prepare for oral argument would appear to be unreasonable for almost any case. When we consider, in addition, the particulars of this litigation, the excessive nature of Professor Tribe’s claim becomes unmistakable. As indicated above, most of what defendants presented in their briefs had been argued at earlier stages of the litigation, and lead counsel is an acknowledged authority in constitutional law, not a novice.

In addition, the decision appealed from expressly did not consider the Due Process claim and held for Grendel’s solely on the Establishment Clause claim. In their jurisdictional statement, the defendants did not discuss the Due Process claim. Professor Tribe first broached the subject in his motion to affirm and then responded to defendants’ straightforward repetition of their earlier Due Process arguments by devoting almost half of the substantive portion of his brief to a complete reorganization and rewriting of the arguments that had already been well briefed for the district court and this court.

While reasonable, cautious treatment of the problem might have called for a protective

summary of Grendel’s previously made arguments, Professor Tribe’s efforts seem to us to have been excessive. The Supreme Court disposed of the First Amendment issue as briefly as did this court and never discussed the Due Process claim. Finally, when we look at the First Amendment portion of plaintiff’s brief, we note that seven of the fourteen pages are devoted to a historical analysis, largely in four footnotes, summarizing anti-“establishment” attitudes of seventeenth and eighteenth century Americans, and of lessons to be learned from sixteenth, and seventeenth and eighteenth century England.

While this analysis may be fresh and interesting, it was only briefly reflected in a footnote in the Supreme Court’s opinion. Thus again, the questions of essentiality and proportionality arise. Professor Tribe spent 20 hours producing an eighteen-line footnote on three ancient English statutes analogous to that in the instant case ... for a claimed $5,500.

In reviewing this portion of the time claimed, we realize that much of the time invested in studying and reacting to the defendants’ jurisdictional statement, brief, and reply brief can be attributed to the length, vigor, and occasionally misleading nature of the argument therein. But, making allowances for that as well as for the need for careful research, writing, and argument before the Supreme Court, we think that 200 hours is the maximum time that should reasonably have been spent. This represents a reduction of 108 hours from those claimed and found reasonable by the district court.

In total, then, we reduce the hours of Professor Rosenberg from 174 to 150 by deducting the 13 for which he has given no account and 11 for his unjustified “appearances” at various oral arguments. We reduce Professor Tribe’s hours from 640.5 to 468.5 by deducting 171 hours of excessive time spent at the appellate stages of litigation. 3.

Reasonableness of Hourly Rates “ ‘[Reasonable fees’ under § 1988 are to be calculated according to the prevailing market rates in the relevant community”, that is “those prevailing in the community for similar services by lawyers of reasonably comparable skill, experience and reputation.” Blum v. Stenson, — U.S. —, 104 S.Ct. 1541, 1547 & n. 11, 79 L.Ed.2d 891 (1984). In the instant case, we conclude that the district court’s determination of a market rate of $125 per hour for Professor Rosenberg can be sustained, but not the rate of $275 per hour for Professor Tribe.

Initially, we had considerable doubts about the rate awarded Professor Rosenberg, as well. This was particularly true with respect to the early years of the litigation when, as evidence of both his own hourly rates and those of the average partner in Suffolk County showed, a lower rate might have been called for. Nevertheless, deeming it appropriate at this juncture that our decision shall give the benefit of any doubt to the decision below, we conclude that the rate of $125 per hour, which is clearly reasonable for the last three years, can fairly be applied to the early years.

Any excess in the rate for those years should be viewed as a rough offset to . the fact that Professor Rosenberg has had to wait for so many years to receive his compensation. Accordingly, we find ourselves in agreement with the district court’s hourly rate of $125 for Professor Rosenberg and calculate that his award for 150 hours of compensable work comes to $18,750. Turning to the hourly rate of $275 for Professor Tribe, however, we find it to be excessive. The evidence submitted to justify this rate is as follows.

Professor Tribe is a widely known and respected professor of constitutional law, who has taught at Harvard since 1968 and has published numerous books and articles. He has argued a number of important constitutional cases in the federal courts including several before the Supreme Court, where he has had significant success. Of the eleven cases that he listed in his application for fees, however, only three were decided before he argued this case before this court in November of 1980.

We are urged to declare as the “relevant community” for assessing his services one embracing a small group of nationally prominent constitutional law scholars. While accepting without any reservations Professor Tribe’s scholarly reputation in his substantive field, we would hesitate to say that all of the skills of trial and appellate litigation are necessarily subsumed under academic excellence. We need not deal with this question, however, because there was no evidence of comparable rates and awards for other such scholars.

We therefore confine our evaluation to the Boston market and to Professor Tribe’s own prior record of charges. While this is the first case in which he has applied for an award under 42 U.S.C. § 1988, in his affidavit he stated in general terms that over the years in question his fees had ranged from $165 to $525 per hour. More specifically, he noted that he had contracted for $165 per hour with the California Energy Commission in 1979 to take a complex federal regulatory case which was ultimately argued before the Supreme Court on January 17, 1983, Pacific Gas & Electric Co. v.

State Energy Resources Conservation & Development Commission, 461 U.S. 190, 103 S.Ct. 1713, 75 L.Ed.2d 752 (1983). He also indicated that the City of Boston had paid him $185 per hour during the same period for taking a constitutional case to the Supreme Court. White v. Massachusetts Council of Construction Employers, 460 U.S. 204, 103 S.Ct. 1042, 75 L.Ed.2d 1 (1983). Other fees that he listed ranged from $250 to $525 per hour, but for the most part they were for more recent work as a consultant or specialist for major law firms such as Davis, Polk & Wardwell and Fried, Frank, Harris, Shriver & Kampelman.

Professor Tribe claimed that these rates and the nature of Grendel’s case justified his request of $275 per hour, and the district court awarded him that rate. We find that rate to be without justification in the record. Professor Tribe was not a particularly experienced litigator during this period. The factual elements of the case were so simple that they took only three pages to summarize. At the appellate levels, the legal problems ultimately reduced to a single, clearly understandable issue.

Although in an important and topical area of the law, the legal point that the case made is so fact-specific and, as the subsequent events have illustrated, so easily avoided by legislation, that it is not likely to be of much precedential value. In the two cases that he admitted taking after taking this case, Pacific Gas & Electric and White, Professor Tribe charged only $165 and $185 per hour, respectively. Both of them involved complex questions of federal constitutional or regulatory law.

Both clients were major institutions which could afford to pay top dollar and both agreed to pay Professor Tribe approximately $100 per hour less than what he was awarded here. The merits of both cases were argued before the Supreme Court after it heard this case. All of these facts lead inescapably to the conclusion that Professor Tribe was awarded an excessive hourly rate. In The American Lawyer Guide to Leading Law Firms (1983 Edition), submitted in evidence by Grendel’s, the current billing rates for “highest partners” in 13 of Boston’s leading firms ranged from $80 to $225 per hour.

Having in mind the facts that the time period involved here begins long before 1983 and that Professor Tribe’s time covers the entire range of services from conferences and proofreading to oral argument, we cannot conscientiously allow an overall rate, notwithstanding his academic distinction, of more than $175 an hour. Accordingly, Professor Tribe’s fees award is reduced to $81,987.50 for 468.5 hours. 4. Reasonableness of Expenses Claimed In a very brief discussion, the district court found reasonable all of Professor Tribe’s printing, room and board, travel, and mailing expenses.

Although we agree with the district court as to the latter two types of expenses, we find that it erred with respect to those for printing and room and board. The printing costs for the appeal to this court were listed as totaling $876.51. Professor Tribe conceded during the application hearing that he had failed to file an itemized bill for these costs even though he was notified by the clerk of this court that he had a right to reimbursement for them under Fed.R.App.P. 39 if he filed an application by August 11, 1981. Defendants argue that his failure precludes reimbursement at this time. We agree.

Contrary to plaintiff’s position, we see no reason to apply the Fees Act in such a way as to give delinquent applicants a second chance to recover these costs. Cf. Laffey v. Northwest Airlines, Inc., 572 F.Supp. 354, 384 (D.D.C.1983) (an untimely application for printing costs having been denied in an earlier appeal, district court refused to allow subsequent fees award for those costs). “Reasonable fees”, within the meaning of the Fees Act should not be read to include an expense that has already been effectively ignored or forgiven by the successful plaintiff.

This $876.51 printing expense must therefore be deducted from the award. Second, we consider the $3,306.31 that the district court awarded for Grendel’s costs in printing its motion, brief, and supplemental statement to the Supreme Court. Defendants argue that this cost may not be assessed against them because Rule 50.3 of the Rules of the Supreme Court states that “[t]he expenses of printing briefs, motions, petitions, or jurisdictional statements are not taxable” to the losing appellant or petitioner.

Although plaintiff argues that the Fees Act takes precedence over Rule 50.3, we note that the Fees Act mentions only “reasonable attorney’s fees” and that courts have had to define the extent to which costs other than the basic fee for hours reasonably spent should be included within the meaning of this phrase. Rule 50.3, by contrast, specifically forbids the awarding of printing costs in any case. We must assume that the Supreme Court had a reason for excluding this type of cost from the enumerated list of costs that Rule 50 allows.

Furthermore, we note that the only court found to have considered this issue held that these costs are not taxable under

section 1988. Duncan v. Poythress, 572 F.Supp. 776, 781 (N.D.Ga.1983). We therefore reject Grendel’s claim of $3,306.31 for its Supreme Court printing expenses. Finally, defendants challenge Professor Tribe’s $917.24 charge for three days and two nights of room and board while in Washington to argue before the Supreme Court. During the application hearing Professor Tribe noted that he had reduced the amount he claimed from the $1,543.00 that the hotel charged him because his wife and children stayed with him.

He admitted that his assistant also stayed with them in the three-bedroom suite but added that her assistance during the stay in Washington was necessary. He further explained that he had spent a second night at the hotel, the night after oral argument, in order to prepare and file a supplemental statement before the Court held its post-argument conference. The district court found that these explanations adequately justified his $917.24 claim. We disagree. We find the total figure to be unreasonable on its face.

Specifically, we see no justification for charging for the second night at the hotel, particularly since the supplemental statement referred to by Professor Tribe was not filed until some time after he returned to Boston. In addition, we find no explanation of how he determined what proportion of the bill should be charged to him as opposed to his family. Nor do we see any effort to minimize costs in his taking a three-bedroom suite at an expensive hotel.

Accordingly, we find that the compensable room and board for Professor Tribe and his assistant should not have exceeded $400 for a reasonable stay of two days and one night. In sum, we conclude that the district court’s award of $6,157.18 for Professor Tribe’s expenses should be reduced by $4,700.06. Thus, his total award for expenses is $1,457.12. No objection has been made to Professor Rosenberg’s claimed expenses of $165, which we allow. 5.

Attorney Shapiro’s Award The district court also awarded attorney Shapiro the full amount for which he applied for successful representation of Professors Tribe and Rosenberg in their application for attorney’s fees. The amount awarded was $17,348.73, representing $17,-087.50 in fees (136.7 hours at $125 per hour) and $261.23 in costs. In a two-page memorandum, the court justified this award on the basis of the vigorous opposition to the original fees application, the experience of Shapiro, the quality of his work, and the degree of his success.

Defendants attack this award for a number of reasons, just as they did when Shapiro originally filed his application with the district court. They argue that the number of hours should be reduced to the extent that the professors’ application was exorbitant; to the extent that time was spent by Shapiro familiarizing himself with the case, which would not be necessary if, for example, Professor Rosenberg or Professor Tribe had handled his own fee application; and to the extent that the time records that he kept do not adequately detail what he did during his conferences with Professors Tribe and Rosenberg.

Defendants also challenge the hourly rate of $125 and the expenses of $261.23. Initially, we note two aspects of the district court’s determinations with which we agree wholeheartedly. First, defendants did vigorously litigate the fees application and it was appropriate, therefore, that Professors Tribe and Rosenberg hire a lawyer such as Shapiro with experience in this area of the law to present and argue their application. Second, the expenses award of $261.23 is reasonable on its face.

Beyond these two matters, however, we conclude that the district court abused its discretion by not deducting some of the hours claimed by Shapiro, as well as by setting too high an hourly rate. As to the first issue, the defendants are correct when they state that counsel handling the fee application should be awarded fees only to the extent that he stands in the shoes of original counsel. Shadis v. Beal, 703 F.2d 71, 73 (3d Cir.1983).

In other words, time spent by Shapiro familiarizing himself with the case and obtaining information from Professors Tribe and Rosenberg may be paid for by the professors themselves, but it cannot be considered “reasonable” within the meaning of the Fees Act. This is particularly true in this case where Professors Tribe and Rosenberg failed to keep contemporaneous time records and thereby needlessly increased the difficulty of Shapiro’s task.

Accordingly, we deduct the entire time Shapiro spent initially reviewing the files and record, 8.5 hours, and half of the approximately 20 hours that Shapiro spent conferencing with Professors Tribe and Rosenberg. This later deduction of 10 hours of conferencing is based on our assumption that no more than that should have been needed to establish the hours spent and tasks performed by Professors Tribe and Rosenberg. In addition, we deduct 8 hours that are listed as having been spent upon Ira Karasick unsuccessful application for fees. Since Mr.

Karasick was not entitled to a fees award, granting one to his fees counsel would stand the Fees Act on its head and encourage the filing of nonmeritorious claims for fees. In total then, Shapiro’s claimed 136.7 hours must be reduced by 26.5 hours to 110.2 hours because of time spent learning the case and time spent on Karasick’s unsuccessful application.

With respect to the hourly rate, we simply note that Shapiro himself could show no more than $80 per hour having been awarded to him in previous fees litigation and that his own estimate, extrapolating from the 1979 Massachusetts Bar Association Survey submitted in evidence by defendants, was that such work late in 1983 was generally worth $100 per hour. He then added that he thought his superior experience in civil rights and fees litigation and the fact that the professors were paying him $125 per hour justified the additional $25 per hour.

We accept Shapiro’s conclusion that his prior rates translated, as of 1983, into a rate of $100 an hour. But we cannot justify adding on to this rate, no matter what may have been his special expertise in civil rights and fees litigation, in the light of what we deemed in Gabriele v. Southworth, 712 F.2d 1505 (1st Cir.1983), to be a maximum rate in the early 1980’s, i.e., $60 an hour, for “docu-meriting what a lawyer did and why he or she did it.” Id. at 1507., In sum, Shapiro’s total award must be reduced to $11,281.23 or $261.23 in costs and $11,020.00 (110.2 hours times $100 per hour) in fees. 6.

The CLC’s Liability for Attorney’s Fees Also on appeal is the question of whether the district court erred in assessing the award of fees and expenses equally against the CLC and the ABCC. In analyzing the CLC’s claim that the district court did err, we must look at two questions: first, whether any award could be made against the CLC; and second, assuming some assessment could be made, whether the district court abused its discretion by simply dividing the burden equally. The CLC makes two points which go to the first question of whether any assessment could be made.

First, it claims that it had no choice but to enforce

section 16C once the Church filed its written objection because the statute allowed the CLC no discretion. Second, it argues that it did nothing more than serve as the state’s agent in this situation, acting not on its own behalf but on behalf of the state legislature which enacted

section 16C. Although both of these claims give the CLC’s argument at least a superficial appeal, they ultimately are not persuasive. With respect to the point that CLC lacked discretion to choose whether or not to ignore the Church’s objection under the strict proscription of

section 16C, numerous courts have held that the governmental entity to whom enforcement authority has been delegated can be assessed attorney’s fees even if the statute in question has been found unconstitutional on its face. See, e.g., Venuti v. Riordan, 702 F.2d 6, 8 (1st Cir.1983) (“The facts that the city did not itself enact the law at issue and that some other entity may be more ‘culpable’ or ‘causally responsible’ than the city, do not ... make it ‘unjust’ as a matter of law to assess” attorney’s fees); Rosenfeld v.

Southern Pacific Co., 519 F.2d 527, 530 (9th Cir.1975) (rejecting argument that individual should not be assessed for attorney’s fees since he was merely complying with state statute which was ultimately held to be unconstitutional).

Similarly, this court has previously failed to find any justification for exempting a local entity from liability on the theory that it is nothing more than an agent of the state and has no status as a separate entity. “We see nothing in the city/state relationship that would warrant carving out a special legal rule excepting cities from cost liability when they seek to enforce state statutes.” Venuti, 702 F.2d at 8. Whether one views this situation as an enforcing of the statute or as a “seek[ing] to enforce the statute”, a distinction upon which the CLC seems to place much, makes no difference in our view.

The CLC acted on the Church’s written objection by denying Grendel’s license. It was a defendant in the subsequent litigation and participated in every phase of the long process. Given these facts it is clear that the district court acted well within its discretion by assessing some fees against the CLC. That does not mean, as we turn to the second issue, that the CLC’s proportion of the assessment should have been one half. We start by acknowledging that the law concerning apportionment of fees assessments remains relatively unsettled and that a number of theories for apportioning fees have been advanced.

See generally 2 M. Derfner & A. Wolf, Court Awarded Attorney Fees § 17.03[2] (1984). Among them are the simplest approach of dividing the award equally among the defendants, see, Vulcan Society of Westchester County, Inc. v. Fire Department of White Plains, 533 F.Supp. 1054, 1064 (S.D.N.Y.1982) (in general, fees should be divided equally among principal defendants), and the more sophisticated approaches of apportionment by degree of each defendant’s liability, see, e.g., Jose P. v. Ambach, 669 F.2d 865, 871 (2 Cir.1982); Seymour v.

Hull & Moreland Engineering, 605 F.2d 1105, 1117 (9th Cir.1979), and apportionment by relative time spent litigating against each defendant, e.g., Southeast Legal Defense Group v. Adams, 657 F.2d 1118, 1125 (9th Cir.1981). Each of these theories may be more or less valid in a given case. For example, where discrete injuries have been demonstrably caused by different parties, apportionment on the basis of degree of liability may be most practical and equitable. E.g., Dean v. Gladney, 621 F.2d 1331, 1339-40 (5th Cir.1980), cert. denied, 450 U.S. 983, 101 S.Ct. 1521, 67 L.Ed.2d 819 (1981).

Where, as here, however, several factors are involved, it makes sense to consider each of them before determining how the burden of the fees should be allocated. Turning first to the injury, while we find it to be singular in nature — the illegal denial of a license, that denial was caused by a state statute, enforced initially by a local commission but ultimately by the state commission, and upheld by the state’s highest court. Hence, though the local political entity, the CLC, is culpable, it is the state entity, the ABCC, that more closely represents the real source of the offense here.

Similarly, when we consider the relative time Grendel’s had to spend litigating against the CLC and the ABCC, respectively, and take as a rough measure of their opposition the pages of briefing that each filed and the amount of time each was allocated for oral argument, we find that the ABCC’s opposition exceeded the CLC’s by a ratio of greater than three to one. Finally, when we consider the relative ability of each of these defendants to pay the fees, see, e.g., Rogers v.

International Paper Co., 510 F.2d 1340, 1357 (8th Cir.), vacated and remanded on other grounds, 423 U.S. 809, 96 S.Ct. 19, 46 L.Ed.2d 29 (1975), we again find that the state is in a better position to absorb the cost of the fees award. While we should, and do, give this final factor relatively little weight because the CLC probably could shoulder the burden of the entire award if it was required to, we note that this factor points to the same conclusion that the others do.

Accordingly, while there may be superficial merit to assessing the CLC half of the award because it did participate to some degree at every stage of the litigation, the only factor the district court appears to have considered, discretion requires a more equitable and fair solution. Considering all of the factors cited above, we conclude that an assessment of twenty-five percent against the CLC and seventy-five percent against the ABCC represents such a solution.

While we recognize that even this apportionment only approximates the “best” solution, a clear attempt to achieve that result must be evinced by a district court if its award is to be upheld on appeal. Without intending to be too critical of the district court in this case on this particular issue, we simply note that in this still developing area of the law, district courts should make every effort to achieve the most fair and sensible solution that is possible. C. CONCLUSION In accordance with the above discussion, the order of the district court is modified in the following respects: 1.

The award for Professor Tribe’s fees is reduced to $81,987.50 and the award for his expenses to $1,457.12. 2. The award for Professor Rosenberg’s fees is reduced to $18,750.00, while the award for his expenses is left unchanged at $165.00. 3. The award for Mr. Shapiro’s fees is reduced to $11,020.00, while the award for his expenses is left unchanged at $261.23. 4. Of the award of fees and expenses, in the total amount of $113,640.85, twenty-five percent, or $28,410.21, is assessed against defendant CLC, and seventy-five percent, or $85,230.64, is assessed against defendant ABCC.

The order of the district court is hereby so modified. No costs. . At the time,

section 16C read in relevant part: "Premises ... located within a radius of five hundred feet of a church or school shall not be licensed for the sale of alcoholic beverages if the governing body of such church or school files written objection thereto....” Mass.Gen.Laws Ann. ch. 138, § 16C (West 1974). .

The relevant portion of 16C now reads: "Premises ... located within a radius of five hundred feet of a school or church shall not be licensed for the sale of alcoholic beverages unless the local licensing authority determines in writing and after a hearing that the premises are not detrimental to the educational and spiritual activities of said school or church ....” Mass.Gen.Laws Ann. ch. 138, § 16C (West Supp.1984). . Professor Tribe did submit a selected series of 34 daily entries from monthly calendar sheets covering six months in 1980 and 1981.

Each day's space was slightly over 1 square inch in area, was filled with a scribbled miscellany of personal and professional "things to do”, and was, he testified, “not a record of time spent”. Effectively, therefore, Professor Tribe and Professor Rosenberg followed what was apparently their usual custom of not keeping time records. It is surprising that they did not make an exception in this case, since, according to their own affidavits, their intention from the outset was to apply for attorney’s fees and expenses under the Fees Act if Grendel’s prevailed. .

In saying this we do not mean to imply that a facially exemplary accounting of time need not be carefully reviewed for reasonableness. . In particular, we have little difficulty with the district court's assessment of Mr. Karasick’s secondary role in the case or with the reasonableness of any of the hours claimed to have been spent during the early stages of litigation. . Specifically, Professors Tribe and Rosenberg each claim three hours of oral argument for the motion to dismiss, five hours for the motion for

summary judgment, three hours for the appeal to this court, and eight hours before the Supreme Court. . To be precise Grendel’s filed no new initial brief. It did file a reply brief of four pages. . Even at the district court level, Professors Tribe and Rosenberg began spending time that causes one to pause — 12 hours preparing a two-page motion for partial

summary judgment and a three-page stipulation of facts, 34 hours preparing for oral argument, and analyzing, and writing a four-page reply to the ABCC’s brief in support of defendants’ cross-motions for partial

summary judgment, which, although fifty pages long, was taken almost verbatim from the ABCC’s earlier brief in support of the motion to dismiss. However, we choose not to reduce those hours both because the district court was in a better position to assess their reasonableness and because when viewed against the background of the total time spent on the motion to dismiss and the motions for partial

summary judgment, they cannot be said to be clearly excessive or unproductive. . We note that the professors had agreed to pay Shapiro at a rate of $125 per hour. . While we agree with defendants that Shapiro’s time records are not so clear that we can tell exactly what he was doing during these conferences, we think his submissions provide an adequate basis for us to reasonably estimate the time he needed to familiarize himself with the case. . The CLC’s remaining argument — that it was acting in good faith by enforcing

section 16C — is equally unavailing because good faith is not a relevant factor in the determination of whether attorney’s fees can be assessed against a defendant. See, e.g., Coalition for Basic Human Needs v. King, 691 F.2d 597, 602 (1st Cir.1982).