WEICK, Circuit Judge,
dissenting.
I respectfully dissent. At the time the State caved in and agreed to the consent decree, its counsel, the Attorney General of Ohio, did not have the benefit of the recent double-cell decision in its favor of the Supreme Court in Rhodes v. Chapman, - U.S. -, 101 S.Ct. 2392, 69 L.Ed.2d 59 (1981) or it would never have agreed to the harsh terms in the consent decree which called for closing down of the facility after December 31, 1983, and during the interim, periodic reduction of the number of prisoners and the initiation and maintenance of many programs which undoubtedly will cause the expenditure of huge sums of money. A copy of the consent decree is attached to appellee’s motion to dismiss the appeal from a preliminary injunction in Case No. 79-3547 on the docket of this court.
As indicated in the attached editorial of the Akron Beacon Journal of June 22, 1981, the double-cell ruling won’t end Ohio’s prison woes.
Having found that the State was easy pickings, the attorneys for the plaintiffs sought to capitalize on it by asserting claims for attorneys fees and expenses of $144,946.21 in a case that had been pending only 19 months without trial and with a single court appearance on a preliminary injunction hearing of only a day and a half. In other words, substantially all of the hours claimed by the attorneys was office work claimed to be worth as much as $90.00 an hour.
I disagree with the majority holding that in the review of allowances of attorneys fees on appeal, it is necessary for the State to establish that the district court abused its discretion. In my opinion, it is necessary for the claimant to establish by reliable records and not mere guess or estimate the actual number of hours spent and what was done or accomplished in each of such hours and the fair and reasonable value thereof. Northcross v. Board of Education, 611 F.2d 624 (6th Cir. 1979), cert. denied, 447 U.S. 911, 100 S.Ct. 2999, 64 L.Ed.2d 862 (1980). In my opinion, the record in this appeal is inadequate for appropriate appellate review.
In substance, three of the five groups preparing to litigate this case for 19 months requested fees and expenses totaling $144,-946.21, claiming 2250 hours expended in preparation and settlement of the consent decree. The district court made the following allowances:
American Civil Liberties Union (ACLU) $65,325.84
Ohio State Legal Services Assn. $ 2,632.00
Legal Aid Society of Columbus $16,464.00
Alexander, Ebinger, Fisher, McAlister and Lawrence $32,598.50
Total $117,020.34
The total allowance of $117,020.34 was for about 2250 hours of office work.
The State has not questioned the award to the Alexander law firm, no doubt be*1218cause the district court found that “[Counsel ... has provided excellent records of hourly services which appear to have been recorded contemporaneously with the performance.” The State has challenged the other allowances principally because of lack of proper documentation of the time spent in preparing the case which was never tried. The allowances made to ACLU by the court were as follows:
Ms. Jean Kamp
(1979) 491.75 hours at $90/hr.
14 hours at $45/hr.
9 hours at $63/hr.
(1978) 134 hours at $80/hr.
14 hours at $15/hr.
Total 662.75 hours
Law Clerks
178.75 hours at $30/hr.
14 hours at $15/hr.
Total 192.75 hours
Ohio State Legal Services Association:
Clement W. Pyles, Esq.
(1978) 25 hours at $70/hr.
18 hours at $49/hr.
Total 43 hours
Legal Aid Society of Columbus:
Clement W. Pyles, Esq.
(1979) 151.75 hours at $80/hr.
9 hours at $56/hr.
8 hours at $40/hr.
(1978) 1.25 hours at $70/hr.
Total 170 hours
Law Clerks
113.75 hours at $30/hr.
These findings by the district court as to the number of hours spent were not based on “the excellent records of hourly services” upon which the findings as to the private law firm were based. By affidavit, Ms. Kamp merely estimated the time she spent from the period “from March 1978 through December 7,1978.” She did keep a “log” of the time spent on the case covering the period from December 7, 1978 to December 11, 1978, but unfortunately the “log” is not part of the record in this case. There is no claim made that such log shows what Ms. Kamp did in each hour reported in the log. Nor does the record show what compensation Ms. Kamp received from ACLU or whether that organization kept any records as to her time and, performance.
Also by way of affidavit, Clement W. Pyles estimated the time he and the law clerks spent on the case but he had no records to support it or show what he or they were doing in that time.
At the hearing on the allowance of fees, both Ms. Kamp and Pyles testified as to their estimates of the hours spent on the case but did not support it by any documentary evidence.
A lawyer working for a private client, who sends his or her client a bill for services based on an hourly rate, would certainly have to document his claim if contested by showing not only the number of hours spent but what he did or accomplished in such hour. It makes some difference in the award of compensation as to what the lawyer was doing. Was he talking on the telephone, or conferring with a witness or another lawyer, writing letters, examining law, preparing a brief or motion or appearing in court. In the present case, there was no trial, only one appearance in court, attending a preliminary injunction hearing lasting one day and a half. The work was devoted to effecting a settlement of the litigation, namely the consent decree. Undoubtedly, there were numerous conferences between all of the lawyers. The ACLU and other lawyers also received substantial help and benefit from the Department of Justice lawyers, as the government was made a party plaintiff as amicus curiae. The government claimed no compensation for the use of its attorneys. Help was also given by the college of law of Ohio State University.
It is obvious that counsel for plaintiffs could not prevail against a private client with the showing they made against the State. The State necessarily will have to pay any award of fees out of funds exacted from its citizens by taxation. Because counsel’s fees will be taxed as costs against the State does not entitle them to a wind*1219fall. Also, the work done by the law clerks should have been documented and explained.
Appellants contend that counsel for plaintiffs have failed to “adequately document” the number of hours claimed. I agree. Northcross does not make clear what it means by “adequate documentation.” The decision does indicate that the district court, in determining the number of hours counsel has expended, “must first take into account the affidavits of counsel.” But the decision surely does not mean that a finding of the number of hours worked should be made exclusively on the testimony of counsel as to how many hours were expended and without any documentation. This conclusion is, I believe, supported by three elements of the Northcross opinion. First, this court was primarily concerned that the record on appeal was sufficient to allow it to review the district court’s finding. Where, as here, the finding is based exclusively on the testimony of counsel, without the support of anything resembling billing sheets or office records, this court is in essence precluded from reviewing the district court’s decision. This is so because the district court’s finding rests exclusively on an assessment of counsel’s credibility, something this court is not in a position to measure especially where there is no written record, but primarily after the fact estimates, against which to gauge counsel’s credibility. Secondly, the premise upon which the Northcross decision was based was that counsel “should be paid ‘as is traditional with attorneys compensated by a fee-paying client.’ ” Northcross, supra, at 633. To the extent that the Attorneys’ Fees Awards Act gives counsel the rights of an attorney of a fee-paying client, one would think that the Act imposes on counsel similar responsibilities. I believe it unthinkable that counsel would bill a client, particularly were the client to be billed so substantially, simply by estimating the hours spent on a case and without evidence as to how those hours were spent. Third, Northcross makes it clear that not all hours claimed are always hours for which the defendant should be charged, (see discussion below.) Yet, where as here counsel has not (and perhaps was not in a position to do so) attempted to break down how the hours were spent, defendants, and certainly this court, are in no position to determine whether certain of the hours spent should be compensated for at all or at the rate claimed by counsel. In short, this court is in exactly the same position that Northcross was designed to eliminate: an inadequate record upon which to review a fee award.
It should be noted, that the decisions of courts have indicated that counsel is required to “submit to the court a detailed record of the time spent on the case and the duties performed. The court must secure from the attorneys a full and specific accounting for their time; bills which simply list a certain number of hours and lack such important specifics as dates and the nature of the work performed during the hour or hours performed should be refused.” King v. Greenblatt, 560 F.2d 1024, 1027 (1st Cir. 1977) (citations excluded). See also, City of Detroit v. Grinnell Corp., 495 F.2d 448, 473 (2nd Cir. 1972) quoting, Lindy Brothers Builders, Inc. v. American Radiator & Standard Sanitary Corp., 487 F.2d 161, 167 (3rd Cir. 1973). (these latter cases are not cases involving § 1988, but as Northcross indicates, the legislative history of § 1988 provides that fee awards are to be governed “by the same standards which prevail in other types of equally complex Federal litigation, such as antitrust cases.” Northcross, supra, at 633, quoting Senate Report No. 94-1011, reprinted in 1976 U.S. Code Cong. & Admin.News p. 5908). A specific procedure for handling the problem of documentation is set forth in Miller, Attorneys’ Fees in Class Actions at 343-60 (1980, Fed. Judicial Center Publication).
As mentioned, one of the reasons requiring adequate documentation of the number of hours claimed by counsel is to allow the defendants, the district court, and this court to determine the extent to which certain hours claimed should be compensated for, or compensated for at a rate lower than the one claimed by counsel. Northcross, supra, at 636, specifically indicated that “hours *1220may be cut for duplication, padding or frivolous claims.” (The court further indicated that a percentage of the fee may be cut for duplication without a finding of the exact amount of such duplication. In the instant case, the fee award was decreased by a small amount to reflect duplication). Further, the court suggested that in determining the award the district court should consider the efficiency of the time spent. That is, the court indicated that the trial court should assess the time spent for specified services and determine whether “necessary services performed by attorneys . .. could reasonably have been performed by less expensive personnel [i. e. law clerks or paralegals]”. Id. at 637. Inefficient use of an attorney’s time should not, the court suggested, be compensated for at an attorney’s rate but at a law clerk’s or paralegal’s rate. Id. Accord, King v. Greenblatt, supra, at 1027. Here, the district court did not (and on the present record, probably could not) determine whether the hours claimed by counsel were not challengeable under this standard. While this is an issue which this court ought to address, it suggests another issue which this court never expressly addressed: The extent to which the district court is to evaluate the reasonableness of the time claimed by counsel.
Northcross suggests, but does not in so many words say so, that counsel should not be compensated for hours worked which were excessive. Other courts have been more explicit than this court. See e. g., Gagne v. Maher, 594 F.2d 336, 345 (2nd Cir. 1979) aff’d, 448 U.S. 1221, 100 S.Ct. 2570, 65 L.Ed.2d 653 (1980); King v. Greenblatt, supra, at 1027; Johnson v. Georgia Highway Express, Inc., 488 F.2d 714, 717 (5th Cir. 1974). This requirement follows simply as a matter of common sense: a reasonable fee based on the number of hours worked cannot be based on an unreasonable number of hours to accomplish a specific task. But it is supported by more than simply common sense. D.R. 2-106(B)(l) of the Code of Professional Responsibility provides, as an ethical principle, that a reasonable fee is one based, in part, on the time required to complete a task, not the time actually spent. To do otherwise, is to reward incompetence. See Sharp v. Hui Wahine, Inc., 49 Haw. 241, 413 P.2d 242 (1966). Not to mention, it provides an incentive for “padding” and discourages quick settlement of cases. See Grinnell, supra, at 471. It should be mentioned, that the results of a survey (the results of which are published in Attorneys’ Fees in Class Actions, supra, at 272-73) indicated that “the working of excessive hours is perceived as a widespread and serious abuse. Well over one-half of the responding judges thought attorneys often or always worked more hours than necessary. Attorneys did not think the problem was that serious, but even one-third of this group thought excessive hours were always or often worked, and only a small fraction would say this abuse never occurred.”
In the instant case, I believe that a closer look is required at the reasonableness of the number of hours worked by counsel. This case was pending for only a year and one-half, there was but a day and one-half of in-court appearances, counsel had the benefit of the United States Attorney General’s office and help from the Ohio State University, and, for the most part, the defendants (or at least it seems) agreed that the facility was not up to par and must be remedied. Nonetheless, the district court awarded fees based on 1850 hours. This figure should be compared to the 4839 hours involved in Northcross, a case involving 19 years of hard-fought litigation, including several trips to this court and the Supreme Court. In short, I believe this court should (1) clearly state that district courts are to carefully “weigh the hours claimed against [their] own knowledge, experience, and expertise of the time required to complete similar activities.” Johnson v. Georgia Highway Express, Inc., supra, at 717, and (2) remand for the district court’s determination of whether the hours awarded were not excessive.
The “fair market value of the attorney’s services” is the second variable in North-cross’ basic formula for fee calculation. The court’s analysis with respect to this issue is as follows:
*1221A reasonable hourly rate: We have indicated in several opinions that it is desirable, whenever possible, to vary the hourly rate awarded depending upon the type of service being provided. Again turning to our mandate to award fees “as is traditional with attorneys compensated by a fee-paying client,” a scale of fees as is used by most law firms is appropriate to use in making fee awards pursuant to Section 1988. The use of broad categories, differentiating between paralegal services, in-office services by experienced attorneys and trial service, would result in a fair and equitable fee.
In determining what the level of compensation for each category of service should be, the court should look to the fair market value of the services provided. In most communities, the marketplace has set a value for the services of attorneys, and the hourly rate charged by an attorney for his or her services will normally reflect the training, background, experience and skill of the individual attorney. For those attorneys who have no private practice, the rates customarily charged in the community for similar services can be looked to for guidance.
Focussing on the fair market value of the attorney’s services will best fulfill the purposes of the Fees Awards Act, by providing adequate compensation to attract qualified and competent attorneys without affording any windfall to those who undertake such representation. The entire purpose of the statutes was to ensure that the representation of important national concerns would not depend upon the charitable instincts of a few generous attorneys.
Northcross, supra, at 638.
The district court’s analysis of this issue consists entirely of the following:
The Court has reviewed the training, background, experience and skill of the attorneys. The Court is persuaded that the hourly rates requested by counsel are reasonable. It is noted that Ms. Kamp’s 1979 hourly rate is $90.00. Her representation herein reflects her accumulation of extensive and specialized knowledge of prison condition matters. In her case the $90.00 per hour is reasonable.
On its face, I believe the district court’s analysis is subject to challenge for a number of reasons. I note at the outset, again, that the record presented on this appeal is simply inadequate to allow this court to discharge its task of judicial review. This court has simply no idea what counsels’ “training, background, experience and skill” might be, and it appears that it was this which formed the basis for the basic hourly rate. Aside from this, the award is subject to the following specific challenges.
(a) The premise of the Northcross decision was that attorneys were to be compensated at a rate determined by the fair market value for their services. That is, in the words of the decision, “the rates customarily charged in the community for similar services.” The district court it would appear did not attempt to determine the prevailing wage rate for attorneys, but based the award simply on his review (on what evidence?) of the “training, background, experience and skill of the attorneys.” Though such factors are relevant, they must, under Northcross, be evaluated in the context of what these factors would mean in the real world: the world in which attorneys are paid by fee-paying clients for their services.
In this regard, the appellants have, during oral argument, claimed that they submitted to the district court the results of a survey by the Columbus Bar Association showing that the customary billing rate of attorneys in the community was substantially lower than that awarded by the district court. This survey has not been made a part of the record on appeal, but my own research confirms appellants’ claims — specifically, two survey results published in two relatively recent issues of the Ohio State Bar Association Report, of which this court may take judicial notice pursuant to Rule 201, F.R. Evid. J. Kozelek, 1980 Economic Survey of the Columbus Bar Association, Vol. LIV Ohio Bar 111 (January 26, *12221981); Lawyers’ Earnings Were Higher in 1980 — Fees up Too!, Vol. LIV Ohio Bar 1052 (June 8(A), 1981). The first, published in the January 26, 1981, issue reports, inter alia, that in 1980 (note that in this case the relevant time period is 1978 and 1979, and as the survey indicates, that rates were greater for 1980 than prior years) in Franklin County, (the County from which this case was “litigated” and from which counsel live and work) the average billing rates were as follows:
Description
years of experience
billing rate
solo practitioner
1-5
$47/hr.
solo practitioner
10-20
$61/hr.
associate in firm with 2 to 10 lawyers
$49/hr.
partners in firm with 2 to 10 lawyers
19- 20
$73/hr.
partners in firm with 2 to 10 lawyers
20- plus
$78/hr.
associate, large firm
$50/hr.
partners, large firm
10-20
$73/hr.
partner, large firm
20-plus
$85/hr.
law clerks
$25/hr.
paralegals
$31/hr.
Based on this survey, Ms. Kamp was awarded at a rate customarily charged for a partner in a large law firm with more than 20 years experience. In short, it appears doubtful that she could command $90 an hour in the local legal community. Mr. Pyles was awarded a basic rate roughly equal to a partner in a small or large law firm who had 10 to 20 years experience. Further, observe that law clerks in this case were billed at a rate greater than that prevailing in 1980.
In another survey, published in the June 8(A), 1981 issue indicated that the average hourly rate of all lawyers in 1980 in the Midwest is $50 an hour.
Aside from these surveys, there is more that indicates that the rate awarded to counsel was unreasonable. In another case, decided by the same district judge, Ms. Kamp’s superior was awarded fees based on an hourly rate of $50/hr. for out-of-court-time and $75/hr. for in-court time. See, Walking Elk Shadow v. George Denton, Case No. C-2-79-999 (S.D.Ohio April 30, 1981) (a copy of which is included in appellants’ “additional citations”). Observe also how carefully counsel in Walking Elk documented his hours and his qualifications, background, etc.)
(b) Not only is the award subject to challenge because the rate was not determined with reference to the local community, but it is clear that the district court simply ignored this court’s admonition in North-cross to consider how the hours were expended in fixing the properly hourly rate. As indicated, above, Kamp was awarded 491.75 hours at $90/hr. in 1979 and 134 hours at $80/hr. in 1978 (there were also some insignificant hours for which the rate was lower). The district court did not — and counsel’s documentation made it almost impossible — consider whether this was for in-court time, or whether it was even for legal work. It appears that this rate applies whether she was copying documents or preparing legal papers. The same can be said for the award to Pyles (at $70 to $80 an hour).
This case was brought to remedy what the State acknowledged was an inadequate facility. Nonetheless, for lack of funds, the State had no choice but to make do for the time being. As the attached Akron Beacon Journal editorial indicates, the State is now considering spending more than $432 million in building new facilities. But finding this money is not something that is particularly easy to do. This court does a disservice in awarding an excessive amount of money to attorneys who claim to want to aid the persons in the State prisons in the form of attorneys’ fees that could be used to build the prisons the State is seeking to build.
I note finally that the allowances were not made to the attorneys who allegedly rendered the services but to the organizations by whom they were employed. I think it clear that this fact should not preclude an award of fees, but it does rep*1223resent what all too commonly the Attorneys’ Fee Awards Act has become: a means by which various groups can secure funding they cannot secure through the political process. As such, I would hold as the Second Circuit has held that it is perfectly proper for the district court to reduce a fee award to reflect the contribution of public funds to a plaintiffs’ attorney. Gagne v. Maher, 594 F.2d 336, 345 (2nd Cir. 1979) aff’d, 448 U.S. 1221, 100 S.Ct. 2570, 65 L.Ed.2d 653 (1980).
For the reasons hereinbefore set forth including an erroneous standard of appellate review, namely, abuse of discretion, I would reverse the allowances made by the district court and remand for further proceedings.
[EXCERPT FROM AKRON BEACON JOURNAL]
* * * * * *
Monday, June 22, 1981
Editorials
DOUBLE-CELL RULING WON’T END OHIO’S PRISON WOES
LAST WEEK’S Supreme Court ruling that it is not unconstitutional to lock two prisoners in a cell designed for only one at Ohio’s maximum-security facility at Lucas-ville relieved at least one worry for Ohio’s penal officials and lawmakers.
It wasn’t at all clear what the state could have done if the decision had gone the other way and Ohio had later faced some court-mandated hard deadline for ending the practice.
But the 8-1 decision by the high court by no means ends Ohio’s prison worries. The state’s prisons are still bulging with far more felons than they were ever intended to house, and something badly needs to be done.
The central question is not “prisoner comfort,” a concern without a large public constituency. It is instead some hard, practical considerations that bear on the manageability and of the prison system — and, in the long term, on public safety.
The recent new rash of prison riots underscores the difficulties penal officials face in trying to maintain order with unmanageable numbers of inmates jamming their prisons. Gross overcrowding causes emotional outbreaks so violent and affecting so many inmates that they defy orderly handling even by the toughest and best-trained disciplinarians.
It also increases the tendency of “correctional” institutions to serve instead as training schools for repeat crime, discharging streams of prison-hardened thugs to seek new victims.
In the long term, Ohio and many other states would be wise to strengthen mechanisms for management of alternatives to imprisonment for offenders who present no real threat to public safety.
But for now the flow of new inmates into Ohio’s prisons, whether or not this is the best way to dispose of their cases, shows no promise of abating. And there simply isn’t room.
The Legislature is now debating an administration proposal for construction of at least five new prisons — all smaller than the one at Lucasville — and a major overhaul of the old one at Mansfield. And at least some of the new prisons would be closer than the existing ones to the major metropolitan centers from which most inmates • come, improving the chances of maintaining the family contacts that offer the greatest promise of inducing offenders to go straight.
The argument of some from Cleveland that one of the prisons should be built there because the construction and operation would provide jobs where they are badly needed is a poor one; this is not a sound basis for prison siting. But Cleveland is indeed a good place for one, because some 17 percent of the state’s prison inmates come from there.
The cost for the overall program will be high. It was first proposed at about $443 million, which has now been whittled in the Legislature to about $432 million. The ultimate figure, like every other detail about *1224the plan, will be open to further debate and modification. And lawmakers know that spending for prisons has little blood-quickening appeal for most of their constituents.
But the long-term real cost of not attacking the problem could be far greater. The proposal or some close equivalent to it ought to get early legislative approval.