delivered the opinion of the court:
The appeal stems from an award of attorney’s fees. The award has been appealed by 51 of Lake County’s taxing districts, who were represented by the petitioning firm in the underlying class action suit.
The law firm of Moss & Bloomberg, Ltd. (hereinafter referred to as M & B), represented the board of education of the Aptakisic-Tripp School District No. 102, a Lake County taxing district, in a suit against Lake County. At the time, the county, like every county in the State, collected taxes for its taxing districts and deposited the funds in interest-bearing accounts. When it paid the taxes over to the taxing districts, the county retained the interest the accounts had earned. The suit alleged that the interest, like the principal, belonged to the taxing districts and that the county’s practice of retaining it violated the Illinois Constitution (Ill. Const. 1970, art. VII, sec 9(a)). M & B had already filed virtually identical suits for other municipal clients in Du Page, Cook, and Will counties.
The Du Page County suit (hereinafter referred to as the Wood Dale case) progressed through the trial court and this court (Board of Commissioners v. County of Du Page (1982), 107 Ill. App. 3d 409) ahead of its companion suits, and ultimately resulted in the county’s practice of retaining interest being declared unconstitutional by the Illinois Supreme Court. (Board of Commissioners v. County of Du Page (1983), 96 Ill. 2d 378 (hereinafter referred to as Wood Dale I).) On remand, the trial court dismissed the Wood Dale case, finding that the practice was permitted by a newly enacted amendment to the Revenue Act of 1939. (Ill. Rev. Stat. 1981, ch. 120, par. 761.) On a second appeal to this court, we reversed that decision. (Board of Commissioners v. County of Du Page (1983), 119 Ill. App. 3d 1085.)
Du Page County then appealed to the supreme court for a second time, and that court held that the county must repay to its taxing districts all interest earned from the date of the Wood Dale I decision. Board of Commissioners v. County of Du Page (1984), 103 Ill. 2d 422 (hereinafter referred to as Wood Dale II).
During the course of the Wood Dale appeals, the Lake, Cook, and *1067Will county cases were also actively progressing. The Wood Dale decisions, however, dictated the outcome of those cases. Therefore, after each appellate decision in the Wood Dale case, this court issued a Rule 23 order in the Lake County case, simply applying the Wood Dale decision to it. Similarly, when the supreme court remanded to the trial court in Wood Dale I and II, this case also returned to the trial court for identical treatment.
After the Wood Dale II decision, classes were certified in each case so that M & B represented all of the taxing districts in each county. M & B’s Lake County suit ultimately resulted in the recovery of $2 million from Lake County to be distributed among the members of the Lake County class. M & B submitted fee petitions in each of the four counties requesting payment from their respective funds. The Cook and Du Page county classes settled with M & B for fees of approximately $400,000 and $575,000, respectively. An agreement was also reached in Will County, and, although the trial court awarded M & B less than the agreed amount, the Appellate Court for the Third District reversed that decision and awarded M & B $218,000, the amount the parties had agreed upon. (Board of Commissioners v. County of Will (1987), 154 Ill. App. 3d 395. Here, however, 51 Lake County taxing districts (hereinafter referred to as objectors) filed objections to M & B’s Lake County petition.
The petition requested payment for 1,403 hours of work performed between May 1981 and October 1984. The trial court heard the testimony of experts for both M & B and the objectors, and of Barry Moss, an M & B partner and one of three attorneys who worked on the case. The court found that M & B had overassessed research time to the case and that “the other counties should have possibly carried a little more of the load of research which was being billed to the Lake County case.” The court reduced the compensable hours to 988, and applied the hourly rate of $127, to which the parties had agreed. It then found that a multiplier of three was appropriate because “the litigation in this case was risky and *** extremely complex. *** The work of Mr. Moss could ultimately all have been for naught.” Based on those figures, the court awarded M & B $376,428.
On appeal, the objectors argue that: (1) M & B has claimed a grossly exaggerated number of hours in this case; (2) 988 hours are not compensable because they were not reasonably and necessarily expended; and (3) the trial court erred in assigning a multiplier of three. In its cross-appeal, M & B contends that the court erred in eliminating 415 hours from its petition.
An attorney’s fee award is a matter properly left to the sound *1068discretion of the trial court (Leader v. Cullerton (1976), 62 Ill. 2d 483, 488), and its decision will not be disturbed unless it resulted from an abuse of discretion. (Lurie v. Canadian Javelin Ltd. (1982), 93 Ill. 2d 231, 239; DeBruyn v. Elrod (1983), 121 Ill. App. 3d 290, 293.) “However, a court of review will not hesitate to reduce the fees awarded if it is the opinion of the court that the fees are unreasonably high.” Leader v. Cullerton (1976), 62 Ill. 2d 483, 488.
In Fiorito v. Jones (1978), 72 Ill. 2d 73, our supreme court set out the proper approach for a court awarding attorney fees in a class action case. The court must first look to the number of hours the attorney has expended on the case, disallowing those which are excessive for the activity performed or of no benefit to the class. (72 Ill. 2d 73, 89; see also Flynn v. Kucharski (1974), 59 Ill. 2d 61.) The court must then determine a reasonable hourly rate for the services performed and multiply that rate by the number of hours expended. The product is the “ ‘lodestar’ computation.” (Fiorito v. Jones (1978), 72 Ill. 2d 73, 90.) The lodestar computation may then be adjusted by applying a weighted multiplier to it. Use of the multiplier is appropriate to compensate the attorney for the “contingent nature of the undertaking and the benefits conferred upon the class.” 72 Ill. 2d 73, 91; Leader v. Cullerton (1976), 62 Ill. 2d 483, 492.
THE NUMBER OF HOURS WORKED
The supreme court stated in Fiorito that “[bjecause of the importance of the time element, it is incumbent upon the attorney to keep detailed time records during the course of the litigation.” (Fiorito v. Jones (1978), 72 Ill. 2d 73, 88; Flynn v. Kucharski (1974), 59 Ill. 2d 61, 67.) The court had given the same admonition earlier in Leader v. Cullerton, (1976), 62 Ill. 2d 483, 489, where it stated that “[i]n fairness, we must to a degree countenance the inadequacy of the time records for services rendered in class actions prior to our decision in Flynn. *** We repeat, however, *** that adequate time records must be kept in class actions ***.”
Looking first at M & B’s fee petition itself, we agree with the objectors that it is seriously deficient. The records do not show court appearances, conferences or phone calls. Rather, M & B simply describes the next pleading due and adds to that description “and related research and work.” Although the petition sets out dates, an attorney’s initials, and the number of hours worked, the descriptions used are very general and are continuously repeated until the date that pleading was filed. Objectors complain that the vast majority of the pleadings filed in Lake County are identical or nearly identical to *1069pleadings filed earlier in other counties, yet M & B has allocated large numbers of hours in this case to preparation of those documents.
We initially recognize that the objectors’ motives in challenging M & B’s fee petition are entirely irrelevant to the issue of whether or not the award of attorney fees was reasonable. (Metropolitan Sanitary District ex rel. O’Keeffe v. Ingram Corp. (1981), 85 Ill. 2d 458, 475; In re Fine Paper Antitrust Litigation (3d Cir. 1984), 751 F.2d 562, 587.) Although tedious, we will analyze some of the descriptions in M & B’s fee petition.
The description “[drafting of pleadings, class action for injunction, accounting, declaratory judgment and other relief, and related research and work” was used 21 times and describes work performed by three different attorneys from May 6 to July 8, 1981, the date the original Lake County complaint was filed, for a total of 124 hours. All but 42.75 of those hours were reported for dates after the identical complaint was filed in Du Page and Cook counties.
The description “[rjeview of motion to strike and dismiss complaint and preparation and drafting of memorandum of law in opposition to defendants’ motion to strike and dismiss” was used 19 times by three attorneys from August 29 to September 25, 1981, for a total of 107 hours. Eighty-eight and one-half of those hours'were reported for dates after a very similar motion was filed in Cook County.
“Drafting of appellate brief, research and related work” was used 39 times by three attorneys from January 11 to April 8, 1982, when the appellate brief was filed, for a total of 187.5 hours. All but 32.25 of those hours were reported for dates after the identical brief was filed in the Wood Dale appeal. We note that to prepare the Lake County reply brief, which was not identical to that filed in any other county, M & B reported only 58.25 hours of work.
The description “[plaintiffs’ cross-motion for summary judgment and research and related work” was used 15 times by three attorneys between August 18 and August 31, 1983, for a total of 86 hours. All of those hours were reported for dates after a very nearly identical motion was filed in the Wood Dale case, and 12 of those hours were reported for August 30 and 31, 1983, although the motion itself was dated August 29,1983.
The description “[d]raft and preparation of notices of appeal, docketing statement and related papers, analysis of applicability to or difference from Wood Dale II research and related work” was used nine times by three attorneys from September 3 to September 16, 1983, for a total of 40.5 hours. All of these hours were reported for dates after a virtually identical notice was filed in the Wood Dale *1070case.
The description “[r]eview and work pertaining to record on appeal; preparation and drafting of appellants’ brief for appellate court and work related thereto” was used 29 times by three attorneys between November 17 and December 29, 1983, for a total of 151 hours. All of these hours were reported for dates after a nearly identical brief was filed in Wood Dale II. M & B reported only 84.5 hours for the drafting of its second Lake County reply brief which was not identical to any prior pleading.
Barry Moss explained these hours by reporting that, while he described the hours by the end product filed, the vast majority of his time was in fact spent on research. He stated that a substantial portion of the time allocated to the initial Lake County brief was spent in determining whether it would be more advantageous to file in another forum or to raise issues not raised in the other counties. He ultimately decided to follow the same course chosen earlier, however. Forty-three of those hours were allocated to Lake County from the substantial research time expended prior to the filing of any pleading in any county. In addition, he testified that, although many of the Lake County pleadings were virtually identical, by design, to those filed in the other counties, much of the time allocated to Lake County was necessitated by the fact that the Lake County State’s Attorney raised important threshold issues not raised by the other counties. Moss considered the issues important to all of the cases because the State’s Attorneys from all four counties were in constant communication with one another, and he believed that the Lake County State’s Attorney had discussed those issues with the others. Moss stated that he researched the issues, but waited for his opponents to raise them rather than raising them himself, because he did not wish to educate his opponents with regard to theories that were potentially harmful to his cases. Many of these issues were in fact never raised in the Wood Dale case and, therefore, never decided by this court or the supreme court.
The objectors placed Barry Moss’ consolidated timesheets and those of another M & B attorney, George Marchetti, into evidence. Moss discarded his daily time records as he consolidated them with those of the other attorneys. Moss’ time records are very sketchy and difficult to read. Like the fee petition, they contain an attorney’s initials, the hours worked, and the date. Most dates contain no descriptions of work performed. Those descriptions that are given are very brief and appear only at the top of a group of dates. Many are just an instruction to his secretary to copy the description provided on a par*1071ticular date on another county’s time records. Like the fee petition, they show no court appearances or meetings. George Marchetti’s records are more detailed. They contain descriptions, however brief, for each date listed and include court appearances and conferences.
Barry Moss also submitted a list of research issues which he maintained throughout the litigation and an “additional analysis of hours billed.” These documents, however, are of little practical utility. They simply describe research topics important to the overall litigation, and while the analysis includes periods during which particular topics were researched, it gives no itemization regarding the number of hours spent on each issue, which might have aided the court in determining whether the research was necessary or beneficial to the class. (See Fiorito v. Jones (1978), 72 Ill. 2d 73, 89 (only hours beneficial to the class are compensable); see also First National Bank v. Edgeworth (1981), 94 Ill. App. 3d 873, 886; In re Fine Paper Antitrust Litigation (3d Cir. 1984), 751 F.2d 562, 596.) It is clear that there is simply no real correlation between the descriptions provided in the fee petition and the work actually performed. Moss’ testimony and supplemental analysis are much too general to cure the deficiencies. They are simply no substitute for detailed contemporaneous time records.
As noted above, the Leader court made allowance for attorneys who had kept inadequate time records prior to its Flynn decision. (Leader v. Cullerton (1976), 62 Ill. 2d 483, 489, citing Flynn v. Kucharski (1974), 59 Ill. 2d 61.) This court has also exercised leniency in awarding fees where the attorney’s records contained only generalized descriptions similar to those used by M & B here. (In re Morgan Washington Home (1982, 108 Ill. App. 3d 245.) In that case the litigation had begun in 1977 and the court noted that “the theory that greater focus should be put upon the hours spent in determining fees gained greater force during the course of the litigation here.” (108 Ill. App. 3d 245, 249.) The same cannot be said of this case. The litigation here began in 1981, three years after the Fiorito decision, and the time for giving class action attorneys the benefit of the doubt when they fail to carefully itemize the hours billed has passed. “The burden of proving the reasonable value of the services performed” clearly rests with the plaintiffs’ attorneys who must “submit to the trial court evidence from which it [can] make a reasoned decision in accordance with the applicable law.” (Fiorito v. Jones (1978), 72 Ill. 2d 73, 93.) M & B’s fee petition contains descriptions so general and unrelated to the actual work performed that the record simply contains no basis for the trial court’s award. See Leader v. Cullerton, (1976), *107262 Ill. 2d 483, 491 (records should enable the court to evaluate the necessity for and quality of time expended); see also National Association of Concerned Veterans v. Secretary of Defense (D.C. Cir. 1982), 675 F.2d 1319, 1327 (record must be sufficient to permit the court to make an independent determination that the hours worked are justified).
We note that one of the primary purposes of itemization is to ensure that there has been no unnecessary duplication of hours (In re Morgan Washington Home (1982), 108 Ill. App. 3d 245, 249) and to permit the court to carry out its obligation to disallow unnecessary, duplicative or inefficient hours. (Fiorito v. Jones (1978), 72 Ill. 2d 73, 89.) Here, M & B simultaneously handled four class action suits, filing virtually identical pleadings dealing with the same or similar issues in four different courts. We accept as true, as apparently did the trial court, M & B’s contention that it did not consolidate the cases for legitimate strategic reasons. M & B was nevertheless required to present the trial court with evidence sufficient to establish that the hours billed to Lake County were not duplications of hours billed to other counties and that common research time was fairly allocated among the classes. 72 Ill. 2d 73, 89 (hours not beneficial to the class must be disallowed); In re Coordinated Pretrial Proceedings in Antibiotic Antitrust Actions (D. Minn. 1975), 410 F. Supp. 704, 705 (to be compensable, hours must represent work actively performed for this class).
M & B has failed to meet that burden. Issues researched and charged to one county were apparently later researched and charged to Lake County, as is evidenced by Moss’ copying of research descriptions from one county’s fee petition to another’s at a later date. In addition, M & B apparently had no articulable method of allocating its research hours. Most of Moss’ testimony with respect to allocation is incoherent, but he did state that “pretty much when I was working in the time frame for that county and when their documents were due, to charge it to that county.” At best, this evidence supports only the conclusion that M & B arbitrarily charged all of its current research to whichever county had the next pleading due and described it as the drafting of that pleading “and related research,” even if the pleading itself was merely the duplicate of an earlier pleading filed in another county. This is not a reasonable method of allocation, and it illustrates M & B’s complete failure to produce evidence to show that the enormous amount of research time charged in this case was either necessary or beneficial to the Lake County class, or to any of the four classes. (See, e.g., Grendel’s Den, Inc. v. Larkin (1st Cir. 1984), 749 *1073F.2d 945 (while a reasonable amount of research time is compensable, unnecessarily exhaustive research is not).) While we agree with M & B that the Lake County class is not entitled to a “free ride” and must bear its fair share of the research burden, it appears that M & B only allocated 43 hours of general research time to it. The fact that this may have been an underassessment does not relieve M & B of the burden of substantiating the hours subsequently charged to this class.
We additionally note that for virtually every description of work performed, M & B reported that two or three attorneys performed that work. There simply was no testimony or other evidence presented from which the court could have assured itself that the attorneys were not duplicating one another’s efforts. For example, between July 23, and August 11, 1982, M & B reported that three attorneys spent 67.25 hours preparing for oral argument before this court, presumably on issues that had already been argued to this court in the Wood Dale case. This is excessive. (See Grendel’s Den, Inc. v. Larkin (1st Cir. 1984), 749 F.2d 945, 953-54 (finding 66 hours for preparation of oral argument excessive).) Nor can we ignore the evidence which established that M & B, and Barry Moss in particular, handled a substantial number of other matters during the course of these class action suits. Yet Moss claimed to have spent 5,300 hours over four years on these four cases alone. Including Moss’ time, M & B charged a total of 7,400 hours on these cases for that period. We find this to be excessive. (Cf. Flynn v. Kucharski (1974), 59 Ill. 2d 61, 67 (finding 3,700 hours for each of two attorneys over 41/a years' excessive where the evidence shows this was not the only matter handled by them during that period).) A trial court is obligated to carefully review a fee petition, and where it contains only a general accounting of hours spent, the court should review it with increased scrutiny. (Waters v. City of Chicago (1981), 95 Ill. App. 3d 919, 927.) The trial court here, however, merely reduced M & B’s compensable hours to 988 without itemizing those disallowed. We therefore conclude that its award was an abuse of discretion.
M & B relies on Prandini v. National Tea Co. (3d Cir. 1978), 585 F.2d 47, for its contention that a court is not justified in reducing an attorney’s hours for having failed to properly allocate them among classes where the record does not establish what work was actually done for each class. M & B thus appears to be arguing that its failure to produce detailed time records precludes a court from reducing its hours. We summarily reject this contention as it overlooks the basic principle that M & B bears the burden of presenting to the trial court evidence sufficient to permit it to “make a reasoned decision in ac*1074cordance with the applicable law.” (Fiorito v. Jones (1978), 72 Ill. 2d 73, 93.) The Prandini case did not hold that an attorney need not properly allocate his hours and demonstrate that he did so, but only that a court may not arbitrarily reduce those hours without making supporting findings of fact. (Prandini v. National Tea Co. (3d Cir. 1978), 585 F.2d 47, 52.) We agree with M & B that the trial court here made an apparently arbitrary reduction of its hours. However, we believe that it abused its discretion not in disallowing 415 hours but in arbitrarily awarding fees for 988 hours. While we acknowledge the recent decision of the Appellate Court for the Third District reversing the trial court’s substantial reduction of M & B’s claimed fees in the Will County case (Board of Commissioners v. County of Will (1987), 154 Ill. App. 3d 395), we find that case inapposite, as the court did not independently determine an appropriate fee award, but simply awarded M & B the amount which the Will County class members had agreed to pay it, finding that the agreement had not been collusive or fraudulent. 154 Ill. App. 3d 395, 400-01.
We have carefully reviewed the fee petition and time records to determine whether there is any basis in the record for a fee award to M & B. (See Action on Smoking and Health v. Civil Aeronautics Board (D.C. Cir. 1984), 724 F.2d 211, 220 (noting that in most cases deficiencies in documentation warrant reduction rather than denial of fees).) We conclude that only the time records kept by George Marchetti will support a fee award. Marchetti’s records contain an itemized, though brief, description for each date worked and list a total of 199.5 hours. The hours he expended are clearly reasonable, and he has listed court appearances and conferences. Most importantly, the descriptions clearly correspond to the work performed and do not conceal countless hours of questionable research time. In addition, the evidence clearly established that M & B allocated 43 hours of prefiling research time to Lake County. We conclude that this time is also compensable. However, 20 of these hours are reflected in George Marchetti’s time records, and we therefore add only 23 hours to his reported time. We therefore reduce M & B’s compensable hours to 222.5, at the stipulated rate of $127 per hour, and conclude that the appropriate lodestar figure is therefore, $28,257.50.
THE MULTIPLIER
The objectors contend that the court erred in awarding a multiplier of three. They cite two United States Supreme Court cases which analyze the use of the lodestar method in the context of statutorily authorized fee awards. (Pennsylvania v. Delaware Valley Citi *1075 zens’ Council (1986), 478 U.S. _, 92 L. Ed. 2d 439, 106 S. Ct. 3088; Blum v. Stenson (1984), 465 U.S. 886, 79 L. Ed. 2d 891, 104 S. Ct. 1541.) In Blum, the court concluded that the quality of representation, the attorney’s skill and experience, the novel or complex nature of the litigation, and the results obtained are all factors reflected in the lodestar figure itself and therefore should not serve as independent bases for an upward adjustment to that amount. (465 U.S. 866, 898-900, 79 L. Ed. 2d 891, 901-03, 104 S. Ct. 1541, 1548-50.) In Delaware Valley the court restated that holding and concluded that there is a presumption that the lodestar figure alone represents a reasonable fee. (Pennsylvania v. Delaware Valley Citizens’ Council (1986), 478 U.S ___, ___, 92 L. Ed. 2d 439, 456, 106 S. Ct. 3088, 3098.) The court concluded that, while an attorney’s superior performance may, in a rare case, justify increasing the lodestar figure, a court’s disallowance of a large number of hours as unproductive or unnecessary is not supportive of a later conclusion that the attorney’s performance was superior. (478 U.S. ___, ___, 92 L. Ed. 2d 439, 457, 106 S. Ct. 3088, 3099.) In neither case, however, did the court hold that an attorney’s having accepted the risk of loss (and of receiving no compensation whatsoever) is an inappropriate basis for enhancing the lodestar figure. (478 U.S. ___, ___, 92 L. Ed. 2d 439, 458, 106 S. Ct. 3088, 3099-4000 (expressly reserving the issue); Blum v. Stenson (1984), 465 U.S. 886, 901 n.17, 79 L. Ed. 2d 891, 903 n.17, 104 S. Ct. 1541, 1550 n.17 (finding a decision on the issue unnecessary under the facts of that case).) We therefore conclude that these cases are inapplicable where, as here, the trial court expressly bases its application of a multiplier on the attorney’s having accepted the risk that his work may go entirely uncompensated. See Fiorito v. Jones (1978), 72 Ill. 2d 73, 91 (authorizing adjustment to the lodestar computation based on the contingent nature of the litigation).
We cannot accept objectors’ argument that the risk in these cases was limited to the Wood Dale litigation, because M & B worked on the four cases simultaneously and ran equivalent risks of nonpayment in all of them. Unlike the situation in Metropolitan Sanitary District ex rel. O’Keefe v. Ingram Corp. (1981), 85 Ill. 2d 458, on which objectors rely, the bulk of the labor (and risk) here was not borne by others litigating the same case in another forum. (85 Ill. 2d 458, 482.) It was carried entirely and exclusively by M & B. Moreover, not only was there a risk of an insubstantial recovery which was not present in Ingram (85 Ill. 2d 458, 483), the evidence here clearly established that the Illinois legislature was actively attempting to protect the counties’ interests, creating a very substantial risk of no recovery at all.
*1076Nor do we agree that our supreme court’s decision in Town of City of Peoria v. O’Connor (1981), 85 Ill. 2d 195, created a high probability of success in these actions. In O’Connor, the court determined only that the taxing bodies, rather than the township collector, owned the interest on real estate tax funds. (85 Ill. 2d 195, 207.) While M & B relied heavily on O’Connor in arguing the Wood Dale I case, it bore the much heavier burden of demonstrating that a county could not retain the interest on tax funds it collected notwithstanding apparent legislative authorization to do so (Board of Commissioners v. County of Du Page (1982), 107 Ill. App. 3d 409, 412), and the additional burden of demonstrating that the interest constituted a fee based on the amount collected, the counties’ retention of which violated the Illinois Constitution (see 107 Ill. App. 3d 409, 414-15). Once these tasks were accomplished, M & B faced the additional obstacles of demonstrating that the legislature was without the authority to sanction the practice (see Board of Commissioners v. County of Du Page (1983), 119 Ill. App. 3d 1085, 1090), and of convincing the supreme court that the situation warranted a retroactive recovery of the retained interest (see Board of Commissioners v. County of Du Page (1984), 103 Ill. 2d 422, 430).
The record thus supports the trial court’s conclusion that M & B was entirely at risk of receiving no compensation throughout this litigation. We therefore find no abuse of discretion and will not disturb the court’s assignment of a multiplier of three. (See Merchandise National Bank v. Scanlon (1980), 86 Ill. App. 3d 719, 729 (a court of review will not substitute its judgment for that of the trial court merely because it might have made a different award).) We therefore affirm the trial court’s decision to award fees, but modify its judgment by reducing the amount of fees awarded to the sum of $84,772.50.
Judgment modified and affirmed.
LINDBERG, P.J., and REINHARD, J., concur.