join, dissenting:
In deciding this appeal, we are constrained to apply a specific statutory provision, section 706(k) of Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e-5(k) (1988), under which the District Court has broad authority to assess the reasonableness of a fee request. In the absence of legal error, the trial judge’s determination as to reasonableness may not be disturbed unless it is an abuse of discretion. Given this highly deferential standard of review, there is no legitimate basis whatsoever for this court to overturn the judgment of the trial judge on the facts of this case.
Furthermore, the majority’s new rule, that contingency awards are never justified, is completely without foundation. Twelve other circuits have reviewed the question at hand, and not one other circuit has adopted a rule that completely bars contingency enhancements.
The plaintiff, Mabel King, was awarded an attorney’s fee pursuant to section 706(k), which reads, in pertinent part, as follows:
In any action or proceeding under this subchapter the court, in its discretion, may allow the prevailing party ... a reasonable attorney’s fee as part of the costs [of bringing the action]_
42 U.S.C. § 2000e-5(k) (1988) (emphasis added). As may be seen from the clear terms of the statute, the “district court is expressly empowered to exercise discretion in determining whether an award is to be made and if so its reasonableness.” Blum v. Stenson, 465 U.S. 886, 902 n. 19, 104 S.Ct. 1541, 1550 n. 19, 79 L.Ed.2d 891 (1984). The Supreme Court has emphasized that it is entirely “appropriate” that the trial judge have broad authority in determining the amount of a fee award “in view of the district court’s superior understanding of the litigation and the desirability of avoiding frequent appellate review of what essentially are factual matters.” Hensley v. Eckerhart, 461 U.S. 424, 437, 103 S.Ct. 1933, 1941, 76 L.Ed.2d 40 (1983).1
In this case, the District Court awarded an attorney’s fee that compensates Ms. King’s counsel for the risk of having taken the case on a contingent-fee basis. In reaching its conclusion that a 50% enhancement over normal hourly rates was “reasonable” compensation in this case, the District Court properly looked to evidence of prevailing market practices to ensure that the fee award was roughly commensurable with what counsel could obtain on the open market. There is no doubt, given the language of the statute, that the District Court’s judgment in this regard is to be reviewed under a highly deferential, abuse-of-discretion standard. See Blum, 465 U.S. *786at 896, 104 S.Ct. at 1547-48; Pennsylvania v. Delaware Valley Citizens’ Council for Clean Air, 478 U.S. 546, 569, 106 S.Ct. 3088, 3100, 92 L.Ed.2d 439 (1986) (“Delaware Valley I”) (Blackmun, J., concurring in part and dissenting in part); City of Riverside v. Rivera, 477 U.S. 561, 586, 106 S.Ct. 2686, 2700, 91 L.Ed.2d 466 (1986) (Powell, J., concurring in the judgment). Under this standard of review, we are not to substitute our judgment of what is “reasonable” for that of the District Court; rather, we are to review the trial court’s judgment only to ensure that it is not founded upon an error of law or a clearly erroneous finding of fact and that there is some evidence in the record upon which the court “rationally could have based its decision.” Heat & Control, Inc. v. Hester Indus., Inc., 785 F.2d 1017, 1022 (Fed.Cir.1986); see also Founding Church of Scientology of Washington, D.C., Inc. v. Webster, 802 F.2d 1448, 1457 (D.C.Cir.1986) (“The abuse-of-discretion standard calls on the appellate department, in a spirit of humility occasioned by not having participated in what has gone before, not just to scrutinize the conclusion but to examine with care and respect the process that led up to it.”), cert. denied, 484 U.S. 871, 108 S.Ct. 199, 98 L.Ed.2d 150 (1987); Gomez v. Chody, 867 F.2d 395, 405 (7th Cir.1989) (“ ‘To find an abuse of discretion, we must conclude that “no reasonable [person] ... could agree with the district court.” ’ ”) (quoting Mumford v. Bowen, 814 F.2d 328, 329 (7th Cir.1986)).
Notwithstanding the latitude vested by Congress in trial courts to craft “reasonable” fee awards, the District of Columbia (“Government”) defendants in this case urge this court to substitute its judgment for that of the trial judge in overturning the award of fees. In following this suggestion, the majority seizes upon the “substantial difficulties” test found in Pennsylvania v. Delaware Valley Citizens’ Council for Clear Air, 483 U.S. 711, 731, 107 S.Ct. 3078, 3089-90, 97 L.Ed.2d 585 (1987) (“Delaware Valley II”); id. at 733, 107 S.Ct. at 3090-91 (O’Connor, J., concurring in part and concurring in the judgment), which purports to measure risk enhancement pursuant to prevailing “market” rates in the relevant legal community. The majority, however, turns the test on its head by converting it to a test whereby an individual plaintiff must establish that she personally encountered difficulty securing competent representation without the promise of a contingency enhancement. The problem with this result, however, is that it defies the premise upon which it is based. If there is a “substantial difficulties” requirement under section 706(k), it does not seek to determine whether a particular plaintiff “actually faced substantial difficulty in retaining counsel.” Morris v. American Nat’l Can Corp., 941 F.2d 710, 715 (8th Cir.1991) (citing McKenzie v. Kennickel, 875 F.2d 330, 337-38 (D.C.Cir.1989); see 875 F.2d at 338 (“Justice O’Connor’s opinion instructs us to adopt a class-wide view of contingent cases; if the unavailability of risk enhancements would have caused plaintiffs to have experienced ‘substantial difficulty’ in locating counsel, then, notwithstanding the particular circumstances of their case, such an enhancement may be granted.”)). Therefore, the trial judge surely did not abuse his discretion in failing to apply the majority’s distorted construction of the so-called “substantial difficulties” test. There is no “actual difficulties” requirement under section 706(k), and this court has no authority to amend the statute to include such a restriction.
Just recently, in rejecting a claim for expert fees as a part of a claim for attorney’s fees, the Supreme Court reminded us that we must enforce fee statutes as written. On this point, Justice Scalia, borrowing a well-known passage from an opinion by Justice Brandéis, said:
[The statute’s] language is plain and unambiguous. What the Government asks is not a construction of a statute, but, in effect, an enlargement of it by the court, so that what was omitted, presumably by inadvertence, may be included within its scope. To supply omissions transcends the judicial function.
West Va. Univ. Hosps., Inc. v. Casey, — U.S. -, 111 S.Ct. 1138, 1148, 113 L.Ed.2d 68 (1991) (quoting Iselin v. United States, *787270 U.S. 245, 250-51, 46 S.Ct. 248, 250, 70 L.Ed. 566 (1926)). In first utilizing an actual difficulties gloss' to section 706(k), and then completely barring contingency enhancements, the majority opinion in this case “transcends the judicial function.” Because there is nothing in the statute or the relevant Supreme Court case law that would support the majority’s conclusion, we dissent.
I.
By now, it should be beyond dispute that the fee-shifting provision of Title VII permits district courts to enhance time-based fee awards to take account of the fact that an attorney has taken a case on a contingent-fee basis. It is, of course, true that, in determining what is “a reasonable attorney’s fee” in any given case, the trial judge normally begins by calculating the prevailing attorney’s so-called “lodestar” fee. As the Supreme Court has explained:
The most useful starting point for determining the amount of a reasonable fee is the number of hours reasonably expended on the litigation multiplied by a reasonable hourly rate. This calculation provides an objective basis on which to make an initial estimate of the value of a lawyer’s services.
Hensley, 461 U.S. at 433, 103 S.Ct. at 1939 (emphasis added). This calculation, however, is only a “starting point” and “does not end the inquiry. There remain other considerations that may lead the district court to adjust the fee upward or downward....” Id. at 434, 103 S.Ct. at 1940; see also Blanchard v. Bergeron, 489 U.S. 87, 94, 109 S.Ct. 939, 944-45, 103 L.Ed.2d 67 (1989); Blum, 465 U.S. at 888, 104 S.Ct. at 1544 (“[adjustments to that [lodestar] fee then may be made as necessary in the particular case”).
Among these “other considerations,” it appears quite certain that Congress intended that the courts would take into account whether a lawyer had taken a case on a fixed- or contingent-fee basis. This can be inferred from Congress’ approving citation of a 1974 Fifth Circuit decision, Johnson v. Georgia Highway Express, Inc., 488 F.2d 714 (5th Cir.1974), which included the contingent nature of payment among 12 factors that trial courts should consider in calculating fee awards. See S.Rep. No. 1011, 94th Cong., ¡2d Sess. 6 (1976); H.R.Rep. No. 1558, 94th Cong., 2d Sess. 8-9 (1976); see also Blum, 465 U.S. at 902-03, 104 S.Ct. at 1550-51 (Brennan, J., concurring) (Congress’ approval of Johnson and related cases makes it “clear ... that Congress authorized district courts to award upward adjustments to compensate for the contingent nature of success”). “Johnson’s ‘list of 12,’ ” the Supreme Court has often observed, “provides a useful catalog of the many factors to be considered in assessing the reasonableness of an award of attorney’s fees....” Blanchard, 489 U.S. at 93, 109 S.Ct. at 944; see also Hensley, 461 U.S. at 429-30, 434 n. 9, 103 S.Ct. at 1937-38, 1940 n. 9 (looking to Fifth Circuit’s Johnson opinion in determining congressional intent with regard to fee awards); Blum, 465 U.S. at 893-95, 104 S.Ct. at 1546-47 (same); id. at 902-03, 104 S.Ct. at 1550-51 (Brennan, J., concurring) (same).
Apart from these indications of congressional intent, the Supreme Court also has acknowledged the propriety of considering the uncertainty of payment in calculating a fee award. Five Justices undoubtedly agreed in Delaware Valley II that “Congress did not intend to foreclose consideration of contingency in setting a reasonable fee under fee-shifting provisions” such as that found in Title VII. 483 U.S. at 731, 107 S.Ct. at 3089-90 (O’Connor, J., concurring in part and concurring in the judgment); see also id. at 739, 107 S.Ct. at 3093-94 (Blackmun, J., dissenting, joined by Brennan, Marshall & Stevens, JJ.) (“Congress envisioned that district courts would take the fact of contingency into account when calculating a reasonable attorney’s fee”). Two years later, in Blanchard, the Court held that, while a plaintiff’s contingent-fee contract with her attorney is by no means dispositive of a subsequent judicial assessment of “a reasonable attorney’s fee” in a case, “[t]he Johnson contingency-fee factor is ... a factor.” 489 U.S. at 93, 109 S.Ct. at 944 (emphasis added).
*788Moreover, perhaps the one rule that has emerged more clearly than any other from the Supreme Court’s pronouncements in this area is that court-ordered attorney’s fees are to reflect prevailing market rates and practices. See, e.g., Missouri v. Jenkins, 491 U.S. 274, 283, 109 S.Ct. 2463, 2469, 105 L.Ed.2d 229 (1989) (“Our cases have repeatedly stressed that attorney’s fees awarded [by a court] ... are to be based on market rates for the services rendered.”); Blum, 465 U.S. at 895, 104 S.Ct. at 1547 (“The statute and legislative history establish that ‘reasonable fees’ ... are to be calculated according to the prevailing market rates in the relevant communi-ty_”); Delaware Valley II, 483 U.S. at 733, 107 S.Ct. at 3090-91 (O’Connor, J., concurring in part and concurring in the judgment); id. at 754, 107 S.Ct. at 3101-02 (Blackmun, J., dissenting). In this way, court-ordered fees will track market forces, fulfilling the congressional purpose of ensuring that attorneys will be available to prosecute Title VII cases and vindicate the fundamental national policies embodied in that statute. See Jenkins, 491 U.S. at 283 n. 6, 109 S.Ct. at 2469 n. 6; Blum, 465 U.S. at 903-04, 104 S.Ct. at 1551 (Brennan, J., concurring); Hensley, 461 U.S. at 447, 103 S.Ct. at 1946-47 (Brennan, J., concurring in part and dissenting in part).
There should be no controversy in the observation that attorneys in the private legal-services market ordinarily demand somewhat greater compensation in exchange for taking a case on a contingent-fee basis. It appears indisputable that “[Ijawyers operating in the marketplace can be expected to charge a higher hourly rate when their compensation is contingent on success than when they will be promptly paid[ ] irrespective of whether they win or lose.” Blum, 465 U.S. at 903, 104 S.Ct. at 1551 (Brennan, J., concurring); see also Berger, Court Awarded Attorneys’ Fees: What Is “Reasonable”?, 126 U.Pa.L.Rev. 281, 324-25 (1977) (“The experience of the marketplace indicates that lawyers generally will not provide legal representation on a contingent basis unless they receive a premium for taking that risk.”). Thus, commentators and courts have long and widely agreed that, in assessing the fair market value of legal services, both inside and outside the court-ordered fee context, some enhancement is required to compensate for the attorney’s assumption of risk in a contingent-fee case. See, e.g., Copeland v. Marshall, 641 F.2d 880, 892-93 (D.C.Cir.1980) (en banc)-, id. at 927 (Wilkey, J., dissenting); Evans v. Sheraton Park Hotel, 503 F.2d 177, 188 (D.C.Cir.1974) (adopting Johnson’s “list of 12”); 2 M. Derfner & A. Wolf, Court Ordered Attorney Fees ¶ 15.01[2][c], at 15-16 (rev. ed. 1990) (“Most courts realize that where payment of a fee is contingent on success an attorney should receive a larger overall fee than where payment is guaranteed regardless of out-come_”) (footnote omitted); id., ¶ 16.-04[4]; S. Speiser, Attorneys’ Fees § 8:10, at 319 (1973) (“The fact that an attorney's employment is undertaken on a contingent basis is a proper factor to be considered in assessing a reasonable compensation for his services, the courts generally taking the view that a larger fee will be authorized where its payment depends upon the attorney’s success than where he is to be paid whether or not his efforts are successful.”) (footnote omitted); Leubsdorf, The Contingency Factor in Attorney Fee Awards, 90 Yale L.J. 473, 501 (1981); Berger, supra, at 324-26.
Furthermore, it is absolutely clear that the lodestar is not the sole measure of a reasonable attorney’s fee. It is true that the Supreme Court has said that “many of the Johnson factors,” such as “ ‘novelty [and] complexity of the issues,’ ‘the special skill and experience of counsel,’ the ‘quality of representation,’ and the ‘results obtained’ from the litigation[,] are presumably fully reflected in the lodestar amount, and thus cannot serve as independent bases for increasing the basic fee award,” see Delaware Valley I, 478 U.S. at 565, 106 S.Ct. at 3098 (quoting Blum, 465 U.S. at 898-900, 104 S.Ct. at 1548-1550); but, in making this observation, the Court has excluded the consideration of “enhancement of the lodestar[ ] based on the likelihood of success[ ] or ... the risk of loss” from any presumption that the lodestar represents a *789reasonable fee, id. 478 U.S. at 568, 106 S.Ct. at 3099-3100.2 Indeed, in Delaware Valley I, the Court reserved until Delaware Valley II the question of when and to what extent contingency enhancements might be awarded. As indicated earlier, a majority of the Court in Delaware Valley II agreed that contingency enhancements may be awarded as a part of a reasonable attorney’s fee. In other words, a majority of the Court in Delaware Valley II declined to apply any “presumption” that the lodestar normally represents a reasonable fee so as to defeat claims of enhancement based on the likelihood-of-success/risk-of-loss factor.3
We recognize that the contingency factor could be accounted for within the initial lodestar calculation. A court could simply enhance the “reasonable” hourly rate used in calculating the lodestar and forgo post-lodestar adjustments. See, e.g., Copeland, 641 F.2d at 893 (“To the extent ... that an hourly rate underlying the ‘lodestar’ fee itself comprehends an allowance for the contingent nature of the availability of fees in Title VII litigation ..., no further adjustment duplicating that allowance will be made.”); Berger, supra, at 325-26. This approach is problematic because there really is “no such thing as a market hourly rate in contingent litigation.” 2 M. Derf-ner & A. Wolf, supra, ¶ 16.04[4][a], at 16-100.15. Accordingly, most courts choose to employ “real” hourly rates in the lodestar calculation — i.e., “the normal hourly charge in the community for noncontingent, contemporaneous payment in litigation of similar complexity and difficulty, by a lawyer with similar experience and reputation,” id., ¶ 16.03[l][a], at 16-47 (footnotes omitted)4 — and only later adjust the product upward to account for the contingency factor. See id., ¶ 16.04[4][a], at 16-100.15-.16. The difference in the mathematical formulas makes no difference in the result, of course, so long as the court is careful to avoid “double-counting” by blending the two approaches. See Copeland, 641 F.2d at 893. The point here is, however, that under the approach followed by most courts — and followed by the District Court in this case — the contingency factor is not subsumed within the initial lodestar calculation; consequently, if the lodestar is not itself adjusted upward, the lawyer’s economic risk will go uncompensated.
The majority rejects the prevailing view that enhancements are available because it can find no governing principle in Delaware Valley II. With no precedent or logic to support its opinion, the majority simply decides that contingency enhancements never should be permitted. This rule is created completely out of new cloth. Neither the plurality, concurring, nor dissenting opinion in Delaware Valley II holds that contingency enhancements are never available. In a sweep of reasoning that defies comprehension, the majority attempts to dismiss Part V of Justice White’s plurality opinion because Justice O'Connor declined to join this portion of the plurality. But, of course, as is clear from the Court’s opinion, Part V remains as written and means what it says: see, e.g., 483 U.S. at 728, 107 S.Ct. at 3088 (“enhancement for the risk of nonpayment should be reserved for exceptional cases where the need and justification ... are readily apparent and are supported by evidence in the record and specific findings by the courts”); id. at 731, 107 S.Ct. at 3089-90 (“[A] fee award should *790be informed by the statutory purpose of making it possible for poor clients with good claims to secure competent help. Before adjusting for risk assumption, there should be evidence in the record, and the trial court should so find, that without risk enhancement plaintiff would have faced substantial difficulties in finding counsel in the local or other relevant market.”). Whatever substantive criteria these statements may stand for, they certainly do not reflect a per se rule against contingency enhancements. In any event, the one thing that is absolutely clear from Delaware Valley II is that the Supreme Court declined to reject the possibility of contingency enhancements.
Furthermore, the majority in this case, having decided that Delaware Valley II does not control interpretation of the fees statute, does absolutely no work to interpret the statute. Rather, the majority relies on the utterly irrelevant proposition that “the question of attorney’s fees must not turn into major litigation itself,” see Delaware Valley II, 483 U.S. at 722, 107 S.Ct. at 3085, to reach a per se rule barring all contingency enhancements. Of course, the majority’s fear could be avoided equally well by routinely granting contingency enhancements. Even a test which looks to whether, in the relevant market, contingency fees are needed to induce attorneys to represent plaintiffs in these actions, will not create much hardship after district and circuit courts establish precedent regarding the major markets. Given that the contingency enhancement question merely asks the court to decide the prevailing market for legal services, it is no more onerous than any other fee inquiry.
The majority’s result also flies in the face of the decisions from every circuit that has considered the issue since Delaware Valley II. In creating a split in the circuits, the majority now causes the D.C. Circuit to stand oddly alone on this question. Of the thirteen circuits applying Delaware Valley II, none — save the D.C. Circuit — has completely ruled out contingency enhancements. See, e.g., Jacobs v. Mancuso, 825 F.2d 559, 561 (1st Cir.1987) (disallowing contingency, not because of per se rule, but because “liability here was so plain ... that, as a practical matter, the risk of not recovering a fee was all but eliminated”); Friends of the Earth v. Eastman Kodak Co., 834 F.2d 295, 298 (2d Cir.1987) (fee enhancement available when “[wjithout the possibility of a fee enhancement ... competent counsel might refuse to represent ... clients thereby denying them effective access to the courts”) (quoting Lewis v. Coughlin, 801 F.2d 570, 576 (2d Cir.1986)); Kelly v. Matlack, 903 F.2d 978 (3d Cir.1990) (“in order to qualify for an enhancement, a plaintiff must establish that without adjustment it would have faced substantial difficulties in finding counsel in the ... relevant market”) (citations omitted); Craig v. Dep’t of Health & Human Servs., 864 F.2d 324, 327 (4th Cir.1989) (dicta) (reading Delaware Valley II to permit fees in “exceptional circumstances”); Alberti v. Klevenhagen, 896 F.2d 927, 936 (5th Cir.) (enhancement available when district court “make[s] the findings required by Justice O’Connor’s concurrence in Delaware Valley II”), reh’g granted in part, 903 F.2d 352 (5th Cir.1990); Perotti v. Seiter, 935 F.2d 761, 765 (6th Cir.1991) (“This court has upheld multipliers for the risk of non-compensation in contingent-fee cases subsequent to Delaware Valley.”) (citing Fite v. First Tennessee Production Credit Ass’n, 861 F.2d 884 (6th Cir.1988)); Soto v. Adams Elevator Equip. Co., 941 F.2d 543, 553 (7th Cir.1991) (following Justice O’Connor’s test as the Delaware Valley II “holding”); Morris v. American Nat’l Can Corp., 941 F.2d 710, 715 (8th Cir.1991) (“Justice O’Connor’s opinion in Delaware Valley II is the current legal standard for awarding contingency enhancements_ We are persuaded that the district court abused its discretion in concluding that [plaintiff] failed to establish that she would have faced substantial difficulties in retaining counsel absent risk enhancement.”); Bouman v. Block, 940 F.2d 1211, 1235-36 (9th Cir.1991) (upholding fee on the basis of district court’s findings matching Justice O’Connor’s test), cert. denied, — U.S. -, 112 S.Ct. 640, 116 L.Ed.2d 658 (1991); *791Smith v. Freeman, 921 F.2d 1120, 1123 (10th Cir.1990) (quoting the Delaware Valley II plurality that “enhancement for the risk of non-payment should be reserved for exceptional cases where the need and justification for such enhancement are readily apparent and are supported by evidence in the record and specific findings by the courts”); Martin v. University of South Alabama, 911 F.2d 604, 610-12 (11th Cir.1990) (following Justice O’Connor’s test); Crumbaker v. Merit Systems Protection Board, 827 F.2d 761, 761 (Fed.Cir.1987) (“the Board on remand shall consider the degree to which the relevant market compensates for contingency and whether any enhancement is necessary to bring the fee within a range that would attract competent counsel”).
Additionally, despite the majority’s assertion that the tests developed in many circuits are “difficult, if not impossible, to meet,” even courts of those circuits have continued to award contingency enhancements in some circumstances. See, e.g., Morris v. American Nat’l Can Corp., 941 F.2d 710, 716 n. 2 (8th Cir.1991); Curry v. Contract Fabricators Inc. Profit Sharing Plan, 891 F.2d 842, 849-50 (11th Cir.1990); Shirley v. Chrysler First, Inc., 763 F.Supp. 856, 860 (N.D.Miss.1991); Vargas v. Calabrese, 750 F.Supp. 677 (D.N.J.1990); Bauman v. Jacobs Suchard, Inc., No. 89 C 5452, 1991 WL 125952, 1991 U.S. Dist. LEXIS 8847 (N.D.Ill.1991).
Thus, setting aside the more contentious question of what degree of enhancement is appropriate, there ought to be no dispute that some upward adjustment of the lodestar fee is permissible where the prevailing attorney assumed the greater risk inherent in contingent-fee cases. There is simply no justification — in the statute, in the case law or in common sense — for the suggestion that contingent-fee lawyers may not be fully compensated for their services.
II.
With regard to the particular facts of this case, there is no basis under the abuse-of-discretion standard of review to overturn the District Court’s decision to allow a 50% enhancement of the lodestar fee submitted by Ms. King’s counsel. Although the Supreme Court’s guidance concerning the appropriate method for calculating a contingency adjustment has been regrettably uncertain, the trial court’s decision in this case is consistent with what standards can be gleaned from recent cases.
The clearest instruction found in the Supreme Court’s cases is that fee awards must be tied to evidence of the fee practices prevailing in the local legal market. See, e.g., Jenkins, 491 U.S. at 283, 285, 109 S.Ct. at 2469; Blum, 465 U.S. at 894-95, 104 S.Ct. at 1546-47; Delaware Valley II, 483 U.S. at 733, 107 S.Ct. at 3090-91 (O’Connor, J., concurring in part and concurring in the judgment); id. at 754, 107 S.Ct. at 3101-02 (Blackmun, J., dissenting). That was done in this case. In reaching its determination, the District Court expressly relied upon attorney affidavits filed by the plaintiff and an earlier decision in the same district in which another trial judge had found that “attorneys in the Washington[,] D.C.[,] community will only accept a fully contingent case if their recovery will be at least double their normal hourly billing rate and will only accept a partially contingent case if their recovery is enhanced by at least 50 percent.” See Palmer v. Shultz, 679 F.Supp. 68, 74 (D.D.C.1988), quoted in King v. Palmer, Civ. Action No. 83-1980, mem. op. at 4, 1988 WL 104970 (D.D.C. Sept. 20, 1988).
Another way to assess the reasonableness of the fee amount awarded by the District Court is pursuant to the two-prong test set forth in Justice O’Connor’s concurring opinion in Delaware Valley II. Although it is unclear what precedential force that opinion should be given, cf. Marks v. United States, 430 U.S. 188, 193, 97 S.Ct. 990, 993-94, 51 L.Ed.2d 260 (1977), it is nonetheless a source of some guidance and it is further indication that the judgment of the District Court should be affirmed in this case. Under the first prong of Justice O’Connor’s test, lower courts would be required to “treat a determination of how a particular market compensates for contingency as controlling future cases involving *792the same market.” 483 U.S. at 733, 107 S.Ct. at 3090 (O’Connor, J., concurring in part and concurring in the judgment). Under the second prong, “the fee applicant bears the burden of proving the degree to which the relevant market compensates for contingency.” Id. Under this second prong, “no enhancement for risk is appropriate unless the applicant can establish that without an adjustment for risk the prevailing party ‘would have faced substantial difficulties in finding counsel in the local or other relevant market.’ ” Id. (quoting from plurality opinion, 483 U.S. at 731, 107 S.Ct. at 3089).
Under the first prong of Justice O’Con-nor’s test, the District Court here reasonably concluded that “the relevant market” does in fact compensate lawyers for assuming the risk of contingent payment. In addition to the evidence cited in Palmer v. Shultz, which went directly to the degree to which the Washington, D.C., legal market customarily compensates for contingency, the court had before it several dozen affidavits from local attorneys swearing either that they generally demand an enhancement over normal hourly rates in order to accept contingent-fee cases or that they refuse such cases altogether because of the risk involved.
Under the second prong of Justice O’Connor’s test, Ms. King was required to show that she “would have faced substantial difficulties in finding counsel” had contingency enhancements not been customarily available in the Washington, D.C., legal market. This proposition, of course, turning as it does on a counterfactual supposition, is difficult to prove. Nonetheless, Ms. King produced an affidavit from her attorney stating that he would not have taken her case without the prospect of a fee enhancement. See Declaration of Robert M. Adler at 2 (Sept. 11, 1987), reprinted in Joint Appendix (“J.A.”) 74, 75. In addition, she produced further affidavits from several Washington, D.C., Title VII plaintiffs’ attorneys corroborating that Ms. King likely would have faced substantial difficulties securing counsel without the promise of a contingency premium.5 These attestations were reinforced by others documenting the general unwillingness of local attorneys to accept such cases absent some likelihood of receiving an enhancement for risk. On this record, we believe that Ms. King carried her burden under both prongs of Justice O’Connor’s test.
Because Title VII and the governing case law clearly permit trial courts to enhance attorney’s fees to compensate for the risk of contingent payment, and because the facts of the instant case satisfy whatever standards can be gleaned from recent Supreme Court cases, it cannot be found that the District Court abused its discretion in shaping the fee award in this case. As we noted at the outset, the standard of review in this case is abuse of discretion. The majority, however, has simply ignored the constraints of appellate review in second-guessing the findings of the trial judge. Indeed, the majority’s approach in this case borders on de novo review, in flat defiance of the Supreme Court’s instruction that “[i]t is central to the awarding of attorney’s fees ... that the district court judge, *793in his or her good judgment, make the assessment of what is a reasonable fee under the circumstances of the case.” Blanchard, 489 U.S. at 96, 109 S.Ct. at 946.
III.
The fundamental problem with this case, as with other fee cases, is that the Supreme Court has yet to give us coherent guidance about how to determine the reasonableness of fee awards under Title VII and similar fee-shifting statutes. In this, we share the majority’s frustration, if not its solution. In particular, we do not yet know under what circumstances to award a contingency enhancement, nor do we know whether the degree of enhancement allowed by the courts merely reflects prevailing market rates or actually creates the relevant market forces by defining the extent to which economic risk will be compensated.6 In our view, the proper answers to these questions ultimately must come from Congress.
Logically, of course, the size of a contingency enhancement should be determined in each case according to the degree of risk actually incurred by the prevailing attorney. It is impossible to determine with any confidence what a “reasonable fee” would be in any particular contingency case without first assessing just how much risk the plaintiff’s lawyer actually assumed.7 See, e.g., S. Speiser, supra, § 8:10, at 320; Berger, supra, at 326. In this regard, the Court’s apparent disapproval of case-by-case risk assessment, see Delaware Valley II, 483 U.S. at 731, 107 S.Ct. at 3089-90 (O’Connor, J., concurring in part and concurring in the judgment); id. at 745-46, 107 S.Ct. at 3097-98 (Blackmun, J., dissenting), serves only to frustrate the lower courts as they struggle to shape fee awards that, consistent with legislative purpose, will be “adequate to attract competent counsel” while stopping short of “producpng] windfalls” for plaintiffs' lawyers, see Blum, 465 U.S. at 897, 104 S.Ct. at 1548 (quoting S. Rep. No. 1011, 94th Cong., 2d Sess. 6 (1976)).
Although we share Justice O’Connor’s concern that risk enhancements not be calculated or awarded in “an arbitrary or unjust” manner, Delaware Valley II, 483 U.S. at 732, 107 S.Ct. at 3090 (O’Connor, J., concurring in part and concurring in the judgment), our experience suggests that the two-prong test enunciated in the concurring opinion in Delaware Valley II cannot be applied without difficulty. The panel opinions in this case and in McKenzie v. Kennickell, 875 F.2d 330 (D.C.Cir.1989), illustrate the two dilemmas inherent in this test. First, although the affidavits in both cases clearly establish that lawyers routinely demand the equivalent of a significantly higher hourly fee if they accept a case on a contingent-fee basis, those same *794affidavits also demonstrate that there is no one “market rate” or contingency fee arrangement for all such cases.8 Instead, lawyers predictably evaluate the case of each potential client independently and decide whether to accept representation and what to charge based on their estimation of the likelihood of success and the amount of fees they will receive if they do, in fact, prevail.9 Consequently, it is arguably unrealistic to assume that a court can determine one specific level of enhancement that the market demands in contingent-fee cases “as a class.”
Second, the so-called “substantial difficulties” test is a confusing and potentially mischievous requirement. For example, the requirements that the court determine the amount of enhancement by looking at contingency cases as a class, and that it look at the circumstances of the particular case only to determine if the plaintiff would have had substantial difficulty in obtaining counsel without the enhancement, force trial judges to award contingency enhancements on an all-or-nothing basis. The judge must either award the supposed “market” risk enhancement or no enhancement at all; there is no leeway for the court to decide that some enhancement is appropriate, but that it should be less than the previously determined “market” rate. Although the awards calculated by this method are concededly uniform, they may be nonsensical insofar as they overcompensate some plaintiffs with not-so-risky cases while inadequately compensating others with especially risky ones.
Consequently, we believe that district judges, in enhancing the lodestar to account for the risk of nonpayment, should have the same discretion that they have in determining other components of the reasonable fee. Although no one formula can be devised for all cases, we think that district courts should consider both the size of the attorney’s investment in the case and the likelihood that that investment would not be recouped. For one thing, in determining whether a contingency enhancement is necessary to enable the plaintiff to secure counsel, the court could consider the amount that the lawyer would lose if the plaintiff did not prevail. See Delaware Valley II, 483 U.S. at 747-48 & n. 12, 107 S.Ct. at 3098 & n. 12 (Blackmun, J., dissenting); Wildman v. Lerner Stores Corp., 771 F.2d 605, 613-14 (1st Cir.1985). The court thus might consider factors such as
what, if any, payment the attorneys would have received had the suit not been successful; what, if any, costs or expenses the attorneys would have incurred if the case had been lost; the extent to which the attorneys were required to compensate associates and to carry overhead expenses without assurance of compensation; and whether other attorneys refused to take the case because of the risk of nonpayment.
Delaware Valley II, 483 U.S. at 748 n. 12, 107 S.Ct. at 3098 n. 12 (Blackmun, J., dissenting). In our view, contingency enhancements are appropriate if the lawyer has to expend a substantial amount of her time and resources to litigate a case and there is a significant risk of not prevailing (and thus of not recouping her investment).
For another thing, the court could consider the plaintiffs likely “ability to prove liability and damages ... [and the legal precedent] either in favor of or against the theories put forth in the case” in order to determine the magnitude of the risk that the plaintiff would not prevail. See Wildman, 771 F.2d at 612. We recognize that *795it is impossible to determine such risk with any mathematical precision, and we are cognizant of the supposed ethical tensions such calculations may engender,10 but we do think that a district court judge, who is familiar with the factual and legal development of the case, can differentiate — after the fact — among cases based on the probability of success in the beginning. As the plurality in Delaware Valley II recognized, in some cases there is very little risk of not prevailing, and no enhancement for contingency, or only a very small one, should be awarded.11 In other cases, the risk of not prevailing would be substantial and the enhancement should be correspondingly enlarged.12
In the absence of further guidance in this area from Congress or the Supreme Court, however, we would adhere to the only secure legislative directive available to us — that assessments of what constitutes a “reasonable attorney’s fee” are to be left to the sound and reasoned discretion of trial judges. A decade ago, this court, sitting en banc, acknowledged the inherent lack of certainty in this enterprise but concluded that such determinations were nonetheless best left, as Congress intended, with trial judges:
To the district court judge falls the task of calculating as closely as possible a contingency adjustment with which fairly to compensate the successful attorney. We have not ... lost sight of the fact that this adjustment is inherently imprecise and that certain estimations must be made. For example, it is difficult in hindsight to determine the risk of failure at the commencement of a lawsuit that ultimately proved to be successful. Thus, we ask only that the district court judges exercise their discretion as conscientiously as possible, and state their reasons as clearly as possible.
Copeland, 641 F.2d at 893 (footnote omitted).13 We can perceive no reason to de*796part from that conclusion today. Because we discern nothing in the District Court’s judgment that would suggest an abuse of discretion justifying reversal, and nothing at all supporting the majority’s new legal rule, we dissent.
. See also Blanchard v. Bergeron, 489 U.S. 87, 96, 109 S.Ct. 939, 945-46, 103 L.Ed.2d 67 (1989) (“It is central to the awarding of attorney’s fees ... that the district court judge, in his or her good judgment, make the assessment of what is a reasonable fee under the circumstances of the case.”).
Although some of these precedents focus upon the parallel fee-shifting provision set out in the Civil Rights Attorney’s Fees Awards Act of 1976, Pub.L. No. 94-559, 90 Stat. 2641, codified at 42 U.S.C. § 1988 (1988), Congress and the Supreme Court have made clear that the fee-shifting provisions of that statute and Title VII are to be interpreted alike. See Hensley, 461 U.S. at 433 n. 7, 103 S.Ct. at 1939 n. 7; S.Rep. No. 1011, 94th Cong., 2d Sess. 4 (1976), U.S.Code Cong. & Admin.News 1976, p. 5908.
. The Court reiterated the presumptive reasonableness of the lodestar fee in Blanchard, but did so there only to rebut the suggestion that a fee arrangement set in a contingent-fee contract should govern as a strict ceiling on a court-ordered fee in the same case. See 489 U.S. at 95, 109 S.Ct. at 945.
. The majority’s extended discussion of whether the five votes that adopt this position constitute a binding majority of the Court seems to us overly pedantic, and mostly irrelevant. It is axiomatic that lower court judges routinely consider and weigh the diverse statements in Supreme Court opinions, especially those propositions garnering a majority, to seek guidance in the disposition of subsequent cases. Indeed, that is precisely what our sister circuits have done in construing Delaware Valley II.
.The Supreme Court acknowledged as much in Blanchard when it stated that the lodestar figure is to be derived by “applying prevailing billing rates to the hours reasonably expended on successful claims.” 489 U.S. at 94, 109 S.Ct. at 944 (emphasis added).
. In one affidavit, attorney George Chuzi stated:
During 1983,1 was personally familiar with most of the attorneys regularly bringing Title VII suits in the District of Columbia on behalf of plaintiffs. Had Mr. Adler not agreed to represent Mrs. King in this case, I am unaware of any other Title VII attorney who would have agreed in 1983 to represent her on a contingency fee basis (even had she agreed to pay up to $5,000 in legal fees). The only way in which I believe that a competent Title VII attorney would have been convinced to seriously consider this representation was if there was a reasonable possibility of receiving an enhanced fee for risk (over and above hourly rates) if Mrs. King prevailed.
Supplemental Declaration of George M. Chuzi at 2 (Dec. 21, 1987), reprinted in J.A. 130, 131; see abo Declaration of David R. Cashdan at 4 (Apr. 25, 1986) ("I believe that it is highly unlikely that I would have agreed to act as sole counsel in this case.”), reprinted in J.A. 112a, 112d; Declaration of Robert B. Fitzpatrick at 2 (Apr. 24, 1986) (“I believe that the chances of prevailing in this case ... were so remote that my firm would not have accepted representation of Mrs. King.”), reprinted in J.A. 173a, 173b; Declaration of Barry H. Gottfried at 2 (July 31, 1987) (“Had the plaintiff sought to retain me to represent her by filing a suit for the claims involved herein I do not believe that I would have accepted representation.”), reprinted in J.A. 176, 177.
. For example, as Judge Williams has pointed out, if courts generally allow a contingency enhancement of 50%, lawyers will have an economic incentive to bring only those cases in which the odds of succeeding on the merits are at least two-to-one; if the court-ordered enhancement figure rises to 100%, lawyers will presumably bring any case in which the chances of winning are at least 50%. See King v. Palmer, 906 F.2d 762, 770 (D.C.Cir.) (Williams, J., concurring in panel decision), vacated & reh’g en banc granted, 906 F.2d 772 (D.C.Cir. 1990).
. In this respect, a risk enhancement arguably should not encompass hours of labor for which compensation was secure. In the instant case, the argument could well be made that the risk of nonpayment incurred by Ms. King’s counsel virtually vanished after this court decided the merits in Ms. King’s favor in 1985. In King v. Palmer, 778 F.2d 878 (D.C.Cir.1985), this court reversed the District Court’s entry of judgment for the defendants and remanded the matter with instructions “to enter judgment for Ms. King and to determine an appropriate remedy." Id. at 882 (footnote omitted). "At a minimum,” we noted, “it appears that the appropriate remedy in this case should include the promotion of Ms. King ..., her receipt of back pay, and a full consideration of any further relief.” Id. at 882 n. 7. Once Ms. King prevailed on the merits of her main claim, the defendants recognize, she qualified for an award of fees for "hours reasonably expended [thereafter] on remedial and similar ancillary matters.” See Brief for Appel-lees/Cross-Appellants at 39. At least once it was established that there would be no review of the merits in the Supreme Court, there remained no risk of nonrecovery. Accordingly, the District Court probably could have, within the bounds of its discretion, eliminated from the risk-enhancement calculation fees for post-1985 work for which lodestar payment was certainly due.
. To the contrary, lawyers claimed to demand contingency enhancements from 33% to 300% of their regular hourly rates. Other lawyers did not specify a percentage, saying only that they needed a "reasonable” risk enhancement to accept a Title VII case on a contingent basis.
. As a result, the lawyer and client may tailor the terms of the contract to address the risks and potential rewards involved in each case. See, e.g., Declaration of Nora A. Bailey at 2 (Aug. 12, 1987), reprinted in J.A. 82, 83; Declaration of John R. Erickson at 2 (Nov. 16, 1987), reprinted in J.A. 168, 169; Declaration of Chester T. Kamin at 2-3 (Aug. 24, 1987), reprinted in J.A. 193, 194-95; Affidavit of David N. Webster at 6 (Sept. 20, 1982) (“The greater the uncertainty of result, the greater the percentage fee may be, albeit always within the limit of reasonableness.”), reprinted in J.A. 289, 294.
.The plurality summarized these concerns in Delaware Valley II:
[Evaluation of the risk of loss creates a potential conflict of interest between an attorney and his client, for in order to increase a fee award, a plaintiffs lawyer must expose all of the weaknesses and inconsistencies in his client’s case, and a defendant’s attorney must either concede the strength of the plaintiffs case in order to keep down the fee award, or "allo[w] the fee to be boosted by the contingency bonus [by] insisting that the plaintiffs victory was freakish.”
483 U.S. at 721-22, 107 S.Ct. at 3084-85 (quoting Leubsdorf, supra, at 483). The latter of these concerns is, of course, not an attorney-client conflict at all, but simply a strategic dilemma for the defendant of the sort that is not unknown in litigation. To the extent that the first represents a genuine potential for attorney-client conflict, we are reassured by the confidence the Supreme Court has previously expressed for lawyers’ ability to put their clients’ interests ahead of their own. See Evans v. Jeff D., 475 U.S. 717, 727-28, 106 S.Ct. 1531, 1537-38, 89 L.Ed.2d 747 (1986) (“Although respondents contend that Johnson, as counsel for the class, was faced with an ‘ethical dilemma’ when petitioners offered him relief greater than that which he could reasonably have expected to obtain for his clients at trial (if only he would stipulate to a waiver of the statutory fee award), and although we recognize Johnson’s conflicting interests between pursuing relief for the class and a fee for the Idaho Legal Aid Society, we do not believe that the ‘dilemma’ was an ‘ethical’ one in the sense that Johnson had to choose between conflicting duties under the prevailing norms of professional conduct. Plainly, Johnson had no ethical obligation to seek a statutory fee award. His ethical duty was to serve his clients loyally and competently.”) (footnote omitted; emphasis in original). If a lawyer can be trusted to serve his client faithfully at the cost of his entire statutory fee award, we feel sure he can be trusted to do so at the cost of some potential diminution of it.
. The plurality in Delaware Valley II voted to reverse the 50% risk enhancement of lodestar fees incurred to enforce a consent decree in part because there was not "a real risk of not persuading the District Court to enforce its own decree." 483 U.S. at 730, 107 S.Ct. at 3089.
. In saying that the district court should have the discretion to tailor the contingency enhancement to the particular case, we are not saying that the experience of other similarly situated plaintiffs can be disregarded. To the contrary, we think that the success rate of other plaintiffs who have filed suits based on similar legal theories is a good indicator of the risk presented in the case.
. The court further observed:
The setting of contingency adjustments is particularly within the expertise of the District Judge, As the Supreme Court said long ago, the District Court “has far better means of knowing what is just and reasonable than an appellate court can have.” Trustees v. *796Greenough, 105 U.S. 527, 537, 26 L.Ed. 1157 (1882).
Copeland, 641 F.2d at 893 n. 24.