with whom Justice Marshall and Justice Blackmun join, concurring in part and concurring in the judgment.
I agree that an award of attorney’s fees under the Equal Access to Justice Act (EAJA) was appropriate in this case, *575and I agree that the courts below did not adhere to the statutory hourly cap on fees. Therefore, I concur in the Court’s judgment affirming the decision to award fees and remanding for a new determination as to the amount. I disagree, however, with some of the Court’s reasoning. While I agree that appellate courts should review district court EAJA fee awards for abuse of discretion, in my view the Government may not prove that its position was “substantially justified” by showing that it was merely “reasonable.” Therefore, although I join Parts I, II, and IV of the Court’s opinion, I do not join Part III.1 Further, because I believe that the Court’s interpretation of the predicate showing for a party to obtain a fee award exceeding the statutory cap — that there existed “a special factor, such as the limited availability of qualified attorneys for the proceedings involved” — is stingier than Congress intended, I do not join Part V of the Court’s opinion.
I
Concerned that the Government, with its vast resources, could force citizens into acquiescing to adverse Government action, rather than vindicating their rights, simply by threatening them with costly litigation, Congress enacted the EAJA, waiving the United States’ sovereign and general statutory immunity to fee awards and creating a limited exception to the “American Rule” against awarding attorneys fees to prevailing parties. S. Rep. No. 96-253, pp. 1-6 (1979) (S. Rep.). Consequently, when a qualified party (as defined in the Act) prevails against the United States in an adversarial proceeding not. sounding in tort, the EAJA prescribes that “a court shall award . . . fees and other expenses . . . unless the court finds that the position of the United States was substantially justified or that special circumstances make an award unjust.” 28 U. S. C. § 2412(d)(1)(A). *576In this, our first EAJA case, we are called upon to consider the phrase “substantially justified.”
The Court begins, as is proper, with the plain meaning of the statutory language. The Court points out that “substantially” is not a word of precise and singular definition. Indeed, the word bears two arguably different relevant definitions: “‘considerable in amount, value, or the like; large’”; and “‘in substance or in the main.’” Ante, at 564. See also Webster’s Third New International Dictionary 2280 (1976) (“considerable in amount, value, or worth”; and “having a solid or firm foundation . . . being that specified to a large degree or in the main”). The Court concludes, and I agree, that, to the extent they are different, Congress intended the latter meaning.
Unfortunately, the Court feels duty bound to go beyond the words enacted by Congress and to fashion its own substitute phrase using what it perceives to be a more legally precise term. The test upon which the Court alights is initially the “‘reasonable basis both in law and fact’” standard, adopted by the courts below. Ante, at 565. While this phrase is often mentioned in the legislative history as the explication of “substantially justified,” this alternative phraseology is inherently no more precise than the statutory language. In fact, it may be less so, for the Court equates it with “the test of reasonableness,” ante, at 568, a standard rejected by Congress and significantly more forgiving than the one actually adopted.
The Senate Judiciary Committee considered and rejected an amendment substituting the phrase “reasonably justified” for “substantially justified.” S. Rep., at 8. Clearly, then, the Committee did not equate “reasonable” and “substantial”; on the contrary, it understood the two terms to embrace different burdens. “Reasonable” has a variety of connotations, but may be defined as “not absurd” or “not ridiculous.” Webster's New Third International Dictionary 1892 (1976). Even at its strongest, the term implies a posi*577tion of some, but not necessarily much, merit. However, as we have seen, “substantial” has a very different definition: “in substance or in the main.” Thus, the word connotes a solid position, or a position with a firm foundation. While it is true “reasonable” and “substantial” overlap somewhat (substantial at its weakest and reasonable at its strongest) an overlap is not an identity. Therefore, although Congress may well have intended to use “substantial” in its weaker sense, there is no reason to believe, and substantial reason to disbelieve (as I will discuss below), that Congress intended the word to mean “reasonable” in its weaker sense.
The underlying problem with the Court’s methodology is that it uses words or terms with similar, but not identical, meanings as a substitute standard, rather than as an aid in choosing among the assertedly different meanings of the statutory language. Thus, instead of relying on the legislative history and other tools of interpretation to help resolve the ambiguity in the word “substantial,” the Court uses those tools essentially to jettison the phrase crafted by Congress. This point is well illustrated by the Government’s position in this case. Not content with the term “substantially justified,” the Government asks us to hold that it may avoid fees if its position was “reasonable.” Not satisfied even with that substitution, we are asked to hold that a position is “reasonable” if “it has some substance and a fair possibility of success.” Brief for Petitioner 13. While each of the Government’s successive definitions may not stray too far from the one before, the end product is significantly removed from “substantially justified.” I believe that Congress intended the E-AJA to do more than award fees where the Government’s position was one having no substance, or only a slight possibility of success; I would hope that the Government rarely engages in litigation fitting that definition, and surely not often enough to warrant' the $10(1 million in attorney’s fees Congress expected to spend over the original EAJA’s 5-year life.
*578My view that “substantially justified” means more than merely reasonable, aside from conforming to the words Congress actually chose, is bolstered by the EAJA’s legislative history. The phrase “substantially justified” was a congressional attempt to fashion a “middle ground” between an earlier, unsuccessful proposal to award fees in all cases in which the Government did not prevail, and the Department of Justice’s proposal to award fees only when the Government’s position was “arbitrary, frivolous, unreasonable, or groundless.” S. Rep., at 2-3. Far from occupying the middle ground, “the test of reasonableness” is firmly encamped near the position espoused by the Justice Department. Moreover, the 1985 House Committee Report pertaining to the EAJA’s reenactment expressly states that “substantially justified” means more than “mere reasonableness.” H. R. Rep. No. 99-120, p. 9 (1985). Although I agree with the Court that this Report is not dispositive, the Committee’s unequivocal rejection of a pure “reasonableness” standard in the course of considering the bill reenacting the EAJA is deserving of some weight.
Finally, however lopsided the weight of authority in the lower courts over the meaning of “substantially justified” might once have been, lower court opinions are no longer nearly unanimous. The District of Columbia, Third, Eighth, and Federal Circuits have all adopted a standard higher than mere reasonableness, and the Sixth Circuit is considering the question en banc. See Riddle v. Secretary of Health and Human Services, 817 F. 2d 1238 (CA6) (adopting a higher standard), vacated for rehearing en banc, 823 F. 2d 164 (1987); Lee v. Johnson, 799 F. 2d 31 (CA3 1986); United States v. 1,378.65 Acres of Land, 794 F. 2d 1313 (CA8 1986); Gavette v. OPM, 785 F. 2d 1568 (CA Fed. 1986) (en banc); Spencer v. NLRB, 229 U. S. App. D. C. 225, 712 F. 2d 539 (1983).
In sum, the Court’s journey from “substantially justified” to “reasonable basis both in law and -fact” to “the test of *579reasonableness” does not crystallize the law, nor is it true to Congress’ intent. Instead, it allows the Government to creep the standard towards “having some substance and a fair possibility of success,” a position I believe Congress intentionally avoided. In my view, we should hold that the Government can avoid fees only where it makes a clear showing that its position had a solid basis (as opposed to a . marginal basis or a not unreasonable basis) in both law and fact. That it may be less “anchored” than “the test of reasonableness,” a debatable proposition, is no excuse to abandon the test Congress enacted.2
II
1 also disagree with the Court’s discussion of the circumstances supporting a fee enhancement beyond the $75-per-hour (adjusted for inflation) cap set by Congress, although I do agree that the lower courts’ judgment in this regard cannot stand. The statute states that courts may not award fees in excess of this cap unless “a special factor, such as the limited availability of qualified attorneys for the proceedings involved, justifies the higher fee.” 28 U. S. C. §2412(d)(2)(A)(ii). The District Court found that there was a limited availability of qualified attorneys here, and also that there were additional special factors warranting an increase. In so deciding, however, the District Court’s and Court of Appeals’ analyses erroneously mirrored the analysis under 42 U. S. C. § 1988, a fee-shifting statute without an hourly rate limitation. Congress clearly meant to contain the potential costs of the EAJA by limiting the hourly rate of attorneys where fees are awarded. Consequently, a consideration of factors like counsel’s customary rate, while perfectly appropriate under § 1988, cannot justify exceeding the EAJA cap. To hold otherwise would render the cap nothing more than *580advisory despite Congress’ expressed intent to permit higher awards only in rare cases.
That said, our job is to decide the meaning of the term: “‘a special factor, such as the limited availability of qualified attorneys.” The Court begins with the single expressed special factor, the “limited availability of qualified attorneys. ” It holds that this phrase refers to an attorney with a required, articulable specialization, and does not refer to the limited availability of attorneys experienced or skilled enough to handle the proceedings involved. The Court reasons that allowing an enhancement for extraordinary skill or experience, even if required, would render the cap nugatory, since those factors merely set the market rate. This tidy analysis is too simplistic.
The most striking aspect of the Court’s holding in this regard is its willingness to ignore the plain meaning and language of the exception. After all, in the rare EAJA case where highly experienced attorneys are truly required, a neophyte lawyer is no more “qualified ... for the proceedings involved” than a nonpatent lawyer is to handle a patent case. The Court’s interpretation might nonetheless be appropriate if the cap would otherwise be actually rendéred meaningless, but that is not the case here. First, we must keep in mind the nature of the cases Congress envisioned would result in a fee award: those in which the Government’s position was not “substantially justified.” This observation takes much of the force from the Court’s reasoning, as it will be a rare case in which an attorney of exceptional skill is necessary and where the Government’s position was weak enough to warrant an EAJA award.
Second, the phrase “limited availability of qualified attorneys,” read in conjunction with “special factor,” reflects a congressional judgment that if the price of lawyers generally exceeds the cap, that trend alone will not justify an increase. Therefore, awarding an enhancement in cases where extráor-*581dinary experience or skill is required does not write the cap out of the statute.
Third, the Court’s economic analysis assumes that the market price for services rendered will always be precisely known, an assumption I cannot share, and one that there is no reason to believe Congress shared. A “reasonable” hourly rate cannot be determined with exactitude according to some preset formulation accounting for the nature and complexity of every type of case. Therefore, courts often assume that an attorney’s normal hourly rate is reasonable, or, in the case of public interest counsel, a reasonable rate is generally the rate charged by an attorney of like “skill, experience, and reputation.” Blum v. Stenson, 465 U. S. 886, 895, n. 11 (1984). Certainly adjustments up or down are appropriate where the fee charged is out of line with the nature of the services rendered. However, such adjustments are often difficult to make given that the “prevailing market rate” is determined by reference to the particular attorney involved rather than to a minimally qualified hypothetical lawyer, ibid., and that the fee determination should not become a “second major litigation.” Hensley v. Eckerhart, 461 U. S. 424, 437 (1983). Moreover, to some extent, even in a simple case higher hourly rates may be offset by fewer hours billed due to counsel's greater efficiency. Absent the statutory cap, these factors would be used in an EAJA analysis as extensively as they are used in a § 1988 analysis. However, a showing that the particular attorney retained normally charges more than the statutory cap will., by itself, avail a fee applicant nothing under EAJA, although it may, by itself, be dispositive under § 1988.
Therefore, the Court is simply wrong when it asserts that if we allow a showing of extraordinary skill or experience (in the rare case where it is required) to justify an enhanced award, then the cap will be rendered meaningless. Far from it. The same logic supporting a “patent lawyer” exception— that when only a fraction of the bar is qualified to handle a *582case, those attorneys may charge a premium for their services — supports an enhancement for skill or experience.
Equally troubling is the Court’s requirement that a “special factor” must not be “of broad and general application:” Ante, at 573. We are given no explanation of or for this limitation, beyond the declaration that it is necessary to preserve the efficacy of the cap. Further, while the Court is willing to say what is not a special factor — everything relied upon below — we are given no example of anything that is a special factor other than the subject-matter specialization already considered as falling within the “limited availability of qualified attorneys for the proceedings involved” example. Having rejected the lower courts’ list of factors in its entirety, it seems as if the Court leaves nothing remáining.
Such a strained interpretation, apparently reading the words “such as” out of the Act, is unnecessary. See Vibra-Tech Engineers, Inc. v. United States, 787 F. 2d 1416 (CA10 1986); Action on Smoking and Health, v. CAB, 233 U. S. App. D. C. 79, 724 F. 2d 211 (1984). Cf. Kungys v. United States, 485 U. S. 759, 778 (1988) (“[N]o provision [of a statute] should be construed to be entirely redundant”). A “special factor” may be readily analogized to the factors we identified in Blum to enhance the lodestar figure under § 1988. In Blum, we held that the lodestar amount (the reasonable hourly rate multiplied by the number of hours billed) is “presumably” the reasonable fee. However, we also held that an upward adjustment may be appropriate “in the rare case where the fee applicant offers specific evidence to show that the quality of service rendered was superior to that one reasonably should expect in light of the hourly rates charged and that the success was exceptional.” 465 U. S., at 899 (internal quotations omitted).3 Analogizing to the EAJA *583context, the lodestar would be calculated by multiplying the reasonable rate (as capped) by the number of hours billed. That amount would presumably be the proper award. However, where a factor exists that would justify an enhancement of the lodestar amount under § 1988, an enhancement of the EAJA award might also be appropriate. Unlike the lower courts’ approach, this rule would not read the cap out of the statute, for as we predicted in Blum, a lodestar enhancement would be appropriate only in “the rare case.”
Although the Blum enhancers constitute more than the situation where there is a limited availability of qualified counsel, the statute expressly allows more to be considered. The Court’s miserly refusal to accede to this statutory command is unjustified and unwarranted. I therefore concur only in the judgment as to the fee calculation.
Because the purposes of the EAJA are different from those of Federal Rule of Civil Procedure 37 and from those served by the “substantial evidence” test used to review agency determinations, I believe the meanings given the term “substantial” in those contexts do not govern here.
We left open whether the contingent nature of the fee could also justify an enhancement. However, much for the reasons stated by the Court, that question is not'pertinent to an EAJA case. It is one thing to say that a contingent-fee enhancement is necessary to compensate an attorney *583when victory is uncertain, it is another thing entirely to say that such an enhancement is necessary to compensate an attorney when the lack of substantial justification is uncertain.