DYK, Circuit Judge.

This case presents the question of when a district court may reduce the "lodestar" calculation of reasonable attorneys' fees to account for the "amount involved and results obtained" or other factors. Although the district court here did an exemplary job, we conclude that two errors require a remand. First, while the district court may reduce the lodestar figure to account for the "amount involved and results obtained" and other factors in rare and exceptional circumstances, we conclude that the district court erred here by taking these factors into account after calculating the lodestar figure, rather than as a part of the lodestar calculation itself. We also hold that the district court should have used forum rates in determining the reasonable hourly rate for the lodestar calculation. Accordingly, we vacate and remand for further proceedings consistent with this opinion.

BACKGROUND

The United States has waived its sovereign immunity with respect to constitutional claims, including government takings claims arising under the Fifth Amendment. See 28 U.S.C. §§ 1346(a)(2), 1491(a)(1). The United States Court of Federal Claims has exclusive jurisdiction over such claims where the amount in controversy is greater than $10,000, § 1491(a)(1) (the "Tucker Act"), but shares jurisdiction with the district courts where the amount in controversy does not exceed $10,000, § 1346(a)(2) (the "Little Tucker Act"). In actions brought under the Tucker Act or the Little Tucker Act in which a plaintiff is awarded compensation for the taking of property, the Uniform Relocation Assistance and Real Property Acquisition Policies Act of 1970 ("URA") provides for the recovery of "such sum as will in the opinion of the court or the Attorney General reimburse such plaintiff for his reasonable costs, disbursements, and expenses, including reasonable attorney ... fees, actually incurred because of such proceeding." 42 U.S.C. § 4654(c).1

This case involves takings compensation claims brought by appellants against the United States. On May 23, 2000, following the transfer of their compensation claims to the United States District Court for the Eastern District of Texas, Plaintiff-Appellant Ashburn Bywaters and other named plaintiffs (collectively, "appellants"), represented by counsel based in Washington, DC, filed an amended class action complaint on behalf of themselves and all others similarly situated, alleging that they were the owners of interests in land constituting part of a railroad corridor (the "Chaparral rail corridor") that had been converted for trail use by the Interstate Commerce Commission pursuant to the National Trails System Act ("Trails Act"), 16 U.S.C. § 1247(d).2 The Trails Act is designed to preserve railroad rights-of-way by converting them into recreational trails. Actions by the government pursuant to the Trails Act can result in takings liability where the railroad acquired an easement from the property owner, the railroad's use of the property ceased, and the government's action under the Trails Act prevented reversion of the property to the original owner. See Preseault v. United States, 100 F.3d 1525, 1550-52 (Fed.Cir.1996) (en banc); see also Caldwell v. United States, 391 F.3d 1226, 1228 (Fed.Cir.2004).

On August 25, 2000, the district court certified a class consisting of all persons who owned an interest in land constituting the Chaparral rail corridor extending from Farmersville, Texas, to Paris, Texas, that was converted to trail use pursuant to the Trails Act, and whose claims did not exceed $10,000 per claim. On April 17, 2003, the government stipulated to takings liability with respect to those claims for segments of the Chaparral rail corridor in which the railroad acquired only an easement.3 From 2003 to 2009, the parties cooperated to determine the amount of just compensation to be paid to the members of the class.

On July 31, 2009, the parties proposed a settlement agreement that resolved all issues in the case, except for the amount of attorneys' fees and costs to be awarded under the URA. The district court approved the proposed settlement after finding that the proposed settlement would secure 100% of the just compensation due to class members with eligible claims, subject to the $10,000 jurisdictional cap of the Little Tucker Act. Under the settlement, appellants' total recovery was $1,241,385.36, including pre-judgment interest.

Following settlement, appellants filed a claim for attorneys' fees under the URA, requesting attorneys' fees in the amount of $832,674.99, which included 2,119.69 hours of work from August 1999, when the case was transferred to the Eastern District of Texas, to December 2009. Appellants also urged the district court to determine the appropriate amount of attorneys' fees by applying market rates for the District of Columbia, where appellants' counsel practiced, rather than rates charged by attorneys in the forum where the case was brought (the Eastern District of Texas). In response, the government argued for application of the forum rule. The government also argued for the reduction of the fees claimed based on various grounds, including that the hours claimed were unreasonable in light of the government's stipulation to liability early in the case, and the fact that a fee agreement between appellants and their counsel provided for the award of attorneys' fees calculated at the greater of counsel's regular hourly rate or one third of appellants' total recovery.

The district court, applying Federal Circuit law, determined the amount of attorneys' fees to be awarded under the "lodestar" approach, i.e., by multiplying the number of hours reasonably expended by a reasonable hourly rate. In determining the lodestar figure, the district court first considered the hours requested by appellants and the government's objections and determined that only 18.2 hours spent drafting and filing an amicus brief were unreasonable. Accordingly, the court reduced the amount of hours requested by appellants by 18.2 hours. The court next determined that the relevant market for determining the reasonable hourly rate was the District of Columbia and applied the Updated Laffey Matrix4 to determine the reasonable hourly rates for complex litigation. Accordingly, the district court determined that "multiplying the number of hours reasonably expended by the reasonable hourly rate using the Updated Laffey Matrix" yielded a lodestar figure of $826,044.19. Bywaters v. United States, No. 6:99-CV-451, 2010 WL 3212124, at *4 (E.D.Tex. Aug. 12, 2010) (hereinafter "Attorneys' Fees Order").

However, calculation of the lodestar figure did not end the district court's inquiry. The district court found that the factor of "amount involved and results obtained" was not adequately taken into account in determining a reasonable fee. The district court reasoned that the lodestar figure would yield an award of attorneys' fees that was 66.5% of the total relief awarded to appellants, which was "extremely high considering the amount at stake in this case and the actual results obtained." Id. The district court observed that "[o]verlooking the large disparity between Plaintiffs' award and the lodestar figure would only encourage protracted litigation." Id. The court also concluded that the work performed by appellants' counsel was "administrative in nature and did not require a high level of legal skill." Id. Finally, the court noted that the lodestar figure exceeded the amount calculated under the contingent-fee option in the fee agreement between appellants and their counsel, which provided that appellants' counsel would be compensated at either "the value of [their] professional services at [their] regular hourly rate or by multiplying by one third the amount recovered for the plaintiff class as damages, whichever is greater." Id. The court reduced the calculated lodestar figure by 50%, awarding attorneys' fees in the amount of $413,022.10.

Appellants timely appealed the district court's award of attorneys' fees. We have jurisdiction pursuant to 28 U.S.C. § 1295(a)(2).

DISCUSSION

Generally, our legal system adheres to the "American Rule" under which "each party in a lawsuit ordinarily shall bear its own attorney's fees." Hensley v. Eckerhart, 461 U.S. 424, 429, 103 S.Ct. 1933, 76 L.Ed.2d 40 (1983). However, in certain categories of cases Congress has carved out exceptions to the American Rule and allowed for recovery of attorneys' fees. See Pennsylvania v. Del. Valley Citizens' Council for Clean Air, 478 U.S. 546, 561-62, 106 S.Ct. 3088, 92 L.Ed.2d 439 (1986). The fee-shifting provisions of the URA are one such example. See 42 U.S.C. § 4654. We have not yet had an occasion to interpret the fee-shifting provisions of the URA.

I

As a threshold matter we must first determine whether, in calculating the amount of reasonable attorneys' fees under the URA, we should apply our law or the law of the regional circuit—here, the Fifth Circuit. Notwithstanding this court's exclusive jurisdiction over Tucker Act and Little Tucker Act appeals, see 28 U.S.C. § 1295(a)(2), the government contends that we should apply Fifth Circuit law in reviewing the district court's grant of attorneys' fees under the URA. Specifically, the government argues that the district court's award of attorneys' fees is "entirely dependent on Fed.R.Civ.P. 23" and thus implicates procedural issues, rather than the merits of a takings claim under the Tucker Act. Appellee's Br. 41. The district court's award of attorneys' fees in this case was quite clearly based upon the mandatory fee-shifting provision of the URA. Compare Fed.R.Civ.P. 23(h) ("[T]he court may award reasonable attorney's fees...." (emphasis added)), with 42 U.S.C. § 4654(c) ("The court ... shall determine and award ... reasonable attorney... fees...." (emphasis added)). The award of fees thus depends on construction of the URA and not Rule 23.

Additionally, the government argues that because we do not have exclusive jurisdiction over all claims arising under the URA generally, Federal Circuit law should not apply. The URA provides for the award of "reasonable" attorneys' fees in two separate circumstances. First, attorneys' fees may be awarded where the government initiates a condemnation proceeding that results in either a final judgment that the government may not acquire the property by condemnation or abandonment of the proceeding by the government. § 4654(a). Such cases are litigated in the district courts and appealed to the regional circuits. See, e.g., United States v. 122.00 Acres of Land, 856 F.2d 56, 58-59 (8th Cir.1988) (reviewing a district court's award of fees pursuant to section 4654(a)). Second, attorneys' fees may also be awarded where, as in this case, a property owner brings an inverse condemnation action under the Tucker Act or the Little Tucker Act alleging a government taking under the Fifth Amendment and that action results in an award of compensation for the taking. § 4654(c). We have exclusive appellate jurisdiction in such cases. See 28 U.S.C. § 1295(a)(2). We are concerned here only with the second provision, section 4654(c).

While we do not have exclusive jurisdiction in all cases arising under section 4654 of the URA, the fee-shifting provision at issue here—section 4654(c)—is applicable only to government takings claims brought under the Tucker Act or the Little Tucker Act, cases that are within our exclusive jurisdiction. In Heisig v. United States, 719 F.2d 1153 (Fed.Cir.1983), we held that district courts adjudicating claims under the Little Tucker Act should apply the law of the Federal Circuit, rather than regional circuit law. We noted that "[l]ogic, as well as the express congressional desire for uniformity, dictate that similar standards of review and the precedents of this circuit should obtain in a proceeding in a district court that is substantially identical, except for jurisdictional amount, to one in the Claims Court." Id. at 1156; see also United States v. Hohri, 482 U.S. 64, 71, 107 S.Ct. 2246, 96 L.Ed.2d 51 (1987) ("A motivating concern of Congress in creating the Federal Circuit was the special need for nationwide uniformity in certain areas of the law." (internal quotation marks omitted)). Furthermore, we have consistently applied our law to claims for attorneys' fees under section 285 of the Patent Act because section 285 relates to an area of substantive law within our exclusive jurisdiction. See, e.g., Q-Pharma, Inc. v. Andrew Jergens Co., 360 F.3d 1295, 1299 (Fed.Cir.2004); see also 35 U.S.C. § 285. Here too attorneys' fees were awarded pursuant to a statutory fee-shifting provision that relates only to cases brought pursuant to the Tucker Act and the Little Tucker Act, an area within our exclusive jurisdiction. In light of "the evident congressional desire for uniform adjudication of Little Tucker Act claims" and Tucker Act claims, Hohri, 482 U.S. at 73, 107 S.Ct. 2246, we hold that our law, rather than the law of the regional circuit, should apply to an award of attorneys' fees under section 4654(c).

II

While we have not yet interpreted section 4654(c), the Supreme Court has advised that all federal fee-shifting statutes calling for an award of "reasonable" attorneys' fee should be construed "uniformly." City of Burlington v. Dague, 505 U.S. 557, 562, 112 S.Ct. 2638, 120 L.Ed.2d 449 (1992); see also Indep. Fed'n of Flight Attendants v. Zipes, 491 U.S. 754, 758 n. 2, 109 S.Ct. 2732, 105 L.Ed.2d 639 (1989) ("We have stated in the past that fee-shifting statutes' similar language is `a strong indication' that they are to be interpreted alike." (quoting Northcross v. Memphis Bd. of Educ., 412 U.S. 427, 428, 93 S.Ct. 2201, 37 L.Ed.2d 48 (1973))); Hubbard v. United States, 480 F.3d 1327, 1333 (Fed.Cir.2007). Nothing in the language or legislative history of the URA suggests that it should receive a different construction than other fee-shifting statutes.

Generally, in determining the amount of reasonable attorneys' fees to award under federal fee-shifting statutes, the district court is afforded considerable discretion. See Hensley, 461 U.S. at 437, 103 S.Ct. 1933; see also 42 U.S.C. § 4654(c) (providing for an award of attorney fees that will "in the opinion of the court" reimburse plaintiffs for reasonable expenses actually incurred (emphasis added)). This deference results from "the district court's superior understanding of the litigation and the desirability of avoiding frequent appellate review of what essentially are factual matters." Hensley, 461 U.S. at 437, 103 S.Ct. 1933. In this case, the district court carefully and thoughtfully considered the submissions of the parties, including over forty pages of billing records submitted by appellants in support of their fee application. In calculating the lodestar figure and subsequently reducing that figure, the district court candidly acknowledged the reasons for its decision. It may well be that the amount awarded by the district court will turn out to be the correct amount. While we think the district court's approach was largely correct, we think a remand is nonetheless required because the district court's analysis was incorrect in two respects. First, the district court should have considered the "amount involved and results obtained" as well as the administrative nature of the work and the fee agreement in determining the lodestar figure, rather than applying these factors after calculation of the lodestar figure. Second, the district court was required to apply the forum rule in determining the reasonable hourly rate for the relevant market.

III

We first consider the district court's adjustment to the lodestar figure. In determining the amount of reasonable attorneys' fees under federal fee-shifting statutes, the Supreme Court has consistently upheld the lodestar calculation as the "guiding light of [its] fee-shifting jurisprudence." Perdue v. Kenny A. ex rel. Winn, ___ U.S. ___, 130 S.Ct. 1662, 1672, 176 L.Ed.2d 494 (2010) (quoting Gisbrecht v. Barnhart, 535 U.S. 789, 801, 122 S.Ct. 1817, 152 L.Ed.2d 996 (2002)). Although there is a "strong presumption" that the lodestar figure represents a "reasonable" attorney fee, Dague, 505 U.S. at 562, 112 S.Ct. 2638, the Supreme Court has recognized a district court's discretion to adjust the lodestar figure "upward or downward" based upon other considerations, Del. Valley, 478 U.S. at 564, 106 S.Ct. 3088 (quoting Hensley, 461 U.S. at 434, 103 S.Ct. 1933). However, adjustments to the lodestar figure "are proper only in certain `rare' and `exceptional' cases, supported by both `specific evidence' on the record and detailed findings by the lower courts." Id. at 565, 106 S.Ct. 3088; see also Perdue, 130 S.Ct. at 1673 (reaffirming that enhancements to the lodestar figure may be awarded in only "rare" and "exceptional" circumstances). Adjustments are warranted only where the lodestar figure fails to take into account a relevant consideration. As the Supreme Court recently stated, "an enhancement may not be awarded based on a factor that is subsumed in the lodestar calculation." Perdue, 130 S.Ct. at 1673 (citations omitted). The question is whether the "amount involved and results obtained" in this case warranted an adjustment.

We note initially that the Supreme Court has not always been clear about what is encompassed within the category of "amount involved and results obtained"—that is, whether it refers to the absolute level of success or the proportionate level of success (percentage of recovery on the initial claim). In truth, even though this case involves an adjustment for the absolute level of success, it seems to make little difference in the mandated approach. As the Supreme Court standards have evolved, neither an adjustment for the absolute or proportionate level of success is appropriate absent unusual circumstances. The "amount involved and results obtained" factor was first identified as relevant to the attorney fee inquiry as one of twelve factors—the so-called "Johnson factors"—considered by the Fifth Circuit in Johnson v. Georgia Highway Express, Inc., 488 F.2d 714 (5th Cir.1974).5 In 1983, citing Johnson, the Court initially opined in Hensley that the district court could, in its discretion, adjust the lodestar figure "upward or downward" to account for the "crucial" factor of the "results obtained." 461 U.S. at 434, 103 S.Ct. 1933. Specifically, the Court noted that in considering this factor, the district court should "focus on the significance of the overall relief obtained by the plaintiff in relation to the hours reasonably expended on the litigation," but that there was "no precise rule or formula" for taking this factor into consideration. Id. at 435-36, 103 S.Ct. 1933.

In the years since Hensley, the Supreme Court's view on the degree of discretion afforded district courts in adjusting the lodestar figure has undergone change, thus cabining the district court's ability to adjust the lodestar figure to only "rare" and "exceptional" cases. See Perdue, 130 S.Ct. at 1673. Later cases have made clear that while the "amount involved and results obtained" remains a factor to be considered in determining a reasonable attorney fee, it cannot be a basis for reducing the lodestar figure where it could have been taken into account in calculating the lodestar figure in the first instance. In Blum v. Stenson, 465 U.S. 886, 104 S.Ct. 1541, 79 L.Ed.2d 891 (1984), decided just one year after Hensley, the Court clarified that while "there may be circumstances in which the basic standard of reasonable rates multiplied by reasonably expended hours results in a fee that is either unreasonably low or unreasonably high," the lodestar figure is "presumed" to be reasonable. Id. at 897, 104 S.Ct. 1541. The Court also cautioned against "double counting" factors such as the "amount involved and results obtained" by adjusting the lodestar figure where those factors are fully reflected in the reasonable hourly rate of the attorneys and the reasonable number of hours expended. Id. at 899-900, 104 S.Ct. 1541; see also Dague, 505 U.S. at 562-63, 112 S.Ct. 2638. In particular, the Court noted that "[b]ecause acknowledgment of the `results obtained' generally will be subsumed within other factors used to calculate a reasonable fee, it normally should not provide an independent basis for increasing the fee award." Blum, 465 U.S. at 900, 104 S.Ct. 1541; see also City of Riverside v. Rivera, 477 U.S. 561, 568-69, 106 S.Ct. 2686, 91 L.Ed.2d 466 (1986) (plurality opinion) (recognizing that while the "amount involved and results obtained" may be considered in determining a reasonable attorney fee, a district court is not free to mechanically adjust the lodestar figure downward based on this factor).

Most recently, the Court considered the principles governing the district court's authority to adjust the lodestar figure in Perdue, 130 S.Ct. 1662. In Perdue, the Court once again endorsed the lodestar method, noting that it is "readily administrable; and unlike the Johnson approach, the lodestar calculation is `objective,' and thus cabins the discretion of trial judges, permits meaningful judicial review, and produces reasonably predictable results." Id. at 1672 (internal citations omitted). Quoting Blum with approval, the Court also held that as a rule, the lodestar figure should only be adjusted in "rare" and "exceptional" cases and may not be adjusted "based on a factor that is subsumed in the lodestar calculation." Id. at 1673. A district court seeking to adjust the lodestar figure must justify its deviation with "specific evidence" demonstrating that the factors considered are not adequately subsumed within the lodestar calculation. See Blum, 465 U.S. at 898-900, 104 S.Ct. 1541; see also Perdue, 130 S.Ct. at 1676 ("It is essential that the judge provide a reasonably specific explanation for all aspects of a fee determination, including any award of an enhancement."); Del. Valley, 478 U.S. at 565, 106 S.Ct. 3088 (noting that modifications of the lodestar figure should be supported by "detailed findings" by the lower court). In Perdue, the Court held that the upward adjustment for "results obtained" was not permissible. 130 S.Ct. at 1676. We see no basis for distinguishing between an upward adjustment and a downward adjustment for "results obtained." Neither is permissible absent unusual circumstances.

Applying the standards set forth in Hensley and later cases, we find that this case does not present the sort of "rare" and "exceptional" circumstance where the factor of "amount involved and results obtained" should be considered as a basis for departure from the lodestar figure. The mere fact that the recovery is small in amount is not a circumstance justifying a reduced fee award. See Millea v. Metro-North R.R. Co., 658 F.3d 154, 168 (2d Cir.2011). The district court's conclusory statement that the "`amount involved and results obtained' factor [was] not adequately taken into account" in determining a reasonable fee, Attorneys' Fees Order, 2010 WL 3212124, at *4, is also not sufficient, standing alone, to support a departure from the lodestar figure. While it is legitimate to consider the "amount involved and results obtained" in determining a reasonable attorney fee award, the district court should have considered this factor when determining the reasonable number of hours expended and the reasonable hourly rates of the attorneys. It is axiomatic that attorneys almost inevitably consider the amount involved in a particular case when determining a reasonable number of hours to expend on any given issue or when allocating personnel resources based upon the expertise or experience required. Where only a small amount is at stake, it certainly would not be reasonable to expend countless hours on such a small claim or to commit the most experienced or valued attorney in the firm to work on the case. Thus where the amount involved is small, reductions in the reasonable number of hours expended or the reasonable hourly rate can easily be made to reflect this fact. It is for this reason that the Supreme Court has held that the "results obtained" factor is generally subsumed within the lodestar calculation and thus normally should not provide an independent basis for a departure from the lodestar figure. Blum, 465 U.S. at 900, 104 S.Ct. 1541; see also Perdue, 130 S.Ct. at 1674.

Just as the "amount involved and results obtained" can readily be incorporated into the lodestar figure, so too can the administrative nature of the work and the low level of skill involved, which the district court identified as alternative bases for reducing the lodestar figure.6 The district court's findings with respect to these factors can be fairly reflected by reducing the number of hours reasonably expended and the appropriate hourly rates of the attorneys. See Del. Valley, 478 U.S. at 566, 106 S.Ct. 3088 ("Because considerations concerning the quality of a prevailing party's counsel's representation normally are reflected in the reasonable hourly rate, the overall quality of performance ordinarily should not be used to adjust the lodestar...."); Blum, 465 U.S. at 898-99, 104 S.Ct. 1541 ("The novelty and complexity of the issues presumably were fully reflected in the number of billable hours recorded by counsel and thus do not warrant an ... adjustment.... Neither complexity nor novelty of the issues, therefore, is an appropriate factor in determining whether to increase the basic fee award.").

Finally, the fee agreement between appellants and their counsel in this case is not a proper basis for reducing the lodestar figure, though it may be taken into account in the lodestar calculation. Unlike many contingent-fee agreements, the agreement here provides for appellants' counsel to seek attorneys' fees calculated as the greater of either "the value of [their] professional services at [their] regular hourly rates or by multiplying by one third the amount recovered for the plaintiff class as damages." J.A. 602. Thus, the fee agreement did not cap attorneys' fees as a percentage of the recovery, and cannot be used to limit the recovery of attorneys' fees after determining the lodestar figure.7 We nonetheless think that the agreement may be considered in calculating the lodestar figure.

In conclusion, the district court should have considered the "amount involved and results obtained," as well as the administrative nature of the work and the fee agreement, in determining the reasonable number of hours expended or the reasonable hourly rate. A remand is therefore necessary. On remand, the district court must determine the amount of attorneys' fees, taking into account the "amount involved and results obtained," the administrative nature of the work and the low level of skill involved, and the fee agreement in calculating the lodestar figure rather than by reducing the lodestar figure itself.8

IV

The second issue in this case is whether the district court properly applied hourly rates representative of those charged in the District of Columbia, where appellants' counsel's office was located, rather than applying hourly rates in the forum where the case was brought, the Eastern District of Texas. The Supreme Court has indicated that the reasonable hourly rates to be applied in determining the lodestar figure are the "prevailing market rates in the relevant community." Blum, 465 U.S. at 895, 104 S.Ct. 1541. However, the Supreme Court has been silent on how to determine the "relevant community" under the URA or any other fee-shifting statute. The district court found that the "relevant community" in this case was the District of Columbia. We disagree.

As we have recognized, "the courts of appeals have uniformly concluded that, in general, forum rates should be used to calculate attorneys' fee awards under other fee-shifting statutes." Avera v. Sec'y of Health & Human Servs., 515 F.3d 1343, 1348 (Fed.Cir.2008).9 In Avera, we considered whether, in awarding attorneys' fees under the National Vaccine Injury Compensation Program, 42 U.S.C. §§ 300aa-10 to -34 (2000), the relevant community should be based upon "the prevailing market rate of the forum court ... or the prevailing market rate of the geographic location where the attorney is based." Id. We held that "to determine an award of attorneys' fees, a court in general should use the forum rate in the lodestar calculation." Id. at 1349. However, we recognized a narrow exception to the "forum rule" where "the bulk of the work is done outside of the [forum] in a legal market where the prevailing attorneys' rates are substantially lower." Id.

Contrary to appellants' contention, nothing in Avera suggests that the forum rate should be disregarded when plaintiffs elect to retain counsel who are located outside the forum in a jurisdiction that charges higher rates than the forum rates. In that situation, the forum rate applies absent some unusual justification for departing from it. While we have not yet squarely addressed the issue, we recognize that several circuits have acknowledged an exception to the forum rule where local counsel is either unwilling or unable to take the case.10 We agree that such an exception is appropriate, but we also agree that it is applicable only in unusual situations. Such exceptions are permissible only where supported by specific evidence that no local attorneys possess the "special expertise" necessary to take the case or that no local attorneys were willing to take the case. See McClain, 649 F.3d at 382 (application of out-of-district rates appropriate only where supported by "abundant and uncontradicted evidence" that out-of-district counsel were necessary); Interfaith Cmty., 426 F.3d at 705-06; Barjon v. Dalton, 132 F.3d 496, 501-02 (9th Cir.1997); see also Arbor Hill Concerned Citizens Neighborhood Ass'n v. Cnty. of Albany, 522 F.3d 182, 191 (2d Cir.2008) (holding that the presumption that the forum rule should be applied may be rebutted "only in the unusual case").

In this case, the only evidence to suggest that an exception to the forum rule was applicable is Bywaters's declaration indicating that the local attorney that he had originally hired to represent him was unable to help him in a "complex, specialized area of law" and that the only attorney he could find to represent him "in the whole country" was his current District of Columbia-based counsel. J.A. 539-40. We find Bywaters's conclusory declaration to be insufficient. See Schwarz v. Sec'y of Health & Human Servs., 73 F.3d 895, 907 (9th Cir.1995) (holding that exception to forum rule was inapplicable where the plaintiff's own declaration, the only evidence in support of an exception, showed only that she had difficulty obtaining local counsel).11 There is no evidence to suggest that no local attorneys were competent to handle Bywaters's case. Nor is there any indication that Bywaters conducted a reasonable search for local counsel to handle his case. While appellants were free to engage out-of-district counsel to represent them, the government should not be required to subsidize their decision to do so under these circumstances. See 10 James Wm. Moore et al., Moore's Federal Practice § 54.190[2][b][i][E] (3d ed. 1997 & Supp. 2011). Thus, we hold that an exception to the forum rule was not warranted in this case.12 On remand, the district court should apply the forum rule to determine the reasonable hourly rate, bearing in mind appellants' burden "to produce satisfactory evidence ... that the requested rates are in line with those prevailing" in the forum "for similar services by lawyers of reasonably comparable skill, experience and reputation." Blum, 465 U.S. at 896 n. 11, 104 S.Ct. 1541.

CONCLUSION

For the foregoing reasons, the district court's award of attorneys' fees is vacated and the matter is remanded for further proceedings consistent with this opinion.

VACATED and REMANDED.

PLAGER, Circuit Judge, dissenting.

The majority opinion focuses on two issues in this appeal1—first, whether the trial court correctly determined the amount of attorney fees to which the plaintiffs are entitled; and second, a subset of that question, was the trial court correct in using as the relevant market for pricing attorney services the District of Columbia, rather than Texas. On the first issue, I agree with the majority that the trial judge erred in his final calculation of the amount of reimbursable attorney fees; where we differ is on what to do about it. On the second issue, I disagree with the majority's overriding of the trial court's factual determination that the relevant market for these particular attorney services is the home base of these particular attorneys.

With regard to the first issue—whether the trial court correctly determined the amount of reimbursable attorney fees under this fee-shifting statute—the majority acknowledges the generally fine job the trial court did in sorting through the evidence, with which I agree; the majority recites the applicable law; and the majority concludes that the trial court did err in its final figure. To that point we are in agreement. How that error should be corrected, however, is a matter of dispute between us.

Simply stated, the trial court correctly invoked the controlling lodestar formula; made the required detailed findings regarding the number of hours reasonably expended by the plaintiffs' attorneys; multiplied that by the reasonable hourly rate using the "Updated Laffey Matrix" applicable to District of Columbia attorney services; and came up with a lodestar figure.

The record is undisputed that the trial court determined that the hours claimed (and awarded with small adjustment) were reasonable; the court specifically so found. The trial court properly multiplied those hours by the reasonable rates it had determined, and then arrived at what on the record is a reasonable and correct lodestar award.

Had the trial court stopped there, it would have been fine. Inexplicably, at least to me, the trial court then reduced the lodestar award by 50%. Other than noting that the fee seemed to be "extremely high" in contrast to the overall award on the merits, the only substantive basis for such a reduction the trial court mentioned was the existence of a fee agreement between the plaintiffs and their attorneys that included a contingent fee option. Otherwise there is nothing to suggest or support the reduction in the award.

I agree with my colleagues that "the fee agreement between appellants and their counsel in this case is not a proper basis for reducing the lodestar...." Maj. op. at 1231-32. The fee agreement expressly stated that the attorneys' fee could be based on either regular billing hours or a contingent fee, whichever was greater. We need not spend undue time reciting the obscure law on contingent fee considerations to simply acknowledge that under such an agreement, the possibility of a contingent fee does not serve as the benchmark for the fee award.

The Supreme Court in Perdue stated that any departure from the lodestar by the trial court must evidence "a method [for a different calculation] that is reasonable, objective, and capable of being reviewed on appeal...." Perdue v. Kenny A. ex rel. Winn, ___ U.S. ___, ___, 130 S.Ct. 1662, 1674, 176 L.Ed.2d 494 (2010). Once the contingent fee is removed from the calculation, there is nothing in the trial court's opinion that meets any of those criteria. More importantly, on the record in this case I see no reasoned basis for departing from the lodestar. The attorneys won virtually everything they set out to win. This is the kind of case that fee shifting is intended to encourage—many plaintiffs with small individual claims but a common transgression by the Government. Congress has made clear that the award is to be of the "reasonable attorney ... fees, actually incurred...." 42 U.S.C. § 4654(c). That the award is substantial when compared to the total recovery is neither surprising nor unusual. See Reply Br. 11. Once the hours expended and fees earned were found reasonable, the trial court's work in this phase of the litigation was concluded. By summarily halving the award, the trial court misunderstood its job under the statute, and by acting contrary to law exceeded its authority.

The Supreme Court has wisely said, "[a] request for attorney's fees should not result in a second major litigation. Nor should it lead to years of protracted appellate review." Perdue at 1684 (Kennedy, J., dissenting) (citing Hensley v. Eckerhart, 461 U.S. 424, 437, 103 S.Ct. 1933, 76 L.Ed.2d 40 (1983)). I would simply reverse and reinstate the lodestar award, and not put the trial court to the task of further hearings and this court to the inevitable further appeals.

Regrettably, the majority proposes instead to remand the issue to the trial court to again undertake a calculation of the lodestar, this time however to consider the "amount involved and results obtained" as part of the initial lodestar calculation, rather than as an afterthought. But that won't work either. As the majority itself acknowledges, the Supreme Court has made it abundantly clear that the lodestar controls absent "rare and exceptional circumstances." Maj. op. at 1224. There is nothing rare or exceptional about these circumstances, beyond the trial court's unfortunate abuse of its discretion in arbitrarily reducing the lodestar award which it had determined to be reasonable. On remand, the same lodestar figure, absent the halving, is the only outcome the record could support. To remand is just to make work for the trial court, which it does not need, and to provide an opportunity for another appeal here, likely of little if any value to any one. Accordingly, from the decision to remand for no good purpose I respectfully dissent.

With regard to the second issue—was the trial court correct in using as the relevant market for pricing these attorney services the District of Columbia, rather than Texas—I agree with my colleagues that "in determining the amount of reasonable attorneys' fees to award under federal fee-shifting statutes, the district court is afforded considerable discretion." Maj. op. at 1228. We defer to the trial court in so far as possible, and particularly in its fact finding, given that the trial court is closest to the parties and to the ebb and flow of the litigation, and given the legal standard on appeal: we must find an abuse of discretion to overturn the trial court's determinations.

At the trial, the plaintiffs explained fully about their search for the best counsel for this type of specialized takings litigation. There is no showing that their choice of this particular counsel, located in Washington D.C., was motivated by anything other than a desire to get the best qualified representation. Under the circumstances, particularly when much of the legal work was office work done at the attorneys' home offices, the simple economic notion of opportunity-cost would justify using the attorneys' hometown rates. The trial court made that determination, and found such rates to be reasonable under the fee-shifting statute. When parties seek justice in the courts, especially when the Government is the defendant, neither we nor the Government should be in a position to challenge a plaintiff's reasoned choice of private counsel.

In holding to the contrary, the majority fails to grant the trial court the discretion in fact-finding that the law provides, and, from the rarified heights of an appellate court and a distance of a thousand miles, denies these plaintiffs the right to seek out and employ the best lawyers they could find to handle a complex class-action litigation that has now been in dispute for over ten years. The majority explains this by saying that this is not fact-finding but a misunderstanding of law, and adding, "[t]he plaintiffs of course have the absolute right to choose their own counsel; what they do not have is the right to recover Washington, D.C. rates where competent local counsel are available." Maj. op. at 1234, n. 12. One might have thought that the trial judge was in the best position to ascertain whether there were local lawyers available with equivalent competencies to the plaintiffs' chosen lawyers in this particular field of litigation; ruling that all lawyers are fungible as a matter of law is for me carrying egalitarian-ism a bit too far.

Contrary to my colleagues' position, I would affirm the trial court's decision that the basis for determining a reasonable fee for the services rendered should be the fees charged in the hometown of the plaintiffs' chosen attorneys (Washington, D.C.) and not what some lawyers might charge for services rendered at the forum (in this case Texas). From the majority's contrary holding, I must respectfully dissent.