WISDOM, Circuit Judge:
These condemnation cases present the issue whether compensation should be determined under federal law or under the law of the state where the condemned property is located when a licensee of the Federal Power Commission exercises the power of eminent domain in federal court as authorized by Section 21 of the Federal Power Act, 16 U.S.C. § 814 (1970). We reserved this question in Louisiana v. Lindsey, 5 Cir. 1975, 524 F.2d 934, as did the Supreme Court in Grand River Dam Authority v. Grand-Hydro, Inc., 1948, 335 U.S. 359, 69 S.Ct. 114, 93 L.Ed. 64. We now hold that federal law controls.
I.
The plaintiff-appellee, Georgia Power Company, is a privately owned Georgia utility. It intends to operate a hydroelectric power generating facility with a dam across the Oconee River in Hancock and Putnam Counties, Georgia, at a point known as Lau-rens Shoals. The dam will produce a lake to be known as Lake Wallace. The appellants are Georgia landowners with property which will be inundated by Lake Wallace.1
[1180]*1180Under the Federal Power Act, 16 U.S.C. §§ 791a, et seq. (1970), the Federal Power Commission may issue licenses to certain persons or entities to construct, operate, or maintain various hydroelectric generating facilities. 16 U.S.C. § 797(e). The F.P.C. issued a license to Georgia Power on August 6,1969, for the Lake Wallace Project.2 As a licensee, Georgia Power may exercise the right of eminent domain under Section 21 of the Federal Power Act, 16 U.S.C. § 814 (1970), which provides:
When any licensee cannot acquire by contract or pledges an unimproved dam site or the right to use or damage the lands or property of others necessary to the construction, maintenance, or operation of any dam, reservoir, diversion structure, or the works appurtenant or accessory thereto ... it may acquire the same by the exercise of the right of eminent domain in the district court of the United States for the district in which such land or other property may be located, or in the State courts. The practice and procedure in any action or proceeding for that purpose in the district court of the United States shall conform as nearly as may be with the practice and procedure in similar action or proceeding in the courts of the State where the property is situated: Provided, That United States district courts shall only have jurisdiction of cases when the amount claimed by the owner of the property to be condemned exceeds $3,000.
Exercising this right, Georgia Power brought actions in the District Court for the Middle District of Georgia. Senior Judge W. A. Bootle appointed a three-member Commission, see Fed.R.Civ.Proc. 71A, to determine the amount of compensation due the condemnees. In his instructions to the Commissioners, Judge Bootle set out the federal law of compensation. These rules differ from Georgia law. Instruction No. 20 directs the Commissioners to ignore an increase in value which the Georgia Power project has created in the condemned property; a Georgia court might recognize' such value. See Hard v. Housing Authority, Ga. 1963, 219 Ga. 74,132 S.E.2d 25. Instruction No. 21 allows the Commissioners to offset any recovery for land actually taken with benefits to any remaining property caused by the project; Georgia law prohibits such a set-off. See Ga.Code Ann. § 36-504 (1970). The district court declined to instruct the Commissioners to include a reasonable attorneys’ fee in their award; in a Georgia proceeding a reasonable attorneys’ fee would be allowed. See White v. Georgia Power Company, Ga.1976, 237 Ga. 341, 227 S.E.2d 385.
After these instructions were first made in 1975 several condemnees in cases other than No. 75-4448 filed motions and objections opposing the use of federal law. The district court held a pre-trial hearing on the objections. Judge Bootle made several changes in the phrasing of his instructions, but he regarded federal law as controlling.3 The Commission then heard evidence in No. 75-4448. Its report concluded that the benefits accruing to the land remaining with the landowners exceeded the value of the property taken. As a result, the Commission awarded n'o monetary compensation. The landowners' opposed Georgia Power’s motion to confirm the report on the ground that the offset of benefits was improper. [1181]*1181The court overruled these objections and the condemnees appealed.4
No. 77-1327 began after Judge Bootle’s decision in No. 75-4448 to apply federal law. The condemnees in No. 77-1327 nevertheless moved that the court incorporate Georgia law into its instructions.5 The judge conducted a pre-trial conference and heard argument on the question. He then denied the motion and adhered to his previous order that federal law applies. Judge Bootle certified the question for immediate review under 28 U.S.C. § 1292(b) (1970). This Court allowed the appeal, and consolidated it with No. 75-4448.
II.
Before a federal court may reach the question of applying state or federal common law to an issue before it, the court must determine that the source of the right or authority in question is federal. If the source is not federal, Erie Railroad Co. v. Tompkins, 1938, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188, and the Rules of Decision Act6 direct that state law apply of its own force. Even if the source is federal, the court must follow any congressional directions about the proper law to apply. See Comment, Adopting State Law as the Federal Rule of Decision: A Proposed Test, 43 U.Chi.L.Rev. 823, 826 (1976).
We find that the source of the power to condemn property contained in Section 21 is federal. Eminent domain inheres in sovereignty. See Kohl v. United States, 1876, 91 U.S. 367, 23 L.Ed. 449. Within our federal system the states and the federal government exercise independent powers of eminent domain, and neither can delegate the use of the other’s authority. 91 U.S. at 372-73, 23 L.Ed. 449; Latinette v. City of St. Louis, 7 Cir. 1912, 201 F. 676, 678. Thus, when Congress provided that licensees could exercise “the” power of eminent domain, it was referring to the federal power. Both courts and commentators have described the eminent domain power delegated in Section 21 as the federal power. Federal Power Commission v. Tuscarora Indian Nation, 1960, 362 U.S. 99, 120, 80 S.Ct. 543, 4 L.Ed.2d 584; City of Tacoma v. Taxpayers of Tacoma, 1958, 357 U.S. 320, 340, 78 S.Ct. 1209, 2 L.Ed.2d 1345; Chapman v. Public Utility District No. 1, 9 Cir. 1966, 367 F.2d 163, 167; 7 Moore’s Federal Practice ¶ 71A.10[2]. When it condemns land under Section 21, a licensee acts as the agent of the United States government. Tuscarora Nation of Indians v. Power Authority, 2 Cir. 1958, 257 F.2d 885, 894, vacated as moot sub nom, McMorran v. Tuscarora Nation of Indians, 362 U.S. 608, 80 S.Ct. 960, 4 L.Ed.2d 1009. This delegation of federal power is constitutional. Missouri v. Union Electric Light & Power Co., C.D.Mo.1930, 42 F.2d 692; see Thatcher v. Tennessee Gas Transmission Co., 5 Cir. 1950, 180 F.2d 644, cert. denied, 340 U.S. 829, 71 S.Ct. 66, 95 L.Ed. 609 (upholding identical delegation in the Natural Gas Act, 15 U.S.C. § 717(b) (1970)).
Appellants argue that Section 21 does not delegate federal authority, but merely provides a federal forum for exercise of a state created eminent domain power. Such an interpretation does not accord with decisions upholding the right of a licensee to condemn property that it could not obtain under the applicable state statute. E. g., City of Tacoma v. Taxpayers of Tacoma, 1958, 357 U.S. 320, 78 S.Ct. 1209, 2 L.Ed.2d 1345; First Iowa Hydro-Electric Cooperative v. Federal Power Commission, 1946, 328 U.S. 152, 66 S.Ct. 906, 90 L.Ed. 1143. [1182]*1182There is no suggestion in the statute that Congress intended to create an interstitial power of eminent domain that would function only when state powers were inadequate. In contrast, Section 19, 16 U.S.C. § 812 (1970) expressly limits the FPC’s authority to regulate rates to situations where a state does not have its own regulations. Section 20, 16 U.S.C. § 813 (1970) places similar limits on other powers of regulation.
We are not persuaded by the argument that the $3,000 in controversy requirement shows that Section 21 is a jurisdictional provision. It is true that this requirement parallels the amount in controversy requirement for diversity cases at the time the Federal Power Act was enacted. Congress, however, added the $3,000 requirement late in the consideration of the Federal Power Act. Compare H.R.Rep.No.715, 65th Cong., 2d Sess. 27, H.R.Rep.No.1147, 65th Cong., 3d Sess. 13, and H.R.Rep.No.61, 66th Cong., 1st Sess. 10 (original versions without the $3,000 amount in controversy requirement) with S.Rep.No.180, 66th Cong., 1st Sess. 9 and 16 U.S.C. § 814 (1970) (versions containing $3,000 amount in controversy requirement as added on the House floor). As a result, the requirement cannot be read as an indication of the original purpose of Section 21. Members of Congress do not seem to have believed the amount in controversy amendment significantly changed Section 21; the Senate and Conference Committee Reports, written after the House adoption of the amendment, do not mention it. Furthermore, the House debates establish that the amount in controversy minimum was designed to prevent inconvenience to a landowner who might otherwise need to travel several hundred miles to press a small compensation claim. 58 Cong.Rec. 2240 (1919) (remarks of Mr. Stevenson).
We have not found any congressional directive to apply state law in either the language of Section 21 or its legislative history. Therefore, the second prerequisite to our consideration of the choice of law issue is met. The language in the statute referring to state “practice and procedure” does not govern the choice between substantive rules of compensation. The Supreme Court has held that such language “[does] not, and could not, affect questions of substantive right — such as the measure of compensation — grounded upon the Constitution of the United States”. United States v. Miller, 1943, 317 U.S. 369, 380, 63 S.Ct. 276, 283, 87 L.Ed. 336. Accord United States v. 93.970 Acres of Land, 1959, 360 U.S. 328, 79 S.Ct. 1193, 3 L.Ed.2d 1275. The appellants attempt to distinguish Miller on the ground that the plaintiff was the United States rather than a private licensee. We cannot see why in this context the identity of the plaintiff would change the meaning of virtually identical language.7 [1183]*1183And insofar as Section 21 did require conformity with state practice, it has been superseded by Rule 71A of the Federal Rules of Civil Procedure. United States v. 93.970 Acres of Land, 1959, 360 U.S. 328, 79 S.Ct. 1193, 3 L.Ed.2d 1275; 7 Moore’s Federal Practice ¶ 71A.10[2] at 71A-252 (2d ed. 1975).
There is little discussion of Section 21 in the Committee Reports or the congressional debates, and none of it illuminates the choice of law issue. Appellants argue, however, that other sections of the statute, such as Sections 19 and 20, evidence a congressional respect for state law which includes state rules of compensation in an eminent domain proceeding. We are not ready to make this inference from unrelated provisions in the Federal Power Act without more support in the legislative materials. The legislative history reveals that Congress took care to preserve state rights in the regulation of power and property to forestall a constitutional attack on the Federal Power Act. See, e. g., First Iowa Hydro-Electric Cooperative v. Federal Power Commission, 328 U.S. at 174, 66 S.Ct. 906. As a result, the Federal Power Act contains many explicit acknowledgements of state authority. See, e. g., Id. at 174 n. 19, 66 S.Ct. 906; Sections 4(a) and (c)8 (requires cooperation between the FPC and the executive departments and other agencies of the state and federal governments on specified subject); 4(f)9 (requires notice of application for a preliminary permit to be given to any state or municipality likely to be interested); 7(a)10 (gives preference in issuing permits and licenses to states and municipalities); 10(e)11 (provides that certain licenses to states and municipalities be issued and enjoyed without charge); 1412 (gives states or municipalities as well as federal government an option to acquire a licensed project); 19;13 20;14 and 2715 (a “saving” clause concerning state laws relating to the control, appropriation, use or distribution of water used in irrigation or for municipal or other uses, or any vested right acquired therein). The failure to mention specifically state laws of eminent domain contrasts with these provisions, and suggests, if anything, that Congress did not believe it necessary to follow state eminent domain rules or desire courts to do so.16
[1184]*1184The Committee Reports reveal that some members had concerns over the constitutionality of Section 21, although it is not clear whether they thought it an improper invasion of state prerogatives or an improper delegation of federal authority to private parties. H.R.Rep.No.61, 66th Cong., 1st Sess. 10. At any rate, a majority expressed belief that the provision was constitutional and appropriate, without any indication that such a conclusion would be possible only if federal courts applied state law in Section 21 cases. Id17
The appellants also argue that Congress endorsed the use of state compensation law when it enacted a provision substantially identical to Section 21 in the Natural Gas Act, 15 U.S.C. § 717f(h) (1970), without commenting adversely on certain cases appellants read to mandate adoption of state law for all substantive issues arising in a Section 21 lawsuit. In a similar vein, the appellants urge us to follow these decisions as a matter of stare decisis. In particular, they draw our attention to a line of Eighth Circuit cases beginning with Feltz v. Central Nebraska Public Power and Irrigation District, 8 Cir. 1942, 124 F.2d 57818 and to Oakland Club v. South Carolina Public Service Authority, 1940, E.D.S.C., 30 F.Supp. 334, aff’d, 4 Cir. 1940, 110 F.2d 84.
We do not read these cases as determinative of the issue before us. In Feltz, a FPC licensee condemned land for relocation of a highway. The landowner claimed consequential losses because the property was divided and because of increased distance to the railhead. The jury gave a smaller award than the appraisers, and the court added interest from the date of the taking. Both sides appealed. Among other questions, the court of appeals ruled on the exclusion of evidence of consequential damages, 124 F.2d at 578, and the propriety of interest on a compensation award, 124 F.2d at 584. Without articulating its reasoning, the court indicated that Nebraska law controlled both issues. There is no indication that the parties contested this issue or that the court gave it any serious consideration. Indeed, because of the mixture of issues each party may have thought it to his advantage to rely upon state law. The con-demnees had read that state law to allow consequential damages and the condemnor could cite favorable state law on interest. The specific issue of just compensation raised before this Court was not considered. The other Eighth Circuit cases also lack any discussion of a general choice of law problem or of the specific issue before us. In the absence of any reasoned discussion of the difficult choice of law issue, we cannot accept these cases as persuasive authority or interpret congressional inattention to them as an endorsement of the view that state law should fill all substantive gaps in Section 21.
Oakland Club is not on point. In that case the federal licensee enjoyed state as well as federal power of eminent domain. It filed a condemnation in federal court which the property owner contested on the grounds that the South Carolina procedure was constitutionally defective and that Section 21 authorized the licensee to take only an easement rather than a fee.
The district court upheld the state procedure and interpreted Section 21 to authorize the taking of a fee interest. It went on to make several observations that the appellants have emphasized in this case. First, the court stated that Section 21 is not a “comprehensive, self-contained and exclusive law of eminent domain”. Second, the [1185]*1185court construed Section 21 “not as an exclusive law of eminent domain, . . . but as complementary to the State law, and as enabling the holder of a Federal Power license to exercise in the Federal courts . the substantive rights of eminent domain granted to it under the State law.” 30 F.Supp. at 341. The Court of Appeals approved of these statements and added its own suggestion that Section 21 was meant to give a condemnor “the privileges of eminent domain under the State law . . . ” 110 F.2d 84, 86.
Insofar as the Oakland Club litigation stands for the point that Section 21 does not preempt any applicable state powers of eminent domain, it is irrelevant to the issue before us.19 Insofar as the case supports the argument that Section 21 is solely a jurisdictional provision, we reject it for the reasons already expressed. And insofar as it interprets Section 21 to grant a state rather than a federal power of eminent domain, we find it in conflict with holdings by the Supreme Court. See Federal Power Commission v. Tuscarora Indian Nation, 1960, 362 U.S. 99, 120, 80 S.Ct. 543, 4 L.Ed.2d 584; City of Tacoma v. Taxpayers of Tacoma, 1958, 357 U.S. 320, 340, 78 S.Ct. 1209, 2 L.Ed.2d 1345; see also Kohl v. United States, 1876, 91 U.S. 367, 372-73, 23 L.Ed. 449, 451 (federal government lacks the power to grant the use of a state power of eminent domain). Neither the district court nor the Court of Appeals in Oakland Club considered the specific issue whether state or federal law governs the methodology of determining compensation under Section 21. In sum, the Oakland Club opinions are far too questionable and undear to justify adherence by this Court or to warrant an interpretation of congressional failure to comment upon them as an expression of legislative intent on the choice of law issue before us.20
III.
We now proceed to the principal issue. Because the source of the power delegated by the Federal Power Act is federal, the governing law must be federal. See, e. g., Friendly, In Praise of Erie — And of the New Federal Common Law, 39 N.Y.U.L. Rev. 383, 407 (1964); Mishkin, The Variousness of “Federal Law”: Competence and Discretion in the Choice of National and State Rules for Decision, 105 U.Pa.L.Rev. 797, 798-801 (1957); Comment, Adopting State Law as the Federal Rule of Decision: A Proposed Test, 43 U.Chi.L.Rev. 823, 826-27 (1976). Of course, federal law may be fashioned by adopting parallel state rules. E. g., Clearfield Trust Co. v. United States, 1943, 318 U.S. 363, 367, 63 S.Ct. 573, 87 L.Ed. 838. For instance, the courts usually determine what constitutes “property” by looking to state law. E. g., United States v. Certain Property Located In Borough of Manhattan, 2 Cir. 1965, 344 F.2d 142, 144-[1186]*118645 (Friendly, J.); United States v. 145.30 Acres of Land, 1974 D.La., 385 F.Supp. 699, aff’d, 5 Cir. 1975, 524 F.2d 1231.
Courts apply independent rules, however, when considering the specific issue of compensation in federal condemnations. E. g., United States v. Miller, 1943, 317 U.S. 369, 380, 63 S.Ct. 276, 87 L.Ed. 336; United States v. Certain Property Located in Borough of Manhattan, 2 Cir. 1965, 344 F.2d 142, 146; United States v. City of New York, 2 Cir. 1948, 165 F.2d 526, 528 (L. Hand, J.); United States v. Certain Parcels of Land, 3 Cir. 1944, 144 F.2d 626, 628; United States v. 3,595.98 Acres of Land, 1962 N.D.Cal., 212 F.Supp. 617; 12 Wright & Miller, Federal Practice and Procedure: Civil § 3042 (1973). These cases did not arise in the specific context of a federal condemnation by a Federal Power Act licensee pursuant to Section 21. Nevertheless, we should follow the same rule unless there are circumstances unique to a federal condemnation under Section 21 which require the opposite result.21
The appellants suggest that several such circumstances exist. First, they contend that we should distinguish between cases where the United States is a party and where it is not, applying federal rules of compensation only in the former situation. Since a licensee exercises the federal power of eminent domain, it is not immediately apparent why such a distinction should be made. That the property will be used privately is not reason enough. In Miller, for instance, the United States took land for use by a private railroad. Appellants, however, cite Public Utility District No. 1 v. City of Seattle, 9 Cir. 1967, 382 F.2d 666, 1969, cert. dismissed, 396 U.S. 803, 90 S.Ct. 22, 24 L.Ed.2d 59, and Tacoma v. Taxpayers of Tacoma, 1957, Wash., 307 P.2d 567, rev’d, 357 U.S. 320, 78 S.Ct. 1209, 2 L.Ed.2d 1345, to support a theory that the Federal Power Act delegates less than the full power of eminent domain. As a result, the appellants continue, a licensee should not benefit from the same legal rules as the United States.
In City of Seattle the Court considered whether a licensee must compensate the owner of shorelands and, for power site values, the owner of adjoining uplands, for property needed for a licensed hydroelectric project. The United States would not have paid any compensation if it had condemned the same property. Instead, it could have relied upon the dominant navigational ser--[1187]*1187vitude for its power to utilize without compensation the stream bed and shorelands of navigable waters up to the ordinary high water level. See 382 F.2d at 669; United States v. Twin City Power Co., 1956, 350 U.S. 222, 76 S.Ct. 259, 100 L.Ed. 240. The Ninth Circuit examined the Federal Power Act and determined that it did not delegate the dominant servitude of navigation. Thus, it found that in the circumstances before it a licensee must pay compensation because it, unlike the Government, takes something it does not already have. 382 F.2d at 672. With respect to power site values, the Court determined that the rule precluding assertion of power site value against the United States, United States v. Twin City Power Co., 1956, 350 U.S. 222, 76 S.Ct. 259, 100 L.Ed. 240, goes “hand in hand” with the dominant navigational servitude and thus is not available to a licensee. 382 F.2d at 673-74.
Whether a licensee enjoys the dominant navigational servitude is irrelevant to the issue before us. The Ninth Circuit did not question that Section 21 delegates some of the federal government’s powers. And the opinion does not support the contention that a licensee enjoys anything less than the full federal power to condemn private property. Significantly, when the Ninth Circuit turned to the amount of compensation due the shoreland and upland property owners it relied exclusively on federal cases. See 382 F.2d at 673-74.
Tacoma holds that Section 21 does not give a licensee power to condemn state-owned land. The theory of the case appears to be that' Congress could not delegate such power to a state-created entity, rather than that it did not intend to do so. 207 P.2d at 577. This theory is questionable at best. See First Iowa Hydro-Electric Cooperative v. Federal Power Commission, 1946, 328 U.S. 152, 66 S.Ct. 906, 90 L.Ed. 1143; Kohl v. United States, 1876, 91 U.S. 367, 23 L.Ed. 449; Washington Department of Game v. Federal Power Commission, 9 Cir. 1953, 207 F.2d 391, 1954, cert. denied, 347 U.S. 936, 74 S.Ct. 626, 98 L.Ed. 1087.22 Even if this holding is correct, it does not affect a licensee’s power to acquire privately held land.
Although the Supreme Court has never ruled directly on the scope of the delegation of federal power in Section 21, Federal Power Commission v. Tuscarora Indian Nation, 1960, 362 U.S. 99, 80 S.Ct. 543, 4 L.Ed.2d 584, suggests that the delegation is not a limited one. As part of its case against condemnation by a FPC licensee, the Indian Nation argued that 25 U.S.C. § 17723 applied to the condemnation of its lands. The Court rejected this contention:
But § 177 is not applicable to the sovereign United States nor hence to its licensees to whom Congress has delegated federal eminent domain powers under § 21 of the Federal Power Act.
362 U.S. at 120, 80 S.Ct. at 555 (emphasis added). In its discussion of this point the Court cited decisions in which the United States was a party without any indication they would not be equally applicable to a federal licensee. See 362 U.S. at 120-21, 80 S.Ct. 543.
Similarly, in Grand River Dam Authority v. Grand-Hydro, 1948, 335 U.S. 359, 69 S.Ct. 114, 93 L.Ed. 64, the Supreme Court appears to draw no distinction between the federal government and its licensee under the Federal Power Act: [1188]*1188335 U.S. at 373, 69 S.Ct. at 121. See also Public Utility District No. 1 v. City of Seattle, 9 Cir. 1967, 382 F.2d 666, 675 (Byrne, J., dissenting), 1969, cert. dismissed, 396 U.S. 803, 90 S.Ct. 22, 24 L.Ed.2d 59.
[1187]*1187If either the United States, or its licensee as such, were seeking to acquire this land under the Federal Power Act, it might face different considerations from those stated above.
[1188]*1188In summary, we find no reason to ignore the general rule of following federal law to set compensation in a federal condemnation because the United States is not taking the land directly.
Second, the appellants have argued that the federal interest in the value of property taken in a Section 21 proceeding is so small that the analysis established by Clearfield Trust Co. v. United States, 1943, 318 U.S. 363, 63 S.Ct. 573, 87 L.Ed. 838 and its progeny, including most recently Miree v. DeKalb County, 1977,-U.S.-, 97 S.Ct. 2490, 53 L.Ed.2d 557, compels federal adoption of state law. Even if a sharp distinction between a licensee and the United States cannot be drawn, it may be that the federal interest in the price paid for licensed hydroelectric plants is less than in other government projects. If so, use of state law would be appropriate notwithstanding the usual rule that federal common law of compensation governs in federal condemnation.
A review of Clearfield, Miree, and other federal common law decisions fails to convince us, however, that state law should be followed. That review reveals a changing emphasis of factors and shifting preferences for one law or the other. In Clear-field Trust the Court chose uniform federal common law to govern an action against the United States on a federal check. The federal interest in uniformity for easier administration was sufficient to justify use of uniform federal common law. See 318 U.S. at 366, 63 S.Ct. 573; Comment, Adopting State Law as the Federal Rule of Decision: A Proposed Test, 43 U.Chi.L.Rev. 823, 831 (1976).
United States v. Standard Oil Company, 1947, 332 U.S. 301, 67 S.Ct. 1604, 91 L.Ed. 2067 also took a hospitable stance to the use of federal common law. That case posed the question whether the United States could recover for damages suffered when a serviceman is injured. The Court found first that the source of any legal right to recover should be federal. It then held such federal law should be uniform because there was no affirmative need for diverse results, the issue did not involve local concerns, and a uniform rule would be appropriate in fiscal matters. Id. at 310-11, 67 S.Ct. 1604.
Other cases diverge from this presumption favoring federal law. For instance, Bank of America v. Parnell, 1956, 352 U.S. 29, 77 S.Ct. 119, 1 L.Ed.2d 93 refused to extend the Clearfield approach of uniform law to suits between private parties simply because the subject matter of the controversy happens to be federal commercial paper. Instead, the Court indicated that an immediate federal interest must be involved. Id. at 33-34, 77 S.Ct. 119. And in Wallis v. Pan American Petroleum Corp., 1966, 384 U.S. 63, 86 S.Ct. 1301, 16 L.Ed.2d 369, the Supreme Court indicated that state law should usually govern unless a “significant conflict between some federal policy or interest and the use of state law in the premises” is shown. Id. at 68, 86 S.Ct. at 1304. Accord Miree v. DeKalb County, 1977, - U.S. -, 97 S.Ct. 2490, 53 L.Ed.2d 557. The Court found no such interest in United States v. Yazell, 1966, 382 U.S. 341, 86 S.Ct. 500, 15 L.Ed.2d 404, a suit by the federal government to collect a debt owed the Small Business Administration. The Court held that Texas community property rules would apply because the Small Business Administration had individually negotiated the loan in the context of and with specific reference to Texas law and because the role of the United States as creditor in the case was no different from any private creditor. The Court also noted the intense local interest in family property and the protection of women, as well as the lack of any existing federal common law to apply. Id. at 346-53, 86 S.Ct. 500. In United States v. Little Lake Misere Land Co., 1973, 412 U.S. 580, 93 S.Ct. 2389, 37 L.Ed.2d 187, on the other hand, the Supreme Court refused to adopt state law. The issue in Little Lake was whether a [1189]*1189state statute retroactively making mineral rights imprescriptible should govern federal land acquisitions under the Migratory Bird Conservation Act.24 Adoption would have deprived the federal government of a contract transferring the mineral rights to the land in question ten years after its acquisition. The statute conflicted with the federal program because it deprived the government of its contractual interests and created uncertainty in land transactions under the Act. Id. 382 U.S. at 596-99, 86 S.Ct. 500.
Together these cases produce a balancing test. See generally Comment, Adopting State Law as the Federal Rule of Decision: A Proposed Test, 43 U.Chi.L.Rev. 799 (1976). On one side is the federal interest in carrying out a program in the most efficient and effective manner possible. On the other is a state’s interest in the preservation of its control over local interests, particularly traditional interests such as family law and real property transactions, and in preventing displacement of state law. Of course, the ultimate goal of the creation of federal law by the courts is to carry out the federal program in question. See United States v. Little Lake Misere Land Co., 412 U.S. at 584-601, 93 S.Ct. 2389; United States v. Standard Oil Co., 332 U.S. at 309-11, 67 S.Ct. 1604. Thus, if state law would actually frustrate rather than only hinder a federal program, federal common law must be applied regardless of state interests. See, e. g., United States v. Little Lake Misere Land Co. On the other hand, the Supreme Court has demonstrated a growing desire to minimize displacement of state law. See Miree v. DeKalb County, 1977, - U.S. -, 97 S.Ct. 2490, 53 L.Ed.2d 557.
Each side of this controversy has provided us with weights for its side of the scale. Georgia Power points to the degree of federal involvement in a licensed project. A license is a prerequisite to construction of a project. 16 U.S.C. § 817 (1970). The design, construction, recreation facilities, and ecological impact of a project are subject to approval by the FPC. 16 U.S.C. § 803, (1970), 18 C.F.R. § 4.41 et seq. (1977). The Commission has certain powers to regulate the rates and charges of electricity generated by the project. 16 U.S.C. §§ 812-13 (1970). Georgia Power also points out that the federal government has a financial interest in minimizing acquisition costs for a licensed project since it holds an option to acquire the project when the license expires upon payment of net investment, less certain items. 16 U.S.C. § 807 (1970). In addition, Georgia Power notes that the government may acquire a project in time of national emergency, 16 U.S.C. § 809 •(1970).
The landowners emphasize that this lawsuit involves private parties and that Georgia Power (or any other licensee) is likely to condemn property only in the state where it operates. They also point out that property rights are a traditional state concern and that Georgia has an elaborate law of eminent domain that reflects considered policy judgments a state would want the federal courts to respect. Finally they contend that no substantial rights and duties of the United States hinge on the outcome of this litigation.
The balance tips toward the need for federal law. That this controversy is between private parties is not determinative. Miree v. DeKalb County, - U.S. -, -, 97 S.Ct. 2490, 53 L.Ed.2d 557 (Burger, C. J., concurring); United States v. Little Lake Misere Land Co., 412 U.S. at 592-93, 93 S.Ct. 2389; Bank of America v. Parnell, 352 U.S. at 34, 77 S.Ct. 119. Nor must we automatically apply state law because property is involved. Little Lake involved a property transaction. Moreover, eminent domain is not like most areas of property law, where state regulation has been virtually exclusive, but one where the United States has developed and exercises an independent power. See, e. g., Kohl v. United States, 91 U.S. 367, 23 L.Ed. 449.
One possible reason for adopting state law as the federal rule of decision is the [1190]*1190advantage of knowing what state law says compared with the uncertainty as to rules the federal courts would create. See McKenna v. Wallis, 5 Cir. 1965, 344 F.2d 432, 444 (Wisdom, J., dissenting), rev’d sub nom., Wallis v. Pan American Petroleum Corp., 1966, 384 U.S. 63, 86 S.Ct. 1301, 16 L.Ed.2d 369; Note, The Federal Common Law, 82 Harv.L.Rev. 1512, 1519 (1969). Since federal law on compensation already exists, there is no such reason to adopt state law here.25
Most importantly, federal common law of compensation is the appropriate choice in this case because the use of state law could interfere with achievement of congressional aims. Even when Congress passed the Federal Power Act in 1920 its primary concern was to stimulate private development of America’s hydroelectric power resources. See, e. g., First Iowa Hydro-Electric Cooperative v. Federal Power Commission, 1946, 328 U.S. 152, 180, 66 S.Ct. 906, 90 L.Ed. 1143; 56 Cong.Rec. 9810 (remarks of Mr. LaFollette); H.R.Rep.No.715, 65th Cong., 2d Sess. 15. That goal has become even more significant with the energy shortages of the 1970’s and the adoption of a national policy to reduce dependence on foreign sources of energy. To fulfil that congressional policy, hydroelectric development must be maximized. The use of state rules of compensation when they would produce a substantially higher award for the landowner could retard development and frustrate the goal of full utilization of hydroelectric resources. If we considered only the increased cost to the Lake Wallace project from these appeals the interference with the federal policy might not appear significant. When, however, all the acquisitions for a major project are grouped together, and even more so when many licensed projects are considered as an aggregate, the increase in cost and the accompanying chill on development resulting from adoption of state rules such as those in Georgia would be substantial. State compensation rules could also interfere with the national interest in reducing energy costs because the higher land acquisition expenses would need to be offset by higher utility rates.
The federal government’s option under Section 14 of the Act, 16 U.S.C. § 807, to acquire a project at the expiration of a license creates another United States interest in minimizing the cost of licensed facilities by applying federal valuation rules. See First Iowa Hydro-Electric Cooperative v. Federal Power Commission, 328 U.S. at 172-73, 66 S.Ct. 906; 26 Public Utility District No. 1 v. City of Seattle, 9 Cir. 1967, 382 F.2d 666, 675 (Byrne, J. dissenting) 1969, [1191]*1191cert. dismissed, 396 U.S. 803, 90 S.Ct. 22, 24 L.Ed.2d 59. Congress expressed an interest in keeping this option price down when it passed the Federal Power Act. Rep. Ferris submitted a Minority Report to the original House Bill, because it did not include a provision limiting the option price of projects which had decreased in value to the fair value of the property taken, H.R.Rep. No.715, 65th Cong., 2d Sess. 40 (Minority Report). In Rep. Ferris’ eyes, this omission meant the option price could be so high that no government would be justified in paying it. Id. at 40. Congress eventually added Rep. Ferris’ amendment, thereby 'expressing an interest in controlling acquisition costs. Use of federal compensation standards would reduce the option price and further this congressional desire to make government acquisition a realistic possibility-
For these reasons, we find that nothing in United States v. Clearfield Trust Co. and its progeny gives reason to diverge from the practice of employing federal compensation rules in a federal compensation proceeding. If anything, those cases push toward use of the federal law. As required by Wallis v. Pan American Petroleum Corp., 384 U.S. 63, 86 S.Ct. 1301, 16 L.Ed.2d 369, a significant conflict between the federal policy of the Federal Power Act and the use of Georgia law can be shown.27
The appellants have raised a third argument against federal law in Section 21 cases. They contend that if we apply federal common law, licensees such as Georgia Power will have the choice of a Section 21 suit in federal court with federal rules or a suit in state court governed by state law. [1192]*1192This, they continue, is the type of forum shopping Erie Railroad Co. v. Tompkins condemns. See also Hanna v. Plumer, 1965, 380 U.S. 460, 85 S.Ct. 1136, 14 L.Ed.2d 8; see generally, Ely, The Irrepressible Myth of Erie, 87 Harv.L.Rev. 693 (1974). Erie, of course, does not apply directly to this litigation because the source of the Section 21 condemnation power is federal. E. g., Clearfield Trust Co. v. United States, 318 U.S. 363, 63 S.Ct. 573, 87 L.Ed. 838.
Any broader implications of the Erie decision about the desirability of uniformity between state and federal courts do not apply here either. If a licensee were to bring a Section 21 condemnation action in a state court, that court would be required to follow the same law as a federal court even if that were federal common law. Friendly, In Praise of Erie — And of the New Federal Common Law, 39 N.Y.U.L.Rev. 383, 405 (1964). See also P. Bator, P. Mishkin, D. Shapiro & H. Wechsler, Hart and Wechsler’s The Federal Courts and the Federal System at 825 (1973). If the proceeding in the state court is under the federal power, the defendant may remove. 7 Moore’s Federal Practice ¶ 71A.10[2] at 71A-252 (2d ed. 1975).28 Thus, there would not be different substantive rules of decision depending upon the forum chosen. If a licensee brought a state condemnation action in state court, state law would govern. E. g., Grand River Dam Authority v. Grand Hydro, 335 U.S. at 372-73, 69 S.Ct. 114. But, if a state condemnation action were brought in federal court — assuming there is jurisdiction29 — state law would also govern of its own force under Erie. 12 Wright & Miller, Federal Practice and Procedure: Civil § 3055 (1973). Again, there would be no difference in substantive law according to forum. Insofar as the appellants’ argument rests on the possibility that different substantive rules might be applied in a state action in state court compared with a federal action in federal court, they can find no comfort in Erie. Justice Brandéis said nothing about a need for uniformity in different causes of action based on similar or even identical facts.
IV.
The appellants argue to us that a decision to apply federal common law in this litigation would render Section 21, as applied, unconstitutional. The theory is that a law permitting Georgia Power to “grant” the benefits of Georgia law arbitrarily through its choice of forum denies equal protection to landowners sued under the federal rather than the state statute. This contention was mentioned (though not developed) in several of the original answers to Georgia Power’s condemnation complaint, but otherwise appears to have received no attention below.30 Part of the argument rests on a disputed factual assertion that Georgia Power arbitrarily discriminates among con-demnees by bringing some state suits in state court, where the court would apply the more generous compensation formula, against landowners situated similarly to appellants. The district court had no opportunity to rule on this factual dispute or the equal protection argument. Therefore, we need not consider them. See, e. g., Commercial Credit Business Loans, Inc. v. St. Louis Terminal Field Warehouse, 5 Cir. 1974, 514 F.2d 75, 77; E. E. O. C. v. Stan[1193]*1193dard Forge and Axle Co., 5 Cir. 1974, 496 F.2d 1392, 1394-95; D. H. Overmyer Co. v. Loflin, 5 Cir. 1971, 440 F.2d 1213, 1215, cert. denied, 404 U.S. 851, 92 S.Ct. 87, 30 L.Ed.2d 90. To the extent the equal protection issue is before us, we find it without merit. There can be no contention that any landowner will be awarded constitutionally inadequate compensation if federal law controls since the federal rules satisfy the requirements of the fifth amendment. See, e. g., United States v. Miller, 1943, 317 U.S. 369, 63 S.Ct. 276, 87 L.Ed. 336. Nor can we seriously consider the possibility that there are constitutional difficulties whenever a prosecutor or other plaintiff meeting state action requirements may choose between two or more causes of action with different consequences for defendants. If there were, the Justice Department could not constitutionally choose to bring a civil antitrust suit against one offender and a criminal suit against another.
V.
The appellants have made several other arguments in their briefs and at oral argument which we regard as unnecessary to discuss in this opinion. We have, however, considered these arguments. We find them to be without merit.
NO. 75-4448 — AFFIRMED.
NO. 77-1327 — AFFIRMED AND REMANDED FOR FURTHER PROCEEDINGS.