OPINION OF THE COURT

GARTH, Circuit Judge:

This appeal requires us to determine whether, under the venue provisions of the Outer Continental Shelf Lands Act ("OCSLA"), 43 U.S.C. § 1349(b)(1), the government "may be found" as a defendant, and therefore sued, in the District of Delaware.

This action was brought by the plaintiff oil companies against various government officials under OCSLA. The plaintiffs filed their complaint and alleged venue in the judicial district of Delaware. The district court, in an unreported opinion and order, granted the defendants' motion to dismiss the action for improper venue, holding that even though the venue provisions of the Act were exclusive, the government defendants could not be sued in Delaware because, under the terms of the venue statute, the "may be found" provisions of the statute did not apply to the government. Hence the government defendants could not be "found" in that judicial district. We reverse.

I.

Plaintiffs, Superior Oil Company and Pennzoil Oil Gas, Inc. (hereinafter "Superior"), filed their complaint against Cecil D. Andrus, Secretary of the Interior, and two other government officials. The complaint, brought under OCSLA, challenged the refusal of the Secretary to award Superior a lease on Tract 135 in the Outer Continental Shelf Lease Sale 42. In the complaint Superior alleged that this rejection was arbitrary, capricious, and an abuse of discretion under the Administrative Procedure Act, 5 U.S.C. § 701 et seq. Superior sought a declaratory judgment and a court order requiring the Secretary to issue the lease. The complaint alleged venue under 28 U.S.C. § 1391(e)(4), the general venue statute.

The government defendants moved to dismiss the complaint for improper venue, alleging that the litigation was governed exclusively by the special venue provision of the Act, 43 U.S.C. § 1349(b)(1), and that under that statute, the government defendants were "not found" in Delaware. Section 1349(b)(1) provides for venue in the judicial district where any defendant resides or may be found, or in the judicial district of the State nearest the place where the cause of action arose.

Superior argued that venue was proper under 28 U.S.C. § 1391(e)(4), the general venue statute. In addition, contrary to the government's argument, Superior contended that the government defendants "may be found" in the District of Delaware as that term appears in 43 U.S.C. § 1349(b)(1). Thus, both parties agree on the applicability of the OCSLA venue statute, § 1349(b)(1), but disagree on the interpretation to be given to its terms.

The district court, as we have previously noted, held that 43 U.S.C. § 1349(b)(1) was exclusive but that under its interpretation of § 1349(b)(1), venue would not lie in Delaware with respect to the government defendants.

As we have indicated, this appeal presents but one issue for resolution: Are government defendants subject to the "may be found" venue provision of § 1349(b)(1) so that an action against such defendants may be brought in any judicial district (here Delaware) where they are found?

II.

The government contends that the venue provisions of OCSLA, 43 U.S.C. § 1349(b)(1), are exclusive. Superior, while disputing exclusivity nevertheless maintains that venue is proper in Delaware under either 43 U.S.C. § 1349(b)(1) or 28 U.S.C. § 1391(e)(4). See Brief for Superior at 8. Thus, neither party disagrees that, in an action such as the present one, § 1349(b)(1) is the venue statute to which we should look. In this posture, the need to determine whether § 1349(b)(1) is exclusive is substantially diminished.

If called upon to decide the exclusivity issue, however, we would be persuaded to agree with the district court that § 1349(b)(1) is exclusive. We regard its exclusivity as stemming not only from the special substantive content of the Outer Continental Shelf Lands Act, but also from the breadth of the OCSLA venue provisions which, in large part, incorporate the relevant options of the general venue statute, 28 U.S.C. § 1391(e).

OCSLA defines a body of law uniquely applicable to the seabed, the subsoil, and fixed structures such as artificial island drilling rigs, all of which pertain to the outer continental shelf lands. This body of law clearly is distinct from the substantive law generally applicable within continental limits. See Rodrigue v. Aetna Casualty Surety Co., 395 U.S. 352, 355, 89 S.Ct. 1835, 1837, 23 L.Ed.2d 360 (1969). The fact that OCSLA is distinctive in terms of its subject matter and the geographical areas to which it pertains, is apparent from the Act itself and the discussion in Rodrigue, supra. Rodrigue touches upon various comparisons of OCSLA with other enactments involving more traditional substantive areas. The Supreme Court, in Fourco Glass Co. v. Transmirra Corp., 353 U.S. 222, 228, 77 S.Ct. 787, 791, 1 L.Ed.2d 786 (1957), also dealt with a special substantive area for which Congress had provided particular venue options. The Court there held that the venue statute applicable to patent infringement actions was special because it was "complete, independent and alone controlling in its sphere" because of the nature of such actions.

In the context of the matters to which OCSLA pertains, we conclude that § 1349(b)(1) is also "complete, independent and alone controlling" in the sphere of the outer continental shelf lands. Thus, even if we were to reach Superior's argument (and we do not), that the general venue statute, 28 U.S.C. § 1391(e), supplements the special venue statute, 43 U.S.C. § 1349(b)(1), we would still be persuaded by the principles stated in Fourco:

However inclusive may be the general language of a statute, it `will not be held to apply to a matter specifically dealt with in another part of the same enactment. . . . Specific terms prevail over the general in the same or another statute which otherwise might be controlling.' Ginsberg Sons v. Popkin, 285 U.S. 204, 208 [52 S.Ct. 322, 323, 76 L.Ed. 704]; MacEvoy Co. v. United States, 322 U.S. 102, 107 [64 S.Ct. 890, 893, 88 L.Ed. 1163].

Id. at 228-29, 77 S.Ct. at 791-792.

Moreover, a comparison of 43 U.S.C. § 1349(b)(1) with 28 U.S.C. § 1391(e) reveals that Congress, in enacting the venue provisions of OCSLA, incorporated into § 1349(b)(1) the relevant provisions of the general venue statute.

Section 1349(b)(1) reads:

Jurisdiction and venue of actions (b)(1) Except as provided in subsection (c) of this section, the district courts of the United States shall have jurisdiction of cases and controversies arising out of, or in connection with (A) any operation conducted on the outer Continental Shelf which involves exploration, development, or production of the minerals, of the subsoil and seabed of the outer Continental Shelf, or which involves rights to such minerals, or (B) the cancellation, suspension, or termination of a lease or permit under this subchapter. Proceedings with respect to any such case or controversy may be instituted in the judicial district in which any defendant resides or may be found, or in the judicial district of the State nearest the place the cause of action arose.

(emphasis supplied).

28 U.S.C. § 1391(e), provides, in pertinent part:

(e) A civil action in which a defendant is an officer or employee of the United States or any agency thereof acting in his official capacity or under color of legal authority, or an agency of the United States, or the United States, may, except as otherwise provided by law, be brought in any judicial district in which (1) a defendant in the action resides, or (2) the cause of action arose, or (3) any real property involved in the action is situated, or (4) the plaintiff resides if no real property is involved in the action. . . .

Thus, it can be seen that both statutes provide for venue where a defendant "resides". The "cause of action" option found in 28 U.S.C. § 1391(e) also has its counterpart in 43 U.S.C. § 1349(b)(1).

Venue alternatives (3) and (4) in the general venue statute § 1391(e) depend upon the presence or absence of real property. Such considerations have little, if any, relevance to OCSLA proceedings and thus, understandably, were not included among the venue options of § 1349(b)(1). Hence, with the exception of the real property provisions found in § 1391(e)(3) and (4), which we deem not to be relevant in an OCSLA action, all other provisions of the general statute appear to be included within § 1349(b)(1).

Having concluded that we must focus our attention on § 1349(b)(1), we turn to a consideration of whether, under that statute, the government defendants "may be found" in Delaware.

III.

Superior contends that venue is properly laid in the District of Delaware under 43 U.S.C. § 1349(b)(1) because the defendants, all of whom are federal officers being sued in their official capacities, "may be found" in Delaware as that term is used in the statute. Section 1349(b)(1), as we have observed, provides for three distinct venue alternatives. Venue may be had in: (1) the district in which a defendant resides; or (2) the judicial district in which a defendant may be found; or (3) the judicial district of the State nearest the place the cause of action arose.

In the present case, we are not concerned with the third venue alternative, i. e., the judicial district of the State nearest the place the cause of action arose. Nor are we concerned with the first alternative, i.e., the judicial district in which a defendant resides. It is unchallenged that the official residence of all of the defendants is Washington, D.C., and had Superior sued in that forum, clearly there would have been venue under that provision of the statute. Thus, the critical question has been narrowed down to the appropriate interpretation of the phrase "may be found" — the second venue alternative. Specifically, we must decide whether a federal official, sued in his official capacity, "may be found" in the District of Delaware, or indeed in any judicial district of the United States, within the meaning of § 1349(b)(1).

A.

Superior argues that we should give plain and literal meaning to § 1349(b)(1), and that the starting point in every case involving statutory construction is the language of the statute itself. See e. g. Ernst Ernst v. Hochfelder, 425 U.S. 185, 197, 96 S.Ct. 1375, 1382, 47 L.Ed.2d 668 (1976). Superior then points out that § 1349(b)(1) gives a plaintiff a choice of options insofar as venue is concerned. Here, the statute is explicit and unambiguous in providing that "proceedings . . . may be instituted in the judicial district in which any defendant resides or may be found, or in the judicial district of the State nearest the place the cause of action arose." Superior's argument proceeds that a court must give effect to each and every clause and word of a statute, as it is drawn, and if such effect is given, then, contrary to the government's position, the broad term "may be found" cannot be equated to or with, the narrower term "resides".

Apparently recognizing the force of this logic, the government contends that § 1349(b)(1) has the potential for providing a broader range of venue choices in a suit by the Secretary against a private corporation, than it does in a suit by a private corporation against the Secretary. Indeed, the government argues that it considers the "may be found" provision to apply fully to corporate defendants. Thus, according to the government's interpretation of § 1349(b)(1), if the United States was the plaintiff in an action rather than the defendant, it could sue any other defendant, be it a corporate or foreign defendant, where that defendant could be "found" in terms of constitutional contacts.

The government supports this position with the following policy argument:

[I]nterpreting the language of 43 U.S.C. § 1349(b)(1) to mean that venue for suits against the Secretary lies anywhere Interior owns land or maintains an office would produce the absurd result that every plaintiff would have a choice of suing the Secretary in an OCSLA case in virtually any of the United States judicial districts. Forum shopping would result in suits relating to a lease off the coast of Massachusetts being brought in forums having no nexus with the case, such as Oklahoma or Hawaii.

Brief for government at 7.

Consistent with its analysis, the government concludes that Congress did not intend to permit suits against the Secretary in his official capacity in every district where the Department of Interior has an office. Accordingly, the government argues that the second venue option of § 1349(b)(1), providing for suit where a defendant "may be found", has no application to the government.

The district court, in granting the government's motion to dismiss Superior's complaint, noted that both parties to the action agreed that an objective of the "may be found" venue option in § 1349(b)(1) was to insure that corporate defendants could be sued in any district in which they were conducting continuing business activities. See Opinion of the District Court at A. 34-35. The court then went on to accept the government's argument that the "may be found" venue option cannot be applied in an equally broad manner to a suit against a federal employee in his official capacity. The court observed: "[G]iven the setting in which section 1349(b)(1) was enacted, . . . it [is] highly unlikely that Congress intended such a radical departure from the then existing practice." Opinion of the District Court A. 35.

Although we do not discount the concerns expressed by the government, we suggest that these concerns are more properly addressed to the Congress than to us. In light of the explicit language of the statute, which provides three discrete and distinct venue options, we are not persuaded by the government's position that the government defendants should be treated differently than any other defendant. Moreover, as we demonstrate below, we do not believe the legislative history provides a sufficient basis for interpreting the statute in the manner urged by the government.

B.

Contrary to the government's argument, the statute fails to distinguish between nongovernment defendants and government defendants. The statute by its express terms does not draw that distinction. Nor does the legislative history indicate such a distinction, and to the extent that the district court concluded that it did, we believe the district court erred.

In the original OCSLA, enacted in 1953, 43 U.S.C. §§ 1337(j) and 1335(d) provided that persons "complaining of a cancellation of a lease by the Secretary" or of "a negative determination" by the Secretary regarding maintenance or validation of a prior lease, could secure judicial review only in the United States District Court for the District of Columbia, thereby restricting venue, in effect to the place where the government defendant resides. See 43 U.S.C. § 1337(j) (cancellation of lease), and § 1335(d) (negative determination) of the 1953 enactment.

Section 1333(b) of the original Act, which provided general venue for all remaining actions arising out of the statute (presumably excepting lease cancellations and negative determinations), did not limit judicial review to the District of Columbia but provided the same three venue options now provided by § 1349(b)(1). In relevant part, 43 U.S.C. § 1333(b) of the 1953 Act read:

[P]roceedings . . . may be instituted in the judicial district in which any defendant resides or may be found, or in the judicial district of the adjacent state nearest the place where the cause of action arose.

In 1978, Congress deleted the cancellation of lease venue provision, 43 U.S.C. § 1337(j), from the original Act, but added a cancellation of lease venue provision to the present § 1349(b)(1), thereby broadening venue for that purpose. In addressing itself to this change in venue, Congress did not, however, amend or delete § 1335(d) of the original Act, which continued to restrict and still limits to the District of Columbia, review of negative determinations made by the Secretary.

Both parties rely on inferences to be drawn from these Congressional actions to support their arguments that the legislative history bolsters their respective positions. Superior contends that by incorporating the lease cancellation review provision into § 1349(b)(1) and retaining the more restricted venue provision in § 1335(d), Congress intended not only that different results would obtain under the two provisions, but also that the broadened venue provision would apply to the Secretary (a putative federal defendant), since it is the Secretary who has the substantive authority to cancel a lease under 43 U.S.C. § 1337(o). Thus, Superior argues that Congress specifically contemplated that the "may be found" venue language would apply to government defendants. See Brief for Superior at 20-21.

The government, however, asserts that there is no evidence in either the 1953 Act or the 1978 Amendments that Congress intended that the Secretary could be sued in every judicial district. It contends that a lack of any discussion of the reason for a special venue provision is an indication that Congress did not intend to change the venue status quo. See Brief for Government at 8-9.

We believe that whatever implications may be drawn from Congress's actions in 1978 support the position taken by Superior. Certainly, in distinguishing between the restricted venue provision of § 1335(d), which in the case of a negative determination permitted review, and thus venue, only in Washington, D.C., and the new all-inclusive OCSLA venue statute, § 1349(b)(1), Congress had to have considered the "may be found" venue option to which it has now made lease cancellations subject. Moreover, if Congress had intended government defendants to be treated differently from any other defendant, or if Congress had intended different venue provisions to be applied when the government was a plaintiff rather than a defendant, it could have so specified. Congress could have easily said, for example, that proceedings "may be instituted in the judicial district in which the defendant United States resides or in the judicial district where any defendant other than the United States resides or may be found." By the express terms of § 1349(b)(1), however, Congress made no such distinctions either between plaintiffs and defendants or between private corporations and the government.

C.

The government, in effect, would have us ignore the "may be found" venue option. Superior responds that if the government's argument were to prevail, in effect, the "may be found" venue option would be identical with the "resides" option. If it were, the statute would then read: "Proceedings . . . may be instituted in the judicial district where any defendant resides or resides, or in the judicial district of the State nearest the place the cause of action arose." See Brief for Superior at 19. Such a construction, Superior argues, is patently absurd.

We agree with Superior that neither the government nor a court may construe the words of a statute so as to give no meaning to an express term. Just as we are not at liberty to add to a Congressional enactment, so too we are not at liberty to delete significant terms from a statute. See Arkansas Valley Industries, Inc. v. Freeman, 415 F.2d 713, 718 (8th Cir. 1969). If we were to read § 1349(b)(1) as the government would have us read it, we would be obliterating any difference in content between the terms "resides" and "may be found", and we would be required to interpret those terms as being synonymous. This we decline to do. The cases which have discussed the difference between the terms "reside" and "may be found" have focused on the general venue statute, 28 U.S.C. § 1391(e)(1), which provides for bringing an action where a defendant "resides", but not where a defendant "may be found". Consequently, those cases have rejected venue where it has been asserted on a "may be found" basis. Nevertheless, such cases highlight the very distinction between the terms "resides" and "may be found" which the government here seeks to blur.

In Reuben H. Donnelley Corp. v. FTC, 580 F.2d 264 (7th Cir. 1978), Donnelley filed an action in the Northern District of Illinois against the FTC and each of its five individual commissioners in order to enjoin the commencement of certain administrative hearings. The FTC challenged venue brought under § 1391(e)(1), claiming that the federal defendants in the action did not reside in Illinois even though the FTC maintained a regional office in that locale. The district court held that the FTC was a resident within the meaning of § 1391 on the basis of its regional office. The Seventh Circuit reversed that holding, reasoning that no such expansive interpretation of residency was intended by Congress and that such a holding would, as a practical matter, render superfluous three of the four venue options provided in § 1391(e). See also Hartke v. Federal Aviation Administration, 369 F. Supp. 741, 746 (E.D.N.Y. 1973) (the statutory reference to the district in which a defendant "resides" may not reasonably be construed to include every district where some subordinate has an office); Kings County Economic Community Development Association v. Hardin, 333 F. Supp. 1302, (N.D.Cal. 1971), (court determined that § 1391(e)(1) cannot be interpreted so as to permit venue on the basis that two of the government officials named as defendants had offices within the district in question).

We refuse to ignore an express provision of a statute which Congress has enacted, just as we decline to give a meaning to that provision which would distort the clear language of the statute. We believe there is a fundamental distinction between the terms "resides" and "may be found", and until such time as Congress changes the scope and plain meaning of 43 U.S.C. § 1349(b)(1), we will give it effect as it is drawn. "It is our judicial function to apply statutes on the basis of what Congress has written, not what Congress might have written." United States v. Great Northern Railway Co., 343 U.S. 562, 575, 72 S.Ct. 985, 993, 96 L.Ed. 1142 (1952). (emphasis supplied)

III.

Having concluded that effect must be given to the "may be found" venue option of § 1349(b)(1), the last question to be addressed is whether the government defendants are indeed found in Delaware. We are satisfied that the government defendants here, even though personally they may not be found within the District of Delaware, are officially "found" in that District, by reason of their ultimate responsibility for the services rendered there. See Green v. Carlson, 581 F.2d 669, 676 (7th Cir. 1978), aff'd, Carlson v. Green, 446 U.S. 14, 100 S.Ct. 1468, 64 L.Ed.2d 15 (1980). Although normally we are reluctant to make such a determination ourselves, we do so in this case because of the position taken by the government. The government has never seriously disputed that, if it were held subject to the "may be found" venue provision, it indeed is "found" in the District of Delaware. The government's argument has never been predicated upon "not being found", but instead, the government, as we have indicated, has focused on the policy argument that to allow venue to lie in every judicial district, would be unreasonable and does not reflect congressional intent. It has thus virtually conceded, as indeed it must, that the government defendants are to be "found" in Delaware.

IV.

In rejecting the government's policy argument, we acknowledge that actions arising under OCSLA may, as a practical matter, be brought in every judicial district. As the government has pointed out, once effect is given to the second venue option of § 1349(b)(1), there is no judicial district in which the United States may not be found in an OCSLA proceeding. Such a result is necessarily mandated by the inclusion of the "may be found" option in § 1349(b)(1).

Our holding should not place any greater burdens or any undue strains upon the federal government, for it is evident that in such actions the Department of Justice would appear for the government defendants, and each judicial district in which an action could be brought has a United States Attorney. See, e. g., Stafford v. Briggs, 444 U.S. 527, 544, 100 S.Ct. 774, 784, 63 L.Ed.2d 1 (1980). If, however, we have underestimated the extent of such burdens, it is Congress's function, if Congress deems it appropriate, to strike the balance in favor of the government. As the Supreme Court stated in Tennessee Valley Authority v. Hill, 437 U.S. 153, 194-95, 98 S.Ct. 2279, 2301-2302, 57 L.Ed.2d 117 (1978):

[A court's] appraisal of the wisdom or unwisdom of a particular course consciously selected by the Congress is to be put aside in the process of interpreting a statute. . . . We do not sit as a committee of review, nor are we vested with the power of veto. . . .

* * * * * *

[I]n our constitutional system the commitment to the separation of powers is too fundamental for us to pre-empt congressional action by judicially decreeing what accords with "common sense and the public weal." Our Constitution vests such responsibilities in the political branches.

Moreover, if a particular forum does prove to be inconvenient or burdensome, a district court has broad discretionary power under 28 U.S.C. § 1404(a) to transfer any civil action to any other district where it might have been brought "[f]or the convenience of parties and witnesses, in the interest of [justice.]"

V.

We conclude that the term "may be found" as it is used in § 1349(b)(1) applies to the United States in the same manner as it does to all other defendants and therefore must be given the same effect regardless of whether the government or a private corporation is the defendant. Because the federal government "may be found" in the District of Delaware, venue in this proceeding properly lies under § 1349(b)(1).

The October 22, 1980 order of the district court which dismissed Superior's action by reason of improper venue will be reversed and we will remand this case to the district court for further proceedings.