Opinion for the court filed by Circuit Judge J. SKELLY WRIGHT, in which Chief Judge SPOTTSWOOD W. ROBINSON III, and Circuit Judges WALD, MIKVA, HARRY T. EDWARDS, GINSBURG, and SCALIA join,
Dissenting opinion filed by Circuit Judge BORK, in which Circuit Judges TAMM, WILKEY and STARR join.
J- SKELLY WRIGHT, Circuit Judge:
On October 19, 1979, Yusaf Na’im Salahuddin (then known as Kenneth M. Brown), an inmate of the District of Columbia’s reformatory at Lorton, Virginia, filed suit in the United States District Court for the District of Columbia. The suit alleged that conditions in the reformatory’s “Adjust- , tt >,„ , J . ment Unit, where the plaintiff had previ- , , ’ . , , , , ously been incarcerated, had unconstitu- . , . , „ , , . , Anally deprived inmates. of adequate food, ^^ns, cell space, and educational, religious, legal, and rehabilitative services, and had failed to protect. them from physical assault by guards and other inmates. The suit originally sought declaratory and injunctive relief as well as damages, and it named as defendants a [1500]*1500number of District of Columbia and federal officials and the District of Columbia itself. Here, we are concerned only with the claim seeking damages from the District of Columbia (the District).1
The District Court, on September 2, 1981, granted a summary judgment motion filed by the District. The part of the District Court’s order that concerns us is its holding that the notice of claims provision in the District of Columbia Code, 12 D.C.Code § 309 (1981), applied to plaintiff’s action for unliquidated damages against the District.
In relevant part, 12 D.C.Code § 309 provides:
An action may not be maintained against the District of Columbia for unliquidated damages to person or property unless, within six months after the injury or damage was sustained, the claimant, his agent, or attorney has given notice in writing to the Commissioner [Mayor] of the District of Columbia of the approximate time, place, cause, and circumstances of the injury or damage. * * *
Because plaintiff had not filed any notice that would comply with 12 D.C.Code § 309, the District Court granted summary judgment as to the damage action. Plaintiff appealed and argued that Section 309 did not apply to federal causes of action such as his constitutional tort action.
While the appeal was pending, a panel of this court decided McClam v. Barry, 697 F.2d 366 (D.C.Cir.1983), which held that Section 309’s six-month notice of claims requirement does apply to federal damage actions against the District. In light of McClam, a panel of this court affirmed the District Court’s grant of summary judgment with respect to the damage claims against the District, 704 F.2d 1293. On May 19, 1983, this court granted reconsideration en banc and vacated the judgment of the panel.
We now hold that McClam v. Barry was in error in holding that the six-month notice of claims provision of Section 309 applies to causes of action that, like this constitutional tort claim,2 are creations of federal law.
[1501]*1501I. The Decision in McClam v. Barry
To resolve the issue whether 12 D.C. Code § 309 should apply to federal causes of action, the McClam court correctly focused its inquiry on two possible approaches. First, it asked whether Congress, when it enacted Section 309, meant for that section to apply to federal causes of action. 697 F.2d at 369. But it also realized that even if Congress had no such explicit intent the section might still be applicable to federal actions because federal law does at times borrow state legal provisions to complement federal law. Thus, as a second inquiry McClam asked whether application of Section 309 was “supported by the rationale of the rule * * requiring application of local statutes of limitations to claims based on federal laws that specify no limitations period.” Id. at 370.
We agree that these inquiries are the proper ones. That we make two inquiries reflects the existence of two available rationales that could justify applying Section 309 to federal causes of action: First, that Congress may have intended Section 309 itself to alter the normal rules applicable to federal causes of action, and second, that those normal rules themselves may take account of Section 309 just as they take account of certain other provisions of local law.
The first inquiry thus rests entirely on a rationale of actual congressional intent: Congress, the author of national legislation, may at times intend that its District of Columbia legislation modify national policies with respect to the District of Columbia. The second inquiry begins with an alternative view of congressional intent: Congress, which is constitutionally responsible for providing a comprehensive body of local law for the governance of the District of Columbia, may see itself as passing local legislation akin to that which any other local legislature might pass. The rationale for applying the provision then rests on established judicial practices concerning when state or local provisions of law should be used as rules of federal law. This approach is in part a judicial positing of congressional intent and in part an examination of the provision’s role in state and local law. For example, the rule that federal law borrows from local law when there is no federal statute of limitations in part “rests on deference to the local balancing of interests embodied in statutes of limitations — a balancing of the interest in allowing the prosecution of valid claims against the interest in preventing the prosecution of claims that, because of delay, cannot be fairly litigated.” Id. at 370 (citing Board of Regents v. Tomanio, 446 U.S. 478, 487, 100 S.Ct. 1790, 1796, 64 L.Ed.2d 440 (1980); Johnson v. Railway Express Agency, 421 U.S. 454, 464, 95 S.Ct. 1716, 1722, 44 L.Ed.2d 295 (1975)).
The McClam court apparently concluded that either of the two inquiries would justify the applicability of Section 309 to federal actions. In answer to the first, it concluded that “the D.C.Code provision was an intentional limitation [by Congress] on the right to bring even federal causes of action for damages against the District of Columbia.” 697 F.2d at 369. In answer to the second, McClam concluded that “[t]he same balancing of interests lies behind the D.C.Code notice provision” as lies behind traditional state statutes of limitations, id. at 370, and that Section 309 should thus be borrowed so that federal law as applied in the District of Columbia would reflect that local balance.
While we agree that McClam made the right inquiries, we disagree with each of its answers.
II. Congressional Intent and 12 D.C.Code § 309
12 D.C.Code § 309 was enacted by Congress in 1933, at least partially in response to this court’s en banc decision and recommendation in District of Columbia v. Leys, 63 F.2d 646, 648 (D.C.Cir.1932) (on rehearing), cert. denied, 289 U.S. 756, 53 S.Ct. 787, 77 L.Ed. 1500 (1933). See Pub.L. No. 72-385, 47 Stat. 1370 (1933). The language and purposes of the statute were broad. As McClam points out, the language contains no limitations regarding the types of [1502]*1502actions covered. 697 F.2d at 369. Similarly, the Act’s stated purposes — to provide the District an opportunity to investigate claims when all evidence is still fresh, to allow the District to seek out early settlement of meritorious claims, and generally to protect the District’s revenues from unreasonable suits — would certainly be fulfilled all the more by giving the Act as broad a reading as its language would allow. See id. See generally H.R.Rep. No. 2010, 72d Cong., 2d Sess. (1933) (describing purposes of Act). We do not, however, find these factors conclusive; other factors argue for a narrower view.
Nothing in the history of the Act indicates that Congress ever considered the question whether the Act should apply to federal actions. The Leys opinion, which urged the Act’s passage, concerned a “garden variety” tort suit against the city; the plaintiff had slipped on a defective sidewalk and brought suit long after the event. Similarly, the report accompanying the bill described the problem to be rectified as that of “[n]umerous instances * * * in which notice of injury as a result of alleged defective sidewalks, or other conditions, were brought to the [city’s] attention * * * by a filing of a suit for damages more than two years after the alleged injury occurred.” H.R.Rep. No. 2010, supra, at 1. The conclusion that Congress was motivated by the municipality’s potential liability for everyday torts is strengthened by the fact that a provision was added which allowed written Metropolitan Police reports to fulfill the notice requirement of the Act. Such reports would have been commonly associated with street accidents and similar events.
None of this discussion is meant to imply that the scope of Section 309 should be limited to the sorts of “garden variety” negligence actions described above. We do not question the subsequent cases that have given it a far broader application. See, e.g., Breen v. District of Columbia, 400 A.2d 1058 (D.C.C.A.1979) (applying Section 309 to intentional torts). The point is, however, that there is no evidence that Congress envisioned itself to be acting as other than a local legislature protecting a municipality from the threat of excessive common law tort liability. Not only was the problem that Congress discussed one of routine municipal administration — with no reference to the “unique” or “national” character of the District — but in its legislative history the Act is explicitly analogized to similar legislation passed in the states to govern other municipalities.
The analogy to municipal notice provisions existing throughout the nation was raised by this court when it first suggested congressional passage of the already introduced notice provision. The Leys opinion argued that “[provisions of this character appear to be included in [the] bill recently introduced into [Congress], and are to be found in city charters in Virginia and elsewhere.” 63 F.2d at 648. Congress similarly equated the provision with laws commonly passed by state governments. The report accompanying the bill pointed out that “[similar statutes are in effect in 32 States” and sought to justify the length of time allowed for giving notice by comparing it to the provisions in effect in “other jurisdictions.” H.R.Rep. No. 2010, supra, at 2. All of this persuades us that Congress simply saw itself to be acting as a local legislature when it passed Section 309.
Under the Constitution, Congress has authority to act as the local legislature for the District of Columbia, and thus Congress frequently enacts legislation applicable only to the District and tailored to meet local needs. Absent evidence of contrary congressional intent, such enactments should be treated as local law, interacting with federal law as would the laws of the several states. See, e.g., Sullivan v. Murphy, 478 F.2d 938, 971-973 (D.C.Cir.), cert. denied, 414 U.S. 880, 94 S.Ct. 162, 38 L.Ed.2d 125 (1973). Therefore, we do not interpret the Congress that passed this apparently local legislation as having intended that it should burden federal causes of action any more than would an analogous state ordinance.
[1503]*1503III. Federal “Borrowing” Doctrine
A. An Overview of the Doctrine
The central claim of the District is that even if 12 D.C.Code § 309 is best understood as purely local law, courts should “borrow” it as a rule of federal law when the District is sued. The doctrine of federal borrowing of local law is well established. To effectuate the goals of federal law, courts, when faced with deficiencies in federal schemes, look to other sources of law to “borrow” appropriate provisions. As the Supreme Court has said: “There will often be no specific federal legislation governing a particular transaction * * *. But silence * * * in federal legislation is no reason for limiting the reach of federal law * * *. [T]he inevitable incompleteness [of federal enactments] means that interstitial federal lawmaking is a basic responsibility of the federal courts.” United States v. Little Lake Misere Land Co., 412 U.S. 580, 593, 93 S.Ct. 2389, 2397, 37 L.Ed.2d 187 (1973).
One of the principal sources of law that courts turn to to fill the deficiencies in federal schemes is state law. For example, where a federal action contains no statute of limitations courts ordinarily look to state law as a possible source of federal law. See generally DelCostello v. Int’l Bhd of Teamsters, 462 U.S. 151, 103 S.Ct. 2281, 76 L.Ed.2d 476 (1983). Where it would be consistent with the federal policies underlying the action, courts borrow the most appropriate state limitations provisions. See Burnett v. Grattan, — U.S. —, —, 104 S.Ct. 2924, 2928, 82 L.Ed.2d 36 (1984); Johnson v. Railway Express Agency, supra, 421 U.S. at 462, 464, 95 S.Ct. at 1721, 1722. Indeed, courts have borrowed not only the limitations provisions themselves, but also state tolling policies, which “are * * * integral part[s] of a complete limitations policy” and which thus give the limitations provisions their meaning under state law. Tomanio, supra, 446 U.S. at 488, 100 S.Ct. at 1797. State provisions other than statutes of limitations and tolling policies, but analogous to them, have also been borrowed to fill gaps in federal law, see, e.g., Robertson v. Wegmann, 436 U.S. 584, 98 S.Ct. 1991, 56 L.Ed.2d 554 (1978) (borrowing state survivorship of actions rule), though the practice has been less common and the courts have been somewhat less consistent. See generally P. Bator, P. Mishkin, D. Shapiro & H. Wechsler, Hart & Wechsler’s The Federal Courts and the Federal System 829 (2d ed. 1973) (apart from statutes of limitations and their incidents, “[t]he authorities are extremely sparse” concerning the extent to which state “ ‘procedural’ and ‘quasi-procedural’ rules[,]” particularly those “which may affect outcome[,]” should be borrowed); id. at 199-200 (Supp.1981) (discussing the inconsistent results in cases concerning the extent of borrowing state procedural provisions other than “statutes of limitations and their incidents”); see also Hill, State Procedural Law in Federal Nondiversity Litigation, 69 Harv.L.Rev. 66, 91-94 (1955) (reasons for applying state procedural law are usually weak when dealing with rules other than statutes of limitations).
These borrowing practices are the result of at least two considerations. First, as mentioned above, they reflect the fact that federal law is frequently “deficient” in that it does not supply the complete legal framework necessary to the fair adjudication of federal causes of action. Although federal law may establish rights, and rights of action for the vindication of those rights, federal law does not always include all the “ ‘procedural’ or ‘quasi-procedural’ ” elements that are generally considered necessary to the fair litigation of its causes of action. Where these are absent, it certainly cannot be assumed that the interests normally embodied in those elements were intended to play no part in the federal scheme. See DelCostello, supra, 462 U.S. at —, 103 S.Ct. at 2287 (where there is no statute of limitations, “we do not ordinarily assume that Congress intended that there be no limit on actions at all”); Tomanio, supra, 446 U.S. at 488, 100 S.Ct. at 1797 (the absence of a statute of limitations cannot be understood as placing policies of [1504]*1504repose in disfavor); see also Hill, supra, 69 Harv.L.Rev. at 91-92. Thus the courts are faced with the necessity of formulating some rule to fill the deficiency in the federal scheme and thereby to effectuate federal policy. In this context, the borrowing doctrine is the alternative to judicial formulation of a uniform federal rule.
The borrowing doctrine also rests on deference to the balances that local law has struck with respect to the interests normally embodied in the elements necessary to, but absent from, the federal scheme. See Johnson v. Railway Express Agency, supra, 421 U.S. at 463-464, 95 S.Ct. at 1721-1722 (“[A]ny statute of limitations * * * reflects a value judgment concerning the point at which the interests in * * * protecting valid claims are outweighed by the interests in prohibiting the prosecution of stale ones. * * * In borrowing a state period of limitation * * *, a federal court is relying on the State’s wisdom * * *.”); McClam v. Barry, supra, 697 F.2d at 370. With respect to certain statutes, Congress has explicitly codified-this borrowing doctrine. See 42 U.S.C. § 1988 (1982) (courts should look to state law when federal law is not “suitable to carry [the policies of the Reconstruction Era civil rights statutes] into effect” or when those statutes “are deficient in the provisions necessary to furnish suitable remedies,” so long as rules of state law are not inconsistent with policies embodied in federal law); Burnett v. Grattan, supra, — U.S. at —, 104 S.Ct. at 2928; Johnson v. Railway Express Agency, supra, 421 U.S. at 464, 95 S.Ct. at 1722 (linking 42 U.S.C. § 1988 to general federal borrowing practices); see also Tomanio, supra. But see Eisenberg, State Law in Federal Civil Rights Cases: The Proper Scope of Section 1988, 128 U.Pa.L.Rev. 499 (1980) (giving very different interpretation of Section 1988). Yet neither when codified nor when used as a “fallback rule of thumb” of federal common law, DelCostello, supra, 462 U.S. at —n. 12, 103 S.Ct. at 2287 n. 12, is the doctrine an absolute one to be mechanically applied. State law rules are borrowed out of the need to effectuate federal policies in the face of incomplete federal law, and they are not borrowed if they would incorporate into federal law balances of interests that are inconsistent with the policies underlying the federal action. See 42 U.S.C. § 1988; Tomanio, supra, 446 U.S. at 487, 100 S.Ct. at 1796. As the Supreme Court has recently stated, “Federal courts must be ever vigilant to insure that application of state law poses ‘no significant threat to any identifiable policy or interest.’ ” Burks v. Lasker, 441 U.S. 471, 479, 99 S.Ct. 1831, 1838, 60 L.Ed.2d 404 (1979) (quoting Wallis v. Pan American Petroleum Corp., 384 U.S. 63, 68, 86 S.Ct. 1301, 1304, 16 L.Ed.2d 369 (1966)).
As this overview of borrowing doctrine shows, the issue presented by the doctrine is neither how to incorporate into federal law as much state law as a federal action will tolerate, nor how best to further state policies and goals in the litigation of a federal action. The issue is how to best effectuate the federal policies embodied in a federal action when the action does not itself supply the complete legal framework necessary to the effectuation of those policies. Because the practice of borrowing presupposes a need to fill a deficiency in the federal scheme, a court must first look to see if there is indeed such a deficiency. See Cohen v. Board of Education, 536 F.Supp. 486, 493-495 (S.D.N.Y.1982).
B. The Issue of Deficiency
This case can be resolved on this point. We cannot view the federal scheme for adjudicating constitutional torts as deficient for lack of a notice of claims provision. Nor can we view the District of Columbia’s notice of claims provision as simply embodying certain interests which are not explicitly provided for in the federal scheme, but which nonetheless should be treated as part of that scheme. In short, there is no deficiency in the federal scheme that would lead us to look to 12 D.C.Code § 309 as a possible source of federal law for effectuating federal policies. “State law is to be resorted to in resolving an issue if, and only if, federal law is deficient, [1505]*1505and if, and only if, state law ‘is not inconsistent with the constitution and the laws of the United States.’ ” Jaworski v. Schmidt, 684 F.2d 498, 500 (7th Cir.1982) (iquoting 42 U.S.C. § 1988), cert. denied, 460 U.S. 1015, 103 S.Ct. 1258, 75 L.Ed.2d 485 (1983).
Most of the cases involving the limiting of federal causes of action by borrowing state procedural rules have involved the question of whether to borrow state statutes of limitations, tolling policies, or other rules of response that clearly establish the point at which a cause of action ends. See, e.g., Tomanio, supra (tolling policies); Robertson, supra (survival of action rule); Johnson, supra (statutes of limitations). The courts have emphasized the frequent absence of statutes of limitations from federal enactments, the inherently legislative balancing of interests that goes into the determination of a particular period of limitation (making the judiciary particularly unsuited to make such a determination), and the indispensability of such statutes to our ideals of justice. This last point has been well understood from the beginnings of our federal system. For example, in 1805 Chief Justice John Marshall recognized the indispensability of statutes of limitations when he borrowed a federal criminal statute’s limitations period and applied it to a federal civil action for debt brought under a statute containing no limitations provision.
[I]t deserves some consideration, that if [the law] does not limit actions of debt for penalties, those actions might, in many cases, be brought at any distance of time. This would be utterly repugnant to the genius of our laws. In a country where not even treason can be prosecuted after a lapse of three years, it could scarcely be supposed that an individual would remain for ever liable to a pecuniary forfeiture.
Adams v. Woods, 6 U.S. (2 Cranch) 336, 341, 2 L.Ed. 297 (1805). The point was more recently emphasized by Justice Rehnquist when he argued for borrowing state statutes of limitations and tolling policies.
On many prior occasions, we have emphasized the importance of the policies underlying state statutes of limitations. [They] are not simply technicalities. On the contrary, they have long been respected as fundamental to a well-ordered judicial system. Making out the substantive elements of a claim for relief involves a process of pleading, discovery, and trial. The process of discovery and trial which results in the finding of ultimate facts for or against the plaintiff by the judge or jury is obviously more reliable if the witness or testimony in question is relatively fresh. Thus in the judgment of most legislatures and courts, there comes a point at which the delay of a plaintiff in asserting a claim is sufficiently likely either to impair the accuracy of the fact-finding process or to upset settled expectations that a substantive claim will be barred without respect to whether it is meritorious. * * *
Tomanio, supra, 446 U.S. at 487, 100 S.Ct. at 1796.
By logic and tradition, the absence of a statute of limitations is an egregious gap in a cause of action. Whatever the source of a cause of action, some clear end to one’s liberty to commence suit is necessary. Most decisions concerning the borrowing of state law that would burden federal actions have arisen from this need to borrow provisions, such as statutes of limitations, tolling policies, and survival rules, that serve the exclusive purpose of defining that point where the right to maintain a cause of action ends.3
[1506]*1506The McClam opinion concluded that “[t]he same balancing of interests lies behind the D.C.Code notice provision” as lies behind statutes of limitations — “a balancing of the interest in allowing the prosecution of valid claims against the interest in preventing the prosecution of claims that, because of delay, cannot be fairly litigated.” 697 F.2d at 370. Similarly, the District argues that there are no significant differences between a traditional statute of limitations and the D.C.Code’s notice provision. Although it is true that Congress was clearly concerned with ensuring the freshness of evidence when it passed 12 D.C.Code § 309, that does not resolve the issue. At a minimum, we must note that a notice provision has a very different relationship to a cause of action than that of a statute of limitations, tolling provision, or survival rule.
No one disputes that if 12 D.C.Code § 309 were applied to a federal cause of action, it would be wholly independent of relevant limitations, tolling, and survival provisions, which would all remain applicable. While these latter provisions all operate to define the point of repose, after which expectations become settled, the notice of claims provision does something different. Of course, if it is not complied with, and there are no grounds for waiver, it bars an action, and thus it acts to create settled expectations; but this is simply a penalty for noncompliance. The provision’s function is to compel notice so that the municipal defendant may investigate early, prepare a stronger case, and perhaps reach an early settlement. The need to apply traditional provisions of repose (statutes of limitations, tolling provisions, and survival rules) remains unchanged, since the notice provision has a purpose that is quite distinct.
It is important to our inquiry that a notice of claims provision supplements rather than replaces a traditional statute of limitations and operates in a quite different manner than does such a statute of limitations. Before examining the exact policies served by the provision, we must emphasize that its borrowing does not have the same support in tradition that the borrowing of a statute of limitations would have. This point does not simply rest on a belief in an inherent value to tradition; it rests instead on the fact that borrowing may be explained as a judicial positing of congressional intent. Because statutes of limitations are such universally familiar procedural aspects of litigation, and because they are so generally understood as essential to a fair scheme of litigation, the judiciary is safe in assuming that Congress intended (or at least would have intended) to limit all congressionally created causes of action by statutes of limitations. In other words, the judiciary can safely look at the absence of a limitations provision as a deficiency.4
The judiciary is on less secure ground, however, when it limits statutory or common law causes of action with procedural rules embodying interests that are not as universally understood to be as essential to fair litigation as are the repose interests embodied in traditional statutes of limitations. The absence of such rules in the cause of action cannot as easily be termed a deficiency. Indeed, the ground is less secure even when the rules embody interests similar to those embodied in traditional [1507]*1507statutes of limitations but implement those interests in a less universally accepted manner. While the borrowing of state statutes of limitations is supported by both the need for some limitations provision and the tradition of borrowing from state law to supply it, in most other instances the application of local procedural rules that would significantly inhibit the ability to bring federal actions will be neither as reasonable nor as necessary. Accord Hill, supra, 69 Harv.L.Rev. at 91-94.5
C. The Role of 12 D. C. Code § 309 in Local Law
With these considerations in mind, we turn to the District’s contention that 12 D.C.Code § 309 should, in spite of its differences, be analogized to a local statute of limitations and thus be adopted as a federal rule of law. To evaluate the argument, we must turn to the role that the provision plays in the local legal scheme — both its method of operation and the interests it serves. In addition to examining the provision’s legislative history, which we have done above, we should also examine the District of Columbia courts’ understanding of the provision. See 11 D.C.Code § 102 (1981) (District of Columbia Court of Appeals is “highest court of the District of Columbia”); Thompson v. United States, 548 F.2d 1031, 1035 (D.C.Cir.1976) (D.C. [1508]*1508Court of Appeals rather than this court is final expositor of local law).
In spite of the District’s insistence that the provision is closely akin to a statute of limitations, quite a different image emerges from the opinions of the District of Columbia Court of Appeals. That court has refused to view the provision as similar to a statute of limitations. See Gwinn v. District of Columbia, 434 A.2d 1376 (D.C. C.A.1981).
1. 12 D.C.Code § 309 as a condition precedent to the accrual of rights against the municipality.
In Gwinn, the District of Columbia Court of Appeals decided that Section 309’s notice provision was not subject to those tolling policies that would apply to a statute of limitations. The court’s reasoning is particularly significant. The court argued that compliance with the provision was a condition precedent to the very existence of any “ ‘right of action’ or ‘entitlement to maintain an action’ ” for tort damages against the municipality. Id. at 1376. This distinguished the provision from a statute of limitations and made the statute of limitations’ tolling principles inapposite because they “presuppose[ ] that a ‘right of action’ exists and that the claimant is ‘entitled to maintain’ that action.” Id. Unless notice is given, according to Gwinn, no right of action ever accrues. In effect, under Gwinn, compliance with the provision is considered an element in any damage cause of action against the municipality.
The way Gwinn characterized the operation of the notice provision makes it more difficult to view it as filling a deficiency in federal law. State law is not usually thought to add elements to federal rights of action. A court may safely posit that federal lawmakers, when setting out the elements of a cause of action, do not usually intend that there should forever remain a remedy available once those elements are established. See Tomanio, supra, 446 U.S. at 488, 100 S.Ct. at 1797. It is much more difficult to assume that those lawmakers, when they set out the elements of a federal cause of action, would normally intend for additional and unstated elements to be also considered necessary, and for state law to be consulted to supply those missing (yet necessary) elements of the federal action. In particular, nothing in federal borrowing doctrine leads us to believe that state law can precondition the accrual of federal rights of action. Cf. Campbell v. Haverhill, 155 U.S. 610, 618, 15 S.Ct. 217, 220, 39 L.Ed. 280 (1895) (arguing that a federally created right of action can be limited by a state statute of limitations because “statutes of limitations affect the remedy only, and do not impair the right”); Bomar v. Keyes, 162 F.2d 136, 140-141 (2d Cir.1947) (Hand, J.) (statutes of limitations are normally treated as going to the remedy and not as a condition to the federal right of action), cert. denied, 332 U.S. 825, 68 S.Ct. 166, 92 L.Ed. 400 (1947); see also Cohen v. Board of Education, supra, 536 F.Supp. at 493-495.
2. 12 D. C. Code § 309 as a condition on the District’s waiver of municipal tort immunity.
When we examine the policy reasons given by the District of Columbia Court of Appeals to support its construction and formal characterization of Section 309, we find that those reasons also undermine the argument that it would fill a deficiency in federal law just like a traditional statute of limitations. The District of Columbia Court of Appeals has consistently understood the notice requirement of Section 309 to be a condition placed on the partial waiver of the District’s sovereign immunity. See Gwinn, supra, 434 A.2d at 1378 & n. 3; Kelton v. District of Columbia, 413 A.2d 919, 920 (D.C.C.A.1980); Wilson v. District of Columbia, 338 A.2d 437, 438 & n. 2 (D.C.C.A.1975). As a matter of local common law the District has waived most of its traditional municipal tort immunity, although it still retains some. See generally Chandler v. District of Columbia, 404 A.2d 964, 965 & n. 2 (D.C.C.A.1979) (discussing continued validity of sovereign im[1509]*1509munity in District of Columbia law); Spencer v. General Hospital, 425 F.2d 479 (D.C.Cir.1969) (en banc) (discussing history of immunity, abandoning “governmental-proprietary” test for defining immunity, and adopting “discretionary function” test for defining immunity); Elgin v. District of Columbia, 337 F.2d 152 (D.C.Cir.1964) (same). The view that 12 D.C.Code § 309 was a condition on the waiver of municipal immunity has influenced the District of Columbia Court of Appeals’ construction of that provision. See, e.g., Gwinn, supra, 434 A.2d at 1378.
To the extent that 12 D.C.Code § 309 accommodates plaintiffs’ interests and notions of municipal immunity, it does not correspond to any deficiency in the federal law, nor indeed would it be well-fitted to filling any such deficiency. The role in constitutional tort litigation of state law immunities in general, and municipal immunities in particular, has been the subject of substantial federal law. In Owen v. City of Independence, 445 U.S. 622, 100 S.Ct. 1398, 63 L.Ed.2d 673 (1980), the Supreme Court held that “[b]y including municipalities within the class of ‘persons’ subject to liability for violations of the Federal Constitution and laws, Congress — the supreme sovereign on matters of federal law — abolished whatever vestige of the State’s sovereign immunity the municipality possessed.” Id. at 647-648, 100 S.Ct. at 1413-1414 (footnote omitted). Moreover, the Court explicitly discussed the two principal common law theories of municipal immunity — each of which has, at different times, been the basis of the District’s municipal immunity — and rejected the proposition that either could limit suits under 42 U.S.C. § 1983. 445 U.S. at 644-650, 100 S.Ct. at 1412-1415. Even before Owen clarified the content of federal law, the Seventh Circuit, in an opinion by Judge (now Justice) Stevens, held that the issue of tort immunity was controlled by uniform federal law rather than by borrowing from state law. Hampton v. City of Chicago, 484 F.2d 602, 607 (7th Cir.1973), cert. denied, 415 U.S. 917, 94 S.Ct. 1414, 39 L.Ed.2d 471 (1974). Hampton was quoted with approval and its conclusion was apparently adopted in Owen’s discussion of the inapplicability of municipal immunity concepts to Section 1983 litigation. 445 U.S. at 647 n. 30, 100 S.Ct. at 1413 n. 30. See also Jaworski v. Schmidt, supra, 684 F.2d 498 (law of immunities relating to constitutional torts is federal law unaffected by state provisions). To import notions of municipal immunity from District of Columbia law into federal civil rights law would be to undermine a federal policy rather than to fill a deficiency in the federal scheme.
IV. Conclusion
For the reasons discussed above, we overrule the panel opinion in McClam. We also note that outside of this circuit the overwhelming majority of federal and state courts that have confronted the issue of borrowing notice of claims provisions have reached the same result that we now reach.6 Joining these courts, we hold that appellant’s noncompliance with 12 D.C. [1510]*1510Code § 309 cannot bar his federal claims. Accordingly, it was error to grant the District’s motion for summary judgment.
Reversed and remanded.