COFFIN, Circuit Judge.
The owners and managers of Castle Square, a housing project established under § 221(d) (3) of the National Housing Act, 12 U.S.C. § 1715 l (d) (3), and located in Boston, brought this action seeking a declaratory judgment that Chapter 11 of the Boston Ordinances of 1970 ("the 1970 Ordinance") is unconstitutional under the Supremacy Clause, Article VI, § 2, insofar as it imposes rent control on Castle Square.1 More simply, they contend that the application of the rent control ordinance by the Boston Rent Board interferes with the authority lodged in the Federal Housing Administration (FHA) to regulate maximum rent levels in § 221(d) (3) housing.2
For present purposes, the relevant facts are few. Having obtained permission from the FHA for a rent increase averaging $11 per apartment per month, effective February 1, 1971, appellants announced their intention so to raise rents. On January 4, 1971, the Rent Board informed appellants that their failure to comply with the procedural requirements of the ordinance made the rent increase ineffective. After the district court denied appellants' request for a preliminary injunction, 322 F. Supp. 1126 (D.Mass.1971), on the ground that they had not exhausted administrative remedies, they invoked the Rent Board's rent increase procedures and were allowed the lesser increase of $5 per apartment per month. Still aggrieved, appellants returned to the district court, which then stayed proceedings pending state court determination of the validity of the Rent Board's action, 334 F. Supp. 861. This appeal followed.3
The abstention doctrine announced in Railroad Comm'n of Texas v. Pullman Co., 312 U.S. 496, 61 S. Ct. 643, 85 L. Ed. 971 (1941) and applied recently in Reetz v. Bozanich, 397 U.S. 82, 90 S. Ct. 788, 25 L. Ed. 2d 68 (1970); Fornaris v. Ridge Tool Co., 400 U.S. 41, 91 S. Ct. 156, 27 L. Ed. 2d 174 (1970), and Askew v. Hargrave, 401 U.S. 476, 91 S. Ct. 856, 28 L. Ed. 2d 196 (1971), is an equitable doctrine premised on the "avoidance of needless friction with state policies", 312 U.S. at 500, 61 S. Ct. at 645. See also Wulp v. Corcoran, 454 F.2d 826 (1st Cir. 1972). Under its strictures, a federal court having jurisdiction over a claimed violation of federal rights by state law must nevertheless abstain from deciding that claim where a state court construction of its law may obviate the need for a decision on the federal claim. While ordinarily the federal claims in abstention cases have been purely constitutional ones, the same policies are applicable where, as here, the federal claim is in part statutory. Cf. Metlakatla Indian Community, etc. v. Egan, 363 U.S. 555, 80 S. Ct. 1321, 4 L. Ed. 2d 1397 (1960).
But the mere mechanical possibility that a state court decision might make adjudication of the federal claim unnecessary does not itself make abstention appropriate. The high costs attendant upon abstention in terms of delay and frustration of federal claims are matters of equity as well, and are appropriately weighed in the equitable balance.4 In Pullman itself, Justice Frankfurter wrote not of "friction" but of "needless friction". In declining to order abstention in Wisconsin v. Constantineau, 400 U.S. 433, 91 S. Ct. 507, 27 L. Ed. 2d 515 (1971), the Supreme Court emphasized that "the abstention rule only applies where 'the issue of state law is uncertain.' Harman v. Forssenius, 380 U.S. 528, 534 [, 85 S. Ct. 1177, 1182, 14 L. Ed. 2d 50]", 400 U.S. at 438, 91 S. Ct. at 511, and that "where there is no ambiguity in the state statute, the federal court should not abstain . . .." 400 U.S. at 439, 91 S. Ct. at 511. See also Lavoie v. Bigwood, 457 F.2d 7, at 9 n. 1 (1st Cir., 1972). We see no substitute for a close analysis of the challenged state law.
Appellees argue that abstention was appropriate because Boston's claimed authority under Massachusetts law to regulate the rents of units in § 221(d) (3) housing projects is arguably vulnerable. Our assessment of its authority begins with Chapter 797 of the Acts of 1969 (as amended, "the Boston enabling act"), which gave Boston alone the authority "notwithstanding any provision of law to the contrary . . . by ordinance [to] control the rent for the use or occupancy of housing accommodations in structures having four or more dwelling units . . .."5 In the next session, the legislature passed Chapter 842 of the Acts of 1970, M.G.L.A. c. 40 App., §§ 1-1 to 1-14 ("the general enabling act"), under which all towns of 50,000 or more were authorized to pass rent control ordinances to apply to "controlled rental units". On the same day, the Boston enabling act was amended by Chapter 853 of the Acts of 1970, extending Boston's authority to "structures having three or more dwelling units" (§ 2) and instructing the city that any board which it established to regulate rents must set "maximum rents . . . at levels which will yield landlords a fair net operating income . . .." The 1970 Ordinance was passed by the Boston City Council on November 30, 1970, and repassed over the mayor's veto on December 28.
The case that Massachusetts has conferred clear authority on Boston can be argued on alternative theories.6 The first begins with the assertion that the Boston enabling act was intended to be wholly distinct from the general enabling act, and emphasizes that the ordinance itself purports to derive statutory authority solely from the Boston act. Even passing the danger that the city's interpretation of a state law delimiting its powers may be self-serving, it is notable that the ordinance gives to the Rent Board power, for example, to "refuse an upward adjustment of maximum rent if it determines that the effective [sic] housing accommodation does not comply with the State Sanitary Code or the Boston Building Code or Fire Prevention Code" (§ 4(e)), a power among those enumerated in the general act, § 7(d), M.G.L.A. c. 40 App. § 1-7(d), but not in the Boston act. In this connection, it may be significant that when the sentence in the Boston enabling act which begins with the phrase "notwithstanding any provision of law to the contrary" was amended by c. 863, § 2, in 1970, that emphatic phrase was deleted, arguably to make the Boston enabling act subject to the general enabling act passed on the same day. In turn, in defining "controlled rental units", the general enabling act made reference to the Boston enabling act "and any act in amendment thereof". We are thus not confident that Boston's authority can be determined independently of the general enabling act.
Appellants press more aggressively a second theory, that even if the enabling acts are to be viewed as an integrated scheme, the general enabling act specifically preserves for Boston wider authority than is allowed to other cities in the Commonwealth. If true, this would simplify an otherwise complex issue. Appellants rely on the general enabling act's definition of "controlled rental units", which encompasses
"all rental units except . . . (3) rental units [of] which a governmental unit, agency, or authority . . . (ii) regulates the rents, other than units regulated
(a) under the provisions of this act, or
(b) under the provisions of [the Boston enabling act], or
(c) under the provisions of any other general or special law authorizing municipal control of rental levels for all or certain rental units within a municipality . . .." § 3(b), M.G.L.A. c. 40 App., § 1-3(b).
The conclusion which appellants would have us draw from this draftsman's corkscrew is that no units "regulated . . . (b) under the provisions of [the Boston enabling act]" are "(3) rental units [of] which a governmental unit . . . (ii) regulates the rent", or, stated another way, that the (3) (ii) exception to "controlled rental units" does not apply to Boston.7 They urge, in effect, that (b) is an affirmative grant to Boston to define its own authority restrained only by the Boston enabling act.
But the riddle admits of another solution, suggested by the (a), "under the provisions of this act", with which (b), "under the provisions of [the Boston enabling act]", is parallel. That is, (a) may have been inserted out of an overcautious concern that "(3) rental units [of] which a governmental unit . . . (ii) regulates the rent" might be interpreted as including those units regulated under the authority of the very act setting out the definition. On this reading, the draftsman feared that units regulated under the act might be treated as falling within the (3) (ii) exception to "controlled rental units". Nor, for that matter, does any other interpretation occur to us. If (a) was designed to defuse a supposed "self-destruct" mechanism, then we may plausibly attribute to (b) the related function of insuring that Boston not be deprived of its authority under the general enabling act by the 1969 Boston enabling act or its amendment in 1970.8
So construed, (b) would be no talisman. The (3) (ii) exception might be viewed as an explicit restriction on Boston's authority, in which case a rental unit subject to governmental rent control other than under the Boston enabling act, as, for example, a unit in Castle Square under FHA rent control, would be exempt from regulation. Whether, apart from (b), the enabling acts should be construed to mesh in a way supportive of Boston's claimed authority is surely uncertain and a matter for the state courts.9
Because in our view the foregoing discussion casts more than contrived doubt on Boston's authority to impose rent control on Castle Square, we have no occasion to consider additional issues of state law which might warrant abstention, notably whether, even viewed in isolation, the Boston act would impliedly confer the claimed authority and whether the "fair net operating income" guaranteed to the landlord by both the Boston enabling act and the 1970 Ordinance was effectively safeguarded in the present case. Nor do we by our discussion intend to intimate views as to the proper construction of the ordinance or either enabling act. That is a question properly litigated before the state courts.10 Rather, we have pursued our discussion only so far as was necessary to identify unsettled and difficult issues of state law potentially dispositive of the case. In light of our discussion, we hold that the district court was correct both to abstain pending state court review of the rent board's action and to retain jurisdiction. Doud v. Hodge, 350 U.S. 485, 76 S. Ct. 491, 100 L. Ed. 577 (1956).
1. The ordinance, entitled "An Ordinance Regulating Certain Residential Rents and Evictions", excludes from coverage "housing accommodations which the United States . . . regulates the individual rents thereof" but then excepts from the exclusion "federally subsidized buildings under the . . . 221(d) 3" program. § 1(e) (i).
2. Our opinion in Hahn v. Gottlieb, 430 F.2d 1243 (1970), explains in detail the FHA regulatory scheme and its application to Castle Square.
3. The district court's abstention order is clearly appealable as a final order under 28 U.S.C. § 1291. Idlewild Liquor Corp. v. Epstein, 370 U.S. 713, 715 n. 2, 8 L. Ed. 2d 794, 82 S. Ct. 1294 (1962); see also 9 Moore's Federal Practice para. 110.20 [4. -- 2], at 251.
4. The length of delay and the concomitant danger to federal rights have been recognized by numerous commentators, see, e. g., C. Wright, The Law of Federal Courts 198 (2d ed. 1970), and by the Supreme Court. See, e. g., Reetz v. Bozanich, supra, 397 U.S. at 86, 90 S. Ct. 788; England v. Louisiana State Board of Medical Examiners, 375 U.S. 411, 434-436, 84 S. Ct. 461, 11 L. Ed. 2d 440 (1964) (Douglas, J., concurring).
5. On November 21, 1969, Boston enacted Chapter 10 of its 1969 Ordinances ("the 1969 Ordinance"), the predecessor to its present ordinance. By specific exclusion of "housing accommodations . . . the rents of which are regulated, by the United States . . . or any authority created under the laws thereof", Chapter 10 left Castle Square unregulated. § 1(e) (i). Chapter 10 was expressly repealed by § 14 of the 1970 Ordinance.
6. The absence of state court interpretation in any respect relevant here of either the Boston or the general act does not in itself mandate abstention. Wisconsin v. Constantineau, supra.
7. Since the 1969 Ordinance did not control rents of units in § 221(d) (3) projects, appellants cannot argue that (b) endorsed pre-existing control. See note 5, supra.
8. Likewise, the third of the companion clauses, (c), arguably accords like protection to other cities, a protection made necessary by the passage on the same day of a rent control enabling act for Brookline, St.1970, c. 843.
9. See Marshal House v. Rent Control Bd., 358 Mass. 686, 1971 Mass.Adv.Sh. 161, 170-171, 266 N.E.2d 876, in which the Supreme Judicial Court, after a detailed analysis, concluded that Brookline had authority under a special enabling act, notwithstanding a restriction in the general enabling act, to regulate the rents of owner-occupied two and three family dwellings.
10. Judicial review of rent board decisions is provided by § 10(a) of the general enabling act, M.G.L.A. c. 40, App., § 1-10(a): "Any person who is aggrieved by any action, regulation or order of the board . . . may file a complaint against the board . . . in a district court . . .." But see Marshal House v. Rent Control Bd., 358 Mass. 686, 1971 Mass.Adv.Sh. 161, 164, 266 N.E.2d 876, indicating that a challenge to the validity of the general enabling act, as opposed to its application in a particular case, is properly addressed to the Superior Court under M.G.L.A. c. 231A, § 1, rather than to the District Court. We do not decide in which of these state forums appellees should raise their claims, but assume, until the contrary is demonstrated, that they are not without a state forum.