dissenting:
The majority holds that there is no implied private right of action for a Native American to enforce a violation of the Indian employment preference requirement contained in 25 U.S.C. § 450(e)(b). Because the majority’s position is contrary to settled precedent and effectively eviscerates a statutory preference duly enacted by Congress, I dissent.
I.
The majority’s analysis gives lip service to, but fundamentally misapplies, the four-factor test set forth in Cort v. Ash, 422 U.S. 66, 95 S.Ct. 2080, 45 L.Ed.2d 26 (1975). Beginning with the first prong, there can be little question that § 450e was enacted for the benefit of a special class of which Solomon is a member. The Supreme Court’s language, in both Cort *1202and Cannon v. Univ. of Chicago, 441 U.S. 677, 99 S.Ct. 1946, 60 L.Ed.2d 560 (1979), specifies that the benefit must inhere to “a special class,” not “specific individuals in a special class,” in order to satisfy the test. The operative wording of § 450e does precisely that: It enacts a required employment and training preference in favor of Indian applicants. It strains common sense to construe this provision, otherwise than as referring to individual members of the special class of Indians. Accord Cannon, 441 U.S. at 690-93 & n. 13, 99 S.Ct. 1946 (distinguishing between statutes that speak to the general public benefit, or simply ban discriminatory conduct, and those that are “phrased in terms of the persons benefited”); see also First Pac. Bancorp, Inc. v. Heifer, 224 F.3d 1117, 1123 (9th Cir.2000) (“In our own cases, we have determined that the first factor of the Cort test is satisfied when there is an explicit reference to the individuals for whose benefit the statute was enacted.”).
I also disagree with the majority’s characterization of Mancari as somehow standing for the proposition that an employment preference is per se not an individual right. Simply put, this is not what the case holds. The majority takes one sentence out of context and characterizes it as the holding of the case. Mancari’s holding that Indian preference for employment in the Bureau of Indian Affairs was constitutional had nothing to do with whether the preference was a “right.” Rather, the Court held, in response to the argument that the Indian employment preference requirement was “invidious racial discrimination,” 417 U.S. at 551, 94 S.Ct. 2474, that it was not:
Contrary to the characterization made by appellees, this preference does not constitute “racial discrimination.” Indeed, it is not even a “racial” preference. Rather, it is an employment criterion reasonably designed to further the cause of Indian self-government and to make the BIA more responsive to the needs of its constituent groups- The preference, as applied, is granted to Indians not as a discrete racial group, but, rather, as members of quasi-sovereign tribal entities whose lives and activities are governed by the BIA in a unique fashion .... Here, the preference is reasonably and directly related to a legitimate, nonracially based goal. This is the principal characteristic that generally is absent from proscribed forms of racial discrimination.
Id. at 553, 94 S.Ct. 2474; see also id. at 553 n. 24, 94 S.Ct. 2474 (“The preference is not directed towards a “racial” group consisting of “Indians”; instead, it applies only to members of “federally recognized” tribes. This operates to exclude many individuals who are racially to be classified as “Indians.” In this sense, the preference is political rather than racial in nature.”). The Mancari Court was concerned with whether Indian employment preference constituted invidious discrimination on the basis of race. It did not address whether Indian employment preference conferred a right at all, much less a private right of action in court.
This, then, is the majority’s first mistake: The first Cort factor is patently not in “equipoise.” It weighs in favor of a private right of action.
As to the second Cort factor, I agree with the majority to this extent: The legislative history of § 450e is silent as to enforcement. However, my agreement ends there. The majority suggests that because (I) the legislative history is silent, (2) 25 U.S.C. § 450a expresses an intention to promote Indian self-governance, and (3) Congress provided a private right of action against federal officials explicitly in 25 U.S.C. § 450m-l, the second Cort factor— *1203legislative intent — weighs against implying a private right of action here. I disagree.
First, legislative silence is not enough, by itself, to support an inference against providing a private right of action. See Cannon, 441 U.S. at 694, 99 S.Ct. 1946 (“[I]n situations ... in which it is clear that federal law has granted a class of persons certain rights, it is not necessary to show an intention to create a private cause of action, although an explicit purpose to deny such cause of action would be controlling.”) (internal quotation marks omitted); see also First Pac. Bancorp, 224 F.3d at 1124 (“The absence of a statement of intent to create a remedy does not necessarily mean that no remedy is available. Indeed, if that were the case, the Supreme Court would not have developed a test for an implied private right of action”).
Second, I find the majority’s argument about the references to Indian self-governance in this context no more persuasive than its application to the first Cort factor. Section 450a does discuss the importance of Indian self-determination to “the Indian people as a whole” and “individual Indian tribes,” but it also speaks at the same time about ensuring an orderly transition “to effective and meaningful participation by the Indian people in the planning, conduct, and administration of’ programs and services. In fact, § 450e confers some benefits on “the Indian people as a whole” and others on individual Indian participants in the programs to be administered. It is simply disingenuous to suggest otherwise.
My analysis, therefore, is that the first Cort factor weighs in favor of a private right of action, and the second factor is silent. The law of this Circuit is that, when the first two Cort factors point in conflicting directions, analysis of the remaining two factors is warranted. Oliver v. Sealaska Corp., 192 F.3d 1220, 1224 (9th Cir.1999) (citing Touche Ross & Co. v. Redington, 442 U.S. 560, 575, 99 S.Ct. 2479, 61 L.Ed.2d 82 (1979)).
As to the third Cort factor, as already discussed above, § 450e has at least two important underlying legislative purposes. The text of § 450e(b) speaks directly to the furtherance of one of these purposes by requiring that, “to the greatest extent feasible ... preferences and opportunities for training and employment in connection with the administration of such contracts or grants shall be given to Indians.” While there can be no question that the promotion of tribal autonomy is another purpose of ISDEAA, the text of the statute itself perforce constrains tribal autonomy by imposing the Indian employment preference in the first place. It strains logic and common sense to say that the statutory text, duly enacted by Congress (and not under constitutional or other challenge here), is legitimate, but that the rights explicitly conferred by the statute may not be enforced in federal court because enforcing them, as opposed to merely conferring them, will burden tribal autonomy impermissibly.
This case presents an entirely different scenario from Santa Clara Pueblo v. Martinez, 436 U.S. 49, 98 S.Ct. 1670, 56 L.Ed.2d 106 (1978), on which the majority relies for the proposition that there can be no implied cause of action where finding one would arguably disserve another statutory purpose. At the heart of Santa Clara Pueblo was a question that could not lie closer to the core of the very concept of Indian self-determination: the right of Indian tribes to determine who is a member of a tribe and who is not. The Court held that there was no implied private right of *1204action under the Indian Civil Rights Act1 because creating a federal cause of action in this context “plainly would be at odds with the congressional goal of protecting tribal self-government.” Santa Clara Pueblo, 436 U.S. at 64, 98 S.Ct. 1670 (emphasis added). It is difficult to conceive of an issue more central to tribal self-government than the right of self-determination.
In this case, on the other hand, the majority concludes that the enforcement of a specific federal statutory employment preference so undermines tribal autonomy generally as to warrant eviscerating that preference by judicial fiat. In addition to violating a fundamental canon of statutory construction — the specific trumps the general, see Green v. Bock Laundry Machine Co., 490 U.S. 504, 524, 109 S.Ct. 1981, 104 L.Ed.2d 557 (1989)—this conclusion is simply wrong. If there is a balancing problem with undermining tribal autonomy here, that problem was resolved by Congress in enacting the employment preference. Allowing aggrieved individuals to sue to enforce that preference certainly does not undermine tribal self-government any more than conferring the preference in the first place. Rather, it simply gives effect to Congress’s mandate by allowing enforcement of the preference.
Enforcement of any statute that serves multiple purposes — as most do — will further some objectives while arguably hindering others. To say that no private right of action could ever be warranted in this context is extreme. While the Court in Santa Clara Pueblo may well have been correct that it is not the job of the courts to balance competing legislative objectives of equal weight, that scenario is not this case. Here, we have a broad statute with multiple purposes and a specific statutory provision that requires employment preferences for Indians to advance one of those purposes. A private right of action advances the purpose of § 450e.
Finally, the fourth Cort factor asks whether a federal remedy is appropriate in this context. As the district court in this case recognized, Congress has plenary authority over Native American affairs. See Alaska Chapter, Assoc. Gen. Contractors of Am. v. Pierce, 694 F.2d 1162, 1167 n. 4 (9th Cir.1982) (“The plenary power of Congress to deal with the special problems of Indians is drawn both explicitly and implicitly from the Constitution itself.”) (citing Mancari, 417 U.S. at 551-52, 94 S.Ct. 2474). Moreover, unlike Santa Clara Pueblo, this is not a case where anyone argues that enforcement of the statute is a matter entirely to be “left to the tribes”— HUD has promulgated the administrative enforcement scheme for NAHASDA block grants. Nor does it raise the Santa Clara Pueblo issue of whether core tribal issues should be removed from the tribal context and adjudicated in a federal forum.
Rather, this case raises the issue of whether a tribal housing authority and recipient of federal grant money is conducting its hiring procedures consistently with the requirements of a federal statute. The enforcement in federal court of federal statutes and federal grant schemes is appropriate.
I would hold that three of the four Cort factors weigh in favor of, not against, a private right of action here, and the fourth is silent. I would find a private right of action to enforce the Indian employment preference provision of § 450e.
*1205II.
My remedy would be to remand for consideration of Solomon’s claim. Solomon may not be able to proceed with his claim if he has not exhausted his administrative remedies prior to filing suit. The plain language of 24 C.F.R. § 1000.54 reflects that its administrative scheme is both mandatory and applicable to Solomon’s complaint. The title of the regulation is: (iWhat procedures apply to complaints arising out of any of the methods of providing for Indian preference?” The regulation goes on to specify:
(a) Each complaint shall be in writing, signed, and filed with the recipient.
(b) A complaint must be filed with the recipient no later than 20 calendar days from the date of the action (or omission) upon which the complaint is based.
(c) Upon receipt of a complaint, the recipient shall promptly stamp the date and time of receipt upon the complaint, and immediately acknowledge its receipt.
(d) Within 20 calendar days of receipt of a complaint, the recipient shall either meet, or communicate by.mail or telephone, with the complainant in an effort to resolve the matter. The recipient shall make a determination on a complaint and notify the complainant, in writing, within 30 calendar days of the submittal of the complaint to the recipient. The decision of the recipient shall constitute final administrative action on the complaint.
24 C.F.R. § 1000.54 (2002) (emphases added).2 See also Myers v. Bethlehem Shipbuilding Corp., 303 U.S. 41, 50-51, 58 S.Ct. 459, 82 L.Ed. 638 (1938) (noting “the long settled rule of judicial administration that no one is entitled to judicial relief for a supposed or threatened injury until the prescribed administrative remedy has been exhausted”); McKart v. United States, 395 U.S. 185, 193-95, 89 S.Ct. 1657, 23 L.Ed.2d 194 (1969); Marshall v. Burlington Northern, Inc., 595 F.2d 511, 513 (9th Cir.1979) (“It is well established in administrative law that before a federal court considers the question of an agency’s jurisdiction, sound judicial policy dictates that there be an exhaustion of administrative remedies.”); General Atomics v. United States Nuclear Regulatory Comm’n, 75 F.3d 536, 541 (9th Cir.1996).
Accordingly, as a threshold matter, the district court would have to determine whether Solomon had exhausted his administrative remedies prior to filing suit. The answer is unclear from the record now before us. If the district court finds exhaustion, it would then proceed to the merits of Solomon’s claim.
III.
I fully acknowledge that this case presents a question that is far from straightforward. The fundamental mandate to construe Indian-law statutes liberally in favor of Indians, Montana v. Blackfeet Tribe of Indians, 471 U.S. 759, 766, 105 S.Ct. 2399, 85 L.Ed.2d 753 (1985), does not provide a ready answer when, as here, Indian interests fall on both sides of the issue. However, in the case of the IS-DEAA, and specifically § 450e, Congress spoke clearly in explaining the balance of competing interests. The ISDEAA has *1206multiple purposes, but § 450e implements an employment preference, where feasible, in favor of Indian job applicants. Although not -explicitly authorized, a private right of action is consistent with and furthers the requirement of employment preference for Indians. The majority seeks to substitute its own judgment that tribal self-determination trumps Congress’s mandate of Indians’ right to job preference. Accordingly, I dissent.
. It should also be noted that the Court made this decision against the backdrop of a totally different legislative history and statutory structure than the one at issue here. 436 U.S. at 65-72, 98 S.Ct. 1670.
. The majority writes that we need not be concerned about finding no private right of action in federal court because Indian job applicants have this avenue of remediation open to them. Note that this "remedy” essentially consists of a formal request to the housing authority — the initial decisionmaker — to change its mind as to its own decision. The notion that this administrative option is somehow equivalent to, or a valid substitute for, a private right of action in court is simply wrong.