SAVE OUR VALLEY, Plaintiff-Appellant, v. SOUND TRANSIT (Central Puget Sound Regional Transit Authority); Bob White, Sound Transit Executive Director; Perry Weinberg, Sound Transit SEPA, Responsible Official; Transportation Dept U.S. Federal Transit Administration; Helen Knoll, Federal Transit Administration Regional Administrator Region X, Defendants-Appellees, Sound Transit SEPA Responsible Official, Defendant.
Save Our Valley, a community group, challenges the Central Puget Sound Regional Transit Authority's plan to build a light-rail line through the Rainier Valley south of Seattle, Washington. Save Our Valley argues that the project will have the effect of discriminating against Rainier Valley residents based on race in violation of a Department of Transportation regulation. The primary question before us is whether that Department of Transportation regulation creates an individual federal right that can be enforced under the Civil Rights Act, 42 U.S.C. § 1983. Because we conclude that the regulation does not create such a right, we affirm the district court's summary judgment.
The Central Puget Sound Regional Transit Authority (“Sound Transit”) is charged with building a light-rail line to connect the Northgate area in north Seattle with Sea-Tac Airport in Sea-Tac, Washington. The preferred twenty-one-mile route is proposed to pass through several Seattle neighborhoods, including south Seattle's Rainier Valley, a neighborhood populated predominantly by minority residents. The 4.6-mile segment through Rainier Valley is to be built at street level. Most of the segments through other neighborhoods are to be elevated above street level or to be built underground.
As pertinent to this appeal, Save Our Valley (“SOV”) filed suit under 42 U.S.C. § 1983 against Sound Transit alleging that the street-level alignment through Rainier Valley will cause disproportionate adverse impacts to minority residents, including the taking of residential and commercial properties, the displacement of community facilities, the disruption of businesses, and safety problems.1 SOV alleged that Sound Transit's plan violated a Department of Transportation “disparate impact” regulation-promulgated pursuant to Title VI of the Civil Rights Act of 1964, 78 Stat. 252, as amended, 42 U.S.C. § 2000d et seq.-that prohibits recipients of federal funds (like Sound Transit) from taking actions that have the effect of discriminating on the basis of race.2 SOV argued that this Department of Transportation regulation creates an individual federal right that SOV can enforce under 42 U.S.C. § 1983.
The district court disagreed that the regulation created such a right and granted summary judgment to Sound Transit. It then affirmed-without explanation-the clerk of court's taxation of $5,310.55 in costs against SOV as the losing party pursuant to Rule 54(d). This appeal followed.
The primary question in this appeal is whether the Department of Transportation's disparate-impact regulation creates an individual federal right that can be enforced through a § 1983 action. The answer to that specific question depends upon the answer to a more general question: Can a federal agency's regulations ever create individual rights enforceable through § 1983? We have never ruled on this fundamental question, which has divided our sister circuits. But because of controlling Supreme Court precedent, we hold that an agency regulation cannot create individual rights enforceable through § 1983.
Section 1983 creates a cause of action against any person who, acting under color of state law, abridges “rights, privileges, or immunities secured by the Constitution and laws” of the United States. 42 U.S.C. § 1983. The Supreme Court has held that only violations of rights, not laws, give rise to § 1983 actions. Gonzaga Univ. v. Doe, 536 U.S. 273, 285, 122 S.Ct. 2268, 153 L.Ed.2d 309 (2002); Blessing v. Freestone, 520 U.S. 329, 340, 117 S.Ct. 1353, 137 L.Ed.2d 569 (1997). This makes sense because § 1983 merely provides a mechanism for enforcing individual rights “secured” elsewhere, i.e., rights independently “secured by the Constitution and laws” of the United States. “One cannot go into court and claim a ‘violation of § 1983’-for § 1983 by itself does not protect anyone against anything.” Gonzaga, 536 U.S. at 285, 122 S.Ct. 2268 (quoting Chapman v. Houston Welfare Rights Organization, 441 U.S. 600, 617, 99 S.Ct. 1905, 60 L.Ed.2d 508 (1979)).
The Third, Fourth, and Eleventh Circuits have held that an agency regulation cannot create an individual federal right enforceable through § 1983. See S. Camden Citizens in Action v. New Jersey Dep't. of Envtl. Prot., 274 F.3d 771, 784, (3d. Cir.2001); Smith v. Kirk, 821 F.2d 980, 984 (4th. Cir.1987); Harris v. James, 127 F.3d 993, 1008 (11th Cir.1997). These courts reasoned the same way. They began by surveying the Supreme Court's § 1983 cases. In those cases, the courts noted, the Supreme Court's persistent focus was on tying the claimed right to Congress's intent (if any) to create the right. S. Camden Citizens in Action, 274 F.3d at 788; Harris, 127 F.3d at 1007. This focus on Congress's intent, paired with the Supreme Court's treatment of regulations as mere “administrative interpretations of the statute” in those cases, persuaded the courts that the Supreme Court's § 1983 jurisprudence is founded on the principle that Congress creates rights by statute, and that valid regulations merely “define” or “flesh out” the contents of those rights. S. Camden Citizens in Action, 274 F.3d at 790; Harris, 127 F.3d at 1008-09.3
On the other side of the circuit split, the District of Columbia and Sixth Circuits have held that an agency regulation can create an individual federal right. See Samuels v. District of Columbia, 770 F.2d 184 (D.C.Cir.1985); Loschiavo v. City of Dearborn, 33 F.3d 548 (6th Cir.1994). These courts' holdings were based on a broad reading of the Supreme Court's decision in Maine v. Thiboutot, 448 U.S. 1, 100 S.Ct. 2502, 65 L.Ed.2d 555 (1980), the first to hold that a violation of statutory rights may be remedied through § 1983. Samuels, 770 F.2d at 199. Although Thiboutot involved the violation of a statute, not a regulation, the D.C. Circuit (and, tacitly, the Sixth Circuit) reasoned that Thiboutot 's broad analysis of the “ ‘laws' clause” of § 1983 suggested that § 1983 could be used to remedy violations of all valid federal laws, including regulations. Id.
The Supreme Court has never addressed this issue directly, so no single Supreme Court precedent controls our decision in this case. Nonetheless, we begin our analysis with two recent Supreme Court decisions-Alexander v. Sandoval, 532 U.S. 275, 121 S.Ct. 1511, 149 L.Ed.2d 517 (2001), and Gonzaga Univ. v. Doe, 536 U.S. 273, 122 S.Ct. 2268, 153 L.Ed.2d 309 (2002)-that are important because they have strengthened the legal foundation underlying the Third, Fourth, and Eleventh Circuits' holdings and eroded the legal foundation underlying the D.C. and Sixth Circuits' holdings. To evaluate the various circuit courts' holdings (as we do below), one must consider them in the new context of Sandoval and Gonzaga.
In Sandoval, the Court considered a challenge to the Alabama Department of Public Safety's official policy of administering its driver's license examination only in English as violative of Title VI and its implementing regulations. See 532 U.S. at 278-79, 121 S.Ct. 1511. The plaintiffs claimed that the implementing regulations-which were § 602 disparate-impact regulations virtually identical to those in this appeal-created a private right of action. Id. The Court rejected the claim, basing its analysis not on the regulation 's text but on the statute 's text. Id. at 293, 121 S.Ct. 1511. The Court held that only Congress by statute can create a private right of action. Id. at 291, 121 S.Ct. 1511.
Although the Sandoval Court addressed only one kind of federal right-implied rights of action-its reasoning has broader implications. The Court suggested that only Congress by statute can create individual rights of any kind (including, we conclude, rights enforceable through § 1983). Even though the plaintiff alleged that the disparate-impact regulations created the claimed right, the Court never performed any analysis of the regulations themselves (as SOV would have us do in this case). The Court wrote:
The judicial task is to interpret the statute Congress has passed to determine whether it displays an intent to create not just a private right but also a private remedy. Statutory intent on this latter point is determinative.
532 U.S. at 286-87, 121 S.Ct. 1511 (citations omitted) (emphasis added).
Language in a regulation may invoke a private right of action that Congress through statutory text created, but it may not create a right that Congress has not.
532 U.S. at 291, 121 S.Ct. 1511 (emphasis added).
[I]t is most certainly incorrect to say that language in a regulation can conjure up a private cause of action that has not been authorized by Congress. Agencies may play the sorcerer's apprentice but not the sorcerer himself.
532 U.S. at 291, 121 S.Ct. 1511.
These statements refer to the creation of implied rights of action, rather than to the creation of individual rights enforceable through § 1983. But the Court's reasoning applies equally to both kinds of rights. Both implied rights of action and rights enforceable through § 1983 are creatures of federal substantive law. And it is an elementary principle of constitutional law that lawmaking is the province of Congress. As the Court stated in Sandoval, “Like substantive federal law itself, private rights of action to enforce federal law must be created by Congress.” 532 U.S. at 286, 121 S.Ct. 1511 (emphasis added). Cf. U.S. Const. Art. I, Section 1 (“All legislative powers herein granted shall be vested in a Congress of the United States”); Whitman v. Am. Trucking Ass'ns, Inc., 531 U.S. 457, 472, 121 S.Ct. 903, 149 L.Ed.2d 1 (2001) (“Article I, § 1, of the Constitution vests ‘[a]ll legislative Powers herein granted ․ in a Congress of the United States.’ This text permits no delegation of those powers.”); Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 587, 72 S.Ct. 863, 96 L.Ed. 1153 (1952) (“In the framework of our Constitution, the President's power to see that the laws are faithfully executed refutes the idea that he is to be a lawmaker.”) (quoted with approval in INS v. Chadha, 462 U.S. 919, 954 n. 16, 103 S.Ct. 2764, 77 L.Ed.2d 317 (1983)); A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495, 529, 55 S.Ct. 837, 79 L.Ed. 1570 (1935) (“The Congress is not permitted to abdicate or to transfer to others the essential legislative functions with which it is thus vested.”). Individual rights enforceable through § 1983-like implied rights of action-are creatures of substantive federal law; therefore, they must be created by Congress.4
Any doubt that may have remained as to the genesis of individual rights enforceable through § 1983 after Sandoval was eliminated by the Supreme Court's Gonzaga decision. In Gonzaga, a decision that held that the Family Educational Rights and Privacy Act was not privately enforceable through § 1983, the Court confirmed that individual rights enforceable through § 1983 and implied private rights of action are similar in respects relevant to this appeal:
We ․ reject the notion that our implied right of action cases are separate and distinct from our § 1983 cases. To the contrary, our implied right of action cases should guide the determination of whether a statute confers rights enforceable under § 1983.
536 U.S. at 283, 122 S.Ct. 2268. The Court continued:
We have recognized that whether a statutory violation may be enforced through § 1983 “is a different inquiry than that involved in determining whether a private right of action can be implied from a particular statute.” But the inquiries overlap in one meaningful respect-in either case we must first determine whether Congress intended to create a federal right.
Id. (citation omitted). As in Sandoval, the Court here suggested that federal rights are created by Congress through statutes, not by agencies through regulations. More importantly for our purposes, the Court's reasoning strongly supports our view that individual rights enforceable through § 1983 are similar to implied rights of action in the important respect that both are federal substantive law and that in each case courts are required to “determine whether Congress intended to create a federal right.” Gonzaga, 536 U.S. at 283, 122 S.Ct. 2268 (emphasis added).5 Since only Congress can create implied rights of action (as the Court held in Sandoval ), the Court's Gonzaga holding suggests that only Congress can create rights enforceable through § 1983.
Other Supreme Court decisions also have focused squarely on Congress's intent to create individual rights. See Blessing, 520 U.S. at 341, 117 S.Ct. 1353 (concentrating on Congress's intent to create rights in a statute enforceable through § 1983); Suter v. Artist M., 503 U.S. 347, 357, 112 S.Ct. 1360, 118 L.Ed.2d 1 (1992) (same); Wilder v. Va. Hosp. Ass'n, 496 U.S. 498, 509, 110 S.Ct. 2510, 110 L.Ed.2d 455 (1990) (same); Golden State Transit Corp. v. City of Los Angeles, 493 U.S. 103, 110-11, 110 S.Ct. 444, 107 L.Ed.2d 420 (1989) (same); Wright, 479 U.S. at 430, 107 S.Ct. 766 (same). As the Third Circuit observed, “[T]he Supreme Court [has] refined its analysis to focus directly on Congress' intent to create enforceable rights and to confine its holdings to the limits of that intent.” S. Camden Citizens in Action, 274 F.3d at 784 (holding that § 602 regulations are not enforceable through § 1983).
We believe the Supreme Court's Sandoval and Gonzaga decisions, taken together, compel the conclusion we reach today: that agency regulations cannot independently create rights enforceable through § 1983. Our conclusion should surprise no one, as it results directly from the broader, venerated constitutional law principle that Congress, rather than the executive, is the lawmaker in our democracy.
SOV relies primarily on the Supreme Court's decision in Wright v. City of Roanoke Redevelopment and Housing Authority, 479 U.S. 418, 107 S.Ct. 766, 93 L.Ed.2d 781 (1987), for the proposition that valid federal regulations may create rights enforceable through § 1983. In Wright, the plaintiffs alleged that a local housing authority violated a federal statute (imposing a rent ceiling) and the statute's implementing regulations (requiring public housing authorities to include a reasonable utility allowance in tenants' rent). See id. at 419, 107 S.Ct. 766. The Supreme Court held that the plaintiffs could maintain a suit under § 1983 for violation of the statute and the HUD regulations. See id. at 431, 107 S.Ct. 766. This holding, according to SOV, must mean the HUD regulations created enforceable rights.
We disagree. The mere fact that the Wright Court permitted a plaintiff to maintain a § 1983 suit for violation of the regulations does not necessarily mean the regulations created the claimed right. Rather, it seems that the Court understood that the statute, rather than the HUD regulations, created the right. The Court looked to the regulations only to interpret the scope of the right that Congress had conferred through the statute. This conclusion is evident in the Court's focus on Congress's intent in enacting the statute, rather than on the agency's intent in promulgating the regulations. The Court referred to “the benefits Congress intended to confer on tenants.” 479 U.S. at 432, 107 S.Ct. 766 (emphasis added). It also stated that “HUD's view is entitled to deference as a valid interpretation of the statute.” Id. at 430, 107 S.Ct. 766 (emphasis added). As the Third Circuit stated (in a case similar to this one):
[T]he Wright Court located the alleged right in the statutory provision and then relied upon the implementing regulations to define and interpret that right ․ Wright does not hold that a regulation alone-i.e., where the alleged right does not appear explicitly in the statute, but only appears in the regulation-may create an enforceable federal right.
S. Camden Citizens in Action, 274 F.3d at 783. The Third Circuit's analysis of Wright echoed an earlier decision by the Eleventh Circuit: “We conclude that the Wright majority did not hold that federal rights are created either by regulations ‘alone’ or by any valid administrative interpretation of a statute creating some enforceable right.” Harris, 127 F.3d at 1008.
Moreover, as the four dissenting Justices observed in Wright, the Wright majority did not reach the question of whether an agency regulation could create a right. Plus, the dissenters noted, the suggestion that a regulation could create a right would have been “troubling.” Justice O'Connor, joined by Chief Justice Rehnquist, Justice Powell, and Justice Scalia, wrote:
In the absence of any indication in the language, legislative history, or administrative interpretation of the Brooke Amendment that Congress intended to create an enforceable right to utilities, it is necessary to ask whether administrative regulations alone could create such a right. This is a troubling issue not briefed by the parties, and I do not attempt to resolve it here ․ I am concerned ․ that lurking behind the Court's analysis may be the view that, once it has been found that a statute creates some enforceable right, any regulation adopted within the purview of the statute creates rights enforceable in federal courts, regardless of whether Congress or the promulgating agency ever contemplated such a result. Thus, HUD's frequently changing views on how best to administer the provision of utilities to public housing tenants becomes the focal point for the creation and extinguishment of federal “rights.” Such a result, where determination of § 1983 “rights” has been unleashed from any connection to congressional intent, is troubling indeed.
Wright, 479 U.S. at 437-38, 107 S.Ct. 766 (O'Connor, J., dissenting).
Our view that Wright does not stand for the proposition that regulations can create federal rights seems even more persuasive in light of the Supreme Court's subsequent cases. The Court's opinion in Sandoval (where the Wright dissenters were in the majority), for example, shows a dogged focus on Congress's intent to create federal rights. At the same time, it shows an utter neglect of the intent underlying administrative regulations.
SOV cites two post-Wright Supreme Court decisions as confirming its understanding of Wright. In Wilder v. Virginia Hospital Assocation, 496 U.S. 498, 110 S.Ct. 2510, 110 L.Ed.2d 455 (1990), the Court stated that “in Wright ․ we found that the[statute] and its implementing regulations did create rights enforceable under § 1983.” Id. at 511, 110 S.Ct. 2510 (emphasis added). The Court also noted that the rights created were sufficiently specific to satisfy the specificity prong of the three-prong Blessing v. Freestone test (the Supreme Court's test for determining whether a federal statute creates individual federal rights enforceable under § 1983) “[b]ecause the regulations set out guidelines for the housing authorities to follow.” Id. at 511-12, 110 S.Ct. 2510. And in Suter v. Artist M., 503 U.S. 347, 112 S.Ct. 1360, 118 L.Ed.2d 1 (1992), the Court stated that in Wright a “statute ․ in conjunction with regulations ․ created rights under § 1983.” Id. at 362 n. 13, 112 S.Ct. 1360.
Although this language suggests some role for agency regulations in applying the three-prong Blessing test to statutes,6 it does not mean that an agency regulation alone can create a federal right. Rather, the language is consistent with the view that “so long as the statute itself confers a specific right upon the plaintiff, and a valid regulation merely further defines or fleshes out the content of that right, then the statute-‘in conjunction with the regulation’-may create a federal right as further defined by the regulation.” Harris, 127 F.3d 993 at 1009. In any case, the Court's more recent decisions, reasoning that right-making is the exclusive province of Congress, dispel any doubt raised by the language in Wilder and Suter.7
Next, SOV urges that our decision in Buckley v. City of Redding, 66 F.3d 188 (9th Cir.1995), requires us to rule in its favor. SOV argues that Buckley holds that where a regulation “alone satisfies all three requirements [of the Supreme Court's Blessing test], the regulation by itself creates the enforceable right.” But Buckley does not support this broad proposition. The issue in Buckley was whether the Federal Aid in Sport Fish Restoration Act-not its enabling regulations-created an enforceable federal statutory right under § 1983. See Buckley, 66 F.3d at 189-90. We began our opinion, “In this case, we must determine whether the Federal Aid in Sport Fish Restoration Act (‘the Act’) confers a right enforceable under 42 U.S.C. § 1983.” Id. at 189 (citations omitted). “We hold that the Act does confer rights enforceable under section 1983.” Id. (emphasis added). Although we referred to the regulations,8 we were careful to emphasize that the regulations were interpretive. Moreover, consistent with Supreme Court precedent, we focused our inquiry squarely on Congress: “In a section 1983 action, the plaintiff must establish that Congress created an enforceable statutory right.” Id. (emphasis added). In other words, rather than contradicting the view that only Congress can create a right enforceable under § 1983, our opinion in Buckley supports it.
Finally, SOV points out that the D.C. and Sixth Circuits have held that regulations can create individual rights. See Samuels, 770 F.2d at 199; Loschiavo, 33 F.3d at 551. In Samuels, tenants of a federally funded public housing project filed suit against local public housing officials to enforce the United States Housing Act and HUD regulations. See 770 F.2d at 188. The D.C. Circuit opined:
[P]laintiffs clearly allege that the District's public housing officials have violated the applicable HUD regulations, and that allegation alone, we think, states a cognizable section 1983 claim under the circumstances of this case. HUD's grievance procedure regulations clearly have the full force and effect of federal law: they are issued under a congressional directive to implement specific statutory norms and they affect individual rights and obligations․ While Thiboutot involved a statutory violation, the Court's broad analysis of the “laws” clause of section 1983 indicates that section 1983 provides a legal remedy for the violation of all valid federal laws, including at least those federal regulations adopted pursuant to a clear congressional mandate that have the full force and effect of law. Such regulations have long been recognized as part of the body of federal law.
Id. at 199.
The D.C. Circuit's reasoning is flawed. The D.C. Circuit assumed (and misread the Supreme Court's decision in Thiboutot to mean) that the mere fact that a statute or regulation is a “law” within the meaning of § 1983 makes it capable of creating rights. The court stated that “substantive federal regulations issued under Congress' mandate constitute ‘laws' within the meaning of section 1983. We therefore hold that the plaintiffs state a valid section 1983 claim.” Samuels, 770 F.2d at 199. The D.C. Circuit skipped a step. For a statute or regulation to be enforced through § 1983, more is needed than for the statute or regulation to be a “law.” The statute or regulation must be a “law” and it must secure “rights, privileges, or immunities.” 42 U.S.C. § 1983. See also Gonzaga, 536 U.S. at 285, 122 S.Ct. 2268 (“[Section] 1983 merely provides a mechanism for enforcing rights ‘secured’ elsewhere.”). And the Supreme Court has all but held that only certain kinds of laws-not including regulations-can create such rights.
Second, the D.C. Circuit's opinion is founded on an out-dated understanding of the relationship between implied rights of action and rights enforceable under § 1983. The D.C. Circuit wrote, “[S]tatutory section 1983 claims differ significantly from implied rights of action.” Samuels, 770 F.2d at 194. This language is difficult to square with the Supreme Court's recent holding “reject[ing] the notion that our implied right of action cases are separate and distinct from our § 1983 cases.” Gonzaga, 536 U.S. at 283, 122 S.Ct. 2268. Moreover, the D.C. Circuit explained that the critical difference between § 1983 and implied right of action cases is their different burdens of proof. Section 1983 plaintiffs-unlike implied right of action plaintiffs-“do not suffer the burden of demonstrating that Congress specifically intended to preserve the ability of private parties to enforce the relevant provisions of federal law against those officials.” Id. But the court failed to mention that a plaintiff is not entitled to this presumption until after the plaintiff has made the threshold showing that Congress intended to create a federal right. As the Supreme Court would later explain,
Plaintiffs suing under § 1983 do not have the burden of showing an intent to create a private remedy because § 1983 generally supplies a remedy for the vindication of rights secured by federal statutes. Once a plaintiff demonstrates that a statute confers an individual right, the right is presumptively enforceable by § 1983. But the initial inquiry-determining whether a statute confers any right at all-is no different from the initial inquiry in an implied right of action case, the express purpose of which is to determine whether or not a statute “confer[s] rights on a particular class of persons.”
Gonzaga, 536 U.S. at 284-85, 122 S.Ct. 2268 (citations omitted) (emphasis added). Thus, the D.C. Circuit again skipped a step: the “initial inquiry” that is the central question in this appeal.
The Sixth Circuit's reasoning in Loschiavo tracks the D.C. Circuit's reasoning in Samuels, but it is far more conclusory. The Sixth Circuit's analysis consisted of two sentences:
In Maine v. Thiboutot, 448 U.S. 1, 6-8, 100 S.Ct. 2502, 2505-06, 65 L.Ed.2d 555 (1980), the Supreme Court recognized that plaintiffs may use § 1983 to enforce not only constitutional rights, but also those rights defined by federal statutes. As federal regulations have the force of law, they likewise may create enforceable rights.
Loschiavo, 33 F.3d at 551. The Sixth Circuit's holding suffered from the same flaw as the D.C. Circuit's holding. It assumed that if a regulation is a “law” it must be the kind of law enforceable through § 1983.9
A final problem with the Sixth and D.C. Circuits' opinions is that they predated the Supreme Court's decisions in Sandoval and Gonzaga. As we explained above, Sandoval and Gonzaga, taken together, compel the conclusion that only Congress can create an individual federal right. We therefore reject the approach of the Sixth and D.C. Circuits and hold that an agency regulation cannot create an individual federal right.
To summarize the principles we hold apply in this case: Violations of rights, not violations of laws, give rise to § 1983 actions. Gonzaga, 536 U.S. at 285, 122 S.Ct. 2268; Blessing, 520 U.S. at 340, 117 S.Ct. 1353. Plaintiffs suing under § 1983 must demonstrate that a statute-not a regulation-confers an individual right.10 S. Camden Citizens in Action, 274 F.3d at 781; Harris, 127 F.3d at 1008-09. As an agency interpretation of a statute, a regulation may be relevant in determining the scope of the right conferred by Congress. S. Camden Citizens in Action, 274 F.3d at 783. Agency regulations therefore may be considered in applying the three-prong Blessing test. See Buckley, 66 F.3d at 190. But the inquiry must focus squarely on Congress's intent. The paramount consideration is to determine if Congress intended to create the particular federal right sought to be enforced. See Suter, 503 U.S. at 357, 112 S.Ct. 1360; S. Camden Citizens in Action, 274 F.3d at 788.
In this case, our analysis begins and ends with Congress's intent. The Supreme Court already has held that Congress never intended to create the right SOV claims, the right to be free from racially disparate effects. The Court has often repeated that “Title VI itself directly reach[es] only instances of intentional discrimination.” Sandoval, 532 U.S. at 281, 121 S.Ct. 1511; Alexander v. Choate, 469 U.S. 287, 293, 105 S.Ct. 712, 83 L.Ed.2d 661 (1985). Because Title VI does not create the right the plaintiffs seek to enforce, we affirm the district court's judgment dismissing the action.11
A separate, alternative line of analysis requires us to hold that SOV cannot enforce the disparate-impact regulation. Even if a regulation in general could create an individual federal right enforceable through § 1983, it is plain that the disparate-impact regulation at issue here does not create such a right.
As we explained above, the disparate-impact regulation was promulgated by the Department of Transportation based on authority granted it by Congress in Title VI. In § 601 of that title, Congress created a right to be free from intentional discrimination based on race. In § 602, Congress authorized federal agencies to “effectuate the provisions of [§ 601] ․ by issuing rules, regulations, or orders of general applicability.” 42 U.S.C. § 2000d 1. Congress in § 602 did not authorize federal agencies to create new rights. As the Supreme Court held in Sandoval, “[f]ar from displaying congressional intent to create new rights, § 602 limits agencies to ‘effectuat [ing]’ rights already created by § 601.” 532 U.S. at 289, 121 S.Ct. 1511 (2001). The disparate-impact regulation cannot create a new right; it can only “effectuate” a right already created by § 601.And § 601 does not create the right that SOV seeks to enforce, the right to be free from racially discriminatory effects. Sandoval, 532 U.S. at 280, 121 S.Ct. 1511 (“[I]t is ․ beyond dispute-and no party disagrees-that § 601 prohibits only intentional discrimination.”). A regulation cannot “effectuate” a statutory right by creating a new and different right. See Cal. Cosmetology Coalition v. Riley, 110 F.3d 1454, 1460 (9th Cir.1997) (“A regulation may not serve to amend a statute, nor add to the statute ‘something which is not there.’ ”). On this alternative ground, SOV's claim must fail.
SOV asserts that the district court abused its discretion by awarding $5,310.55 in costs to Sound Transit.12 Federal Rule of Civil Procedure 54(d)(1) provides that “costs other than attorneys' fees shall be allowed as of course to the prevailing party unless the court otherwise directs.” Rule 54(d) creates a presumption for awarding costs to prevailing parties; the losing party must show why costs should not be awarded. Stanley v. Univ. of Southern California, 178 F.3d 1069, 1079 (9th Cir.1999); Nat'l Info. Servs., Inc. v. TRW, Inc., 51 F.3d 1470, 1471-72 (9th Cir.1995).
In Stanley, we held that the district court abused its discretion in rejecting a losing civil rights plaintiff's motion to deny costs to the defendant without considering: (1) the plaintiff's limited financial resources and (2) the chilling effect on future civil rights litigants of imposing high costs. Stanley, 178 F.3d at 1079. SOV argues that Stanley 's requirement that the district court “consider” the two Stanley factors means that the court must specify reasons for awarding costs to the prevailing party when a civil rights plaintiff puts those considerations at issue.
Although a district court must “specify reasons” for its refusal to tax costs to the losing party, Assoc. of Mexican-American Educators v. California, 231 F.3d 572, 591 (9th Cir.2000); Subscription Television, Inc. v. Southern Cal. Theatre Owners Ass'n, 576 F.2d 230, 234 (9th Cir.1978), we have never held that a district court must specify reasons for its decision to abide the presumption and tax costs to the losing party. The distinction is critical. A district court deviates from normal practice when it refuses to tax costs to the losing party, and that deviation triggers the requirement to “specify reasons.” As we explained in Association of Mexican-American Educators,
The requirement that district courts give reasons for denying costs flows logically from the presumption in favor of costs that is embodied in the text of the rule; if a district court wishes to depart from that presumption, it must explain why so that the appellate court will be able to determine whether or not the trial court abused its discretion․ Our requirement that a district court give reasons for denying costs is, in essence, a requirement that the court explain why a case is not ordinary.
Assoc. of Mexican-American Educators v. California, 231 F.3d at 592-93 (internal quotation marks omitted). This reasoning suggests, as we hold today, that a district court need not give affirmative reasons for awarding costs; instead, it need only find that the reasons for denying costs are not sufficiently persuasive to overcome the presumption in favor of an award. The presumption itself provides all the reason a court needs for awarding costs, and when a district court states no reason for awarding costs, we will assume it acted based on that presumption. Stanley only held that, in the rare occasion where severe injustice will result from an award of costs (such as the injustice that would result from an indigent plaintiff's being forced to pay tens of thousands of dollars of her alleged oppressor's legal costs), a district court abuses its discretion by failing to conclude that the presumption has been rebutted. No such injustice will result from the award of $5,310.55 in this case.
SOV observes that we have approved several factors that would justify a district court's refusal to award costs to a prevailing party: the losing party's limited financial resources, see National Org. for Women v. Bank of Cal., 680 F.2d 1291, 1294 (9th Cir.1982); misconduct on the part of the prevailing party, see National Info. Servs., 51 F.3d at 1472; the importance of the issues, Assoc. of Mexican-American Educators v. California, 231 F.3d 572, 593 (2000); the importance and complexity of the issues, Id.; the merit of the plaintiff's case, even if the plaintiff loses, Id.; and the chilling effect on future civil rights litigants of imposing high costs. Stanley, 178 F.3d at 1079. SOV argues that these factors weigh in its favor: SOV is a nonprofit organization with limited resources. It has raised issues of great public importance, both to the local community and to civil rights plaintiffs everywhere. The legal questions it has raised are close and complex; indeed, the circuits are split over a major question raised by this appeal. These factors would have justified the district court's decision to deny costs to the prevailing party, had the district court exercised its discretion in that manner. But that does not mean that the district court had to state the reasons that it awarded costs.
The district court might have believed that this relatively small sum-$5,310.55-would not “chill” future civil rights litigation, see Assoc. of Mexican-American Educators, 231 F.3d at 593 (affirming the district court's denial of $216,443.67 in costs to a prevailing defendant); that the prevailing party had engaged in no misconduct that should disqualify it from costs; or that there were other reasons to award costs. But the district court needs no affirmatively expressed reason to tax costs. Rather, it need only conclude that the reasons advanced by the party bearing the burden-the losing party-are not sufficiently persuasive to overcome the presumption. In the circumstances of this case, the presumption itself provided an adequate reason for the district court to award costs. We decline to adopt a rule that would place on district courts the burden of justifying routine awards of costs against losing parties in civil rights cases.
I write separately to clarify two critical issues that the majority has utterly confused: the distinction between rights and rights of action, and the consequence of its assumption that the regulation here is valid. The majority's confusion on the first issue taints its entire opinion and causes it to conclude, incorrectly, that regulations may never create rights enforceable under 42 U.S.C. § 1983. Its confusion on the second issue leads it to conduct the wrong analysis regarding whether the particular regulation here creates a right to be free of disparate impact discrimination. In the end, however, I conclude that under the last relevant Supreme Court opinion, Gonzaga Univ. v. Doe, 536 U.S. 273, 122 S.Ct. 2268, 153 L.Ed.2d 309 (2002), there is simply no escaping the conclusion that the majority's bottom line is correct: The disparate impact regulation-although in my view indubitably a valid legislative regulation-does not create a “right” within the meaning of § 1983, so no cause of action lies under that statute.
The central issue in this appeal is whether the DOT regulation creates a right, enforceable in a private action under § 1983, to be free of disparate impact discrimination. Resolving this issue requires us to answer two distinct questions: first, whether the regulation creates a right; and second, if so, whether § 1983 permits enforcement of that right. The majority answers the first question in the negative because it believes a regulation can never create a right. Yet the majority opinion demonstrates that it does not understand what a right is, and how it differs from a right of action. That elision of two distinct concepts results in the wrong answer to the general question whether agency regulations can create legal relationships properly described as “rights.”
A legal right is an entitlement that inheres in an individual and enables her to make certain demands of other individuals, which demands are backed by the coercive power of the state. It is thus a tripartite relationship between one individual and another and the state.1 This relationship has been alternately described as: a social duty owed from one person to another, see 1 William Blackstone, Commentaries 117 (photo. reprint 1992) (1765); a restriction on each individual's freedom, backed by the coercive power of the state, that harmonizes the individual's freedom with the freedom of everyone else in the community, see Immanuel Kant, “On the Relationship of Theory to Practice in Political Right,” in Kant's Political Writings, 73, 73 (H. Reiss ed.1999) (1793); “a permission to exercise certain natural powers, and upon certain conditions to obtain protection, restitution, or compensation by the aid of the public force,” Oliver Wendell Holmes, Jr., The Common Law, 214 (Dover, 1991) (1881) (emphasis added); a capacity of influencing the acts of another, see Thomas Erskine Holland, Jurisprudence 78 (1908) (emphasis added), or of asserting a claim, see IV Roscoe Pound, Jurisprudence § 118 at 70 (1959) (emphasis added); and, a political trump held by an individual that supercedes a collective goal which might otherwise justify denying the individual what she wishes or imposing some loss upon her, see Ronald Dworkin, Taking Rights Seriously, xi (1977). While each description is linked to a specific conception about the world (e.g., Kant) or the nature of law (e.g., Dworkin), all share the basic idea that a right is a relationship between two individuals and the state.
For present purposes, the essential point is that a “right” is not the same as the authority to bring an enforcement action in court. To the contrary, a cause of action is a specific type of remedy, a procedural vehicle for redressing a violation of a right. Some rights may not be enforceable through such an affirmative remedy in court, and others may not be enforceable in court at all.
This distinction has found expression innumerable times in innumerable ways. The Declaration of Independence is eloquent on this point. It states that:
[A]ll men are created equal, that they are endowed by their Creator with certain unalienable Rights, that among these are Life, Liberty and the pursuit of Happiness. That to secure these rights, Governments are instituted among Men ․, That whenever any Form of Government becomes destructive of these ends, it is the Right of the People to alter or to abolish it, and to institute a new Government, laying its foundation on such principles and organizing its powers in such form, as to them shall seem most likely to effect their Safety and Happiness.
Id. at para. 2. Explicit in this statement are the ideas that rights (at least, the natural rights to life, liberty and the pursuit of happiness) derive from a source independent of the state, and that it is the principal function of governments to “secure these rights.” One method of securing rights, of course, is to provide civil remedies for their violation. The Declaration of Independence is just as explicit, however, in its statement that the specific rights mentioned are “enforceable” even in the absence of civil remedies: they are enforceable by insurrection.
The Bill of Rights provides examples of more direct pertinence.2 The First, Fourth and Seventh Amendments speak respectively of “the right of the people” to peaceably assemble and petition the government for redress of grievances; be secure against unreasonable searches and seizures; and be tried by a jury. See U.S. Const. amends. I, IV, VII. In each case, the text of the amendment suggests that the people possessed the right in question prior to the existence of the central government. The amendments are worded as restrictions on the government's ability to invade rights that the amendment assumes people already have. See e.g. U.S. Const. amend. I (“Congress shall make no law ․ abridging ․ the right of the people peaceably to assemble․”); U.S. Const. amend. IV (“The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated․”). The Ninth Amendment states this expressly. That amendment provides that “[t]he enumeration in the Constitution, of certain rights, shall not be construed to deny or disparage others retained by the people.” U.S. Const. amend. IX. By implication, all of these rights exist prior to and apart from any means of privately redressing their violation, whether that means is through the judiciary or otherwise.
Indeed, the rights enumerated in the Constitution have, since the beginning of the Republic, had significance unrelated to private actions to enforce them. Such rights have been understood to order the relationship between individuals and the government. The Fourth Amendment, for example, provides that the “right of the people to be secure in their persons, houses, papers and effects, against unreasonable searches and seizures, shall not be violated.” U.S. Const. amend. IV. This right limits the government's authority and thus restrains the actions of those individuals who act on its behalf (law enforcement officers) by preventing them from detaining or searching other individuals absent probable cause. See Wayne R. LaFave, Search and Seizure, § 1.1(a) (3d ed.1996). Possessing the “right” against unreasonable searches and seizures provides individuals with the assurance that they may go about their activities unmolested by other individuals acting on behalf of the government. Similarly, the Fifth Amendment provides that “[n]o person ․ shall be compelled in any criminal case to be a witness against himself.” U.S. Const. amend. V. While there has been much scholarly debate over the meaning of this particular provision,3 all of the commentators agree that it in some way limits what one individual may do to another on behalf of the sovereign.
These rights are not merely abstract or ethereal. They have real-life consequences, in and out of court, independent of any individual's ability to vindicate them in a private, affirmative suit in court. Legislatures are obliged to respect them when legislating, and executive officials are obliged to respect them in enforcing the law.
Concomitantly, rights have consequences in court proceedings even absent an affirmative enforcement action. There are evidentiary consequences, for example. Evidence seized in violation of the Fourth Amendment is inadmissible in court, see Mapp v. Ohio, 367 U.S. 643, 655-56, 81 S.Ct. 1684, 6 L.Ed.2d 1081 (1961) and Weeks v. United States, 232 U.S. 383, 398, 34 S.Ct. 341, 58 L.Ed. 652 (1914), as is evidence seized in a manner that “shocks the conscience,” see Rochin v. California, 342 U.S. 165, 72 S.Ct. 205, 96 L.Ed. 183 (1952) (violation of Fourteenth Amendment right to due process of law). The Fifth Amendment precludes a prosecutor from forcing a defendant to testify, see Malloy v. Hogan, 378 U.S. 1, 6, 84 S.Ct. 1489, 12 L.Ed.2d 653 (1964), or from commenting on the defendant's silence, see Griffin v. California, 380 U.S. 609, 613, 85 S.Ct. 1229, 14 L.Ed.2d 106 (1965). The Fourteenth Amendment requires the prosecution in a criminal matter to prove its case beyond a reasonable doubt. See In re Winship, 397 U.S. 358, 364, 90 S.Ct. 1068, 25 L.Ed.2d 368 (1970). In each instance, the right puts specific limits on what evidence the court may receive in a criminal trial, without regard to whether the affected individual may obtain any other form of redress for violation of his right.
Also, courts have long understood that violation of a right may compel a court to invalidate an Act of Congress. In Boyd v. United States, 116 U.S. 616, 638, 6 S.Ct. 524, 29 L.Ed. 746 (1886), for example, the Court held unconstitutional a federal statute that permitted what the Court deemed to be an “unreasonable” seizure of a person's books and papers.
Further, the rights enumerated in the Constitution not only restrict individuals acting on behalf of the government but can also require affirmative conduct of government actors. As now understood, the Fifth Amendment, for example, requires law enforcement officers to inform every person taken into custody of the rights that Amendment guarantees. See Dickerson v. United States, 530 U.S. 428, 440, 120 S.Ct. 2326, 147 L.Ed.2d 405 (2000) (holding that Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966), and the prophylactic warnings it requires, are “constitutionally based”). The Sixth Amendment requires that states provide attorneys to indigent defendants in every criminal case (in which the potential punishment is greater than six months). See Gideon v. Wainwright, 372 U.S. 335, 344-45, 83 S.Ct. 792, 9 L.Ed.2d 799 (1963); Powell v. Alabama, 287 U.S. 45, 68-69, 53 S.Ct. 55, 77 L.Ed. 158 (1932). These rights, true, are enforceable in court, but police officers and states have an obligation to honor those rights whether judicial process is invoked or not.
The Supreme Court's third-party standing cases provide another excellent illustration of this principle. See Powers v. Ohio, 499 U.S. 400, 415, 111 S.Ct. 1364, 113 L.Ed.2d 411 (1991) (criminal defendant may raise statutory and constitutional Equal Protection rights of veniremen excluded from jury panel because of race); Craig v. Boren, 429 U.S. 190, 197, 97 S.Ct. 451, 50 L.Ed.2d 397 (1976) (bar owner could raise Equal Protection rights of his patrons in challenge to law allowing 18 year-old girls to buy alcohol, but denying boys the right until age 21); Barrows v. Jackson, 346 U.S. 249, 257, 73 S.Ct. 1031, 97 L.Ed. 1586 (1953) (defendant sued for breach of racially restrictive covenant could raise Equal Protection rights of those excluded as defense to breach). In each case, the Court has permitted someone other than the right-holder to assert the right in court. Thus even in the absence of the right-holder's attempt to assert the right, it acts to constrain others' conduct.
Individuals therefore possess rights, rights with real-life consequences, independent of the ability to seek redress for violation of those rights through affirmative civil process. Indeed, most of the consequences discussed above were well-recognized before, and in some cases long before, the Supreme Court's 1971 decision in Bivens v. Six Unknown Named Agents of Federal Bureau of Narcotics, 403 U.S. 388, 91 S.Ct. 1999, 29 L.Ed.2d 619 (1971), which permitted a direct private right of action against federal government officials to redress violations of a constitutional right. It would be absurd to say that, until Bivens, individuals did not possess with respect to the federal government, or possess in any meaningful sense, the Fourth Amendment right to be free of unreasonable searches and seizures or the Fifth Amendment right against self-incrimination.
Statutes, also, can create or recognize rights without at the same time providing an affirmative judicial remedy for their enforcement. The National Labor Relations Act (“NLRA”), for example, explicitly confers on employees the right to organize and join unions.4 See Lechmere, Inc. v. National Labor Relations Bd., 502 U.S. 527, 532, 112 S.Ct. 841, 117 L.Ed.2d 79 (1992). The NLRA also prohibits employers from interfering with this right. See 29 U.S.C. § 158(a)(1). Yet, employees may not sue to redress violations of the right. Instead, enforcement of the NLRA is within the exclusive jurisdiction of the National Labor Relations Board (“NLRB”). See Communications Workers of Am. v. Beck, 487 U.S. 735, 742, 108 S.Ct. 2641, 101 L.Ed.2d 634 (1988); San Diego Bldg. Trades Council v. Garmon, 359 U.S. 236, 245, 79 S.Ct. 773, 3 L.Ed.2d 775 (1959). Further, the NLRB's General Counsel has discretion whether to prosecute charges of interference with § 7 rights, see 29 U.S.C. § 153(d), and the NLRB can choose not to assert jurisdiction or investigate a particular species of claim, Garmon, 359 U.S. at 245-46, 79 S.Ct. 773.
Even in the absence of an affirmative enforcement action, the rights the NLRA confers on individuals have consequences: they function as limits on state action. In Nash v. Florida Indus. Comm'n, 389 U.S. 235, 239, 88 S.Ct. 362, 19 L.Ed.2d 438 (1967), the Court held that an employee's § 7 rights precluded a state from denying an employee unemployment benefits because she filed an unfair labor practice charge. “The action of Florida here,” the Court wrote, “like the coercive actions which employers and unions are forbidden to engage in, has a direct tendency to frustrate the purpose of Congress to leave people free to make charges of unfair labor practices.” Id. at 239, 88 S.Ct. 362. Thus the petitioner's rights under the NLRA, although not privately enforceable under the statute and not the subject of an NLRB enforcement proceeding in the particular case, nonetheless functioned to pre-empt an adverse state action.
As the NLRA example shows, statutes can create rights in individuals absent any guarantee that the individual can obtain judicial enforcement. In Alden v. Maine, 527 U.S. 706, 119 S.Ct. 2240, 144 L.Ed.2d 636 (1999), similarly, the Court held that, while Congress could through the exercise of its Article I powers create federal rights that were binding on the states (through the Supremacy Clause), it could not also subject the states to private suit in either state or federal court for violations of those rights. Id. at 754, 119 S.Ct. 2240. In other words, Congress could create a right, but could not ensure a forum in which to vindicate that right. Yet, the inability to bring a private suit did not mean that the petitioners in Alden lacked the rights they sought to vindicate: “[t]he constitutional privilege of a State to assert its sovereign immunity in its own courts does not confer upon the State a concomitant right to disregard the Constitution or valid federal law.” Id. at 754-55, 119 S.Ct. 2240. In other words, the state's immunity from suit did not prevent Congress from creating a right, or discharge the state's obligation to respect the right.
The dissent in Alden took issue with this separation of right and private remedy:
[T]here is much irony in the Court's profession that it grounds its opinion on a deeply rooted historical tradition of sovereign immunity, when the Court abandons a principle nearly as inveterate, and much closer to the hearts of the Framers: that where there is a right, there must be a remedy.
Id. at 811, 119 S.Ct. 2240 (Souter, J. dissenting). The dissent further opined that:
[T]oday the Court has no qualms about saying frankly that the federal right to damages afforded by Congress under the FLSA cannot create a concomitant private remedy. The right was made for the benefit of petitioners; they have been hindered by another of that benefit; but despite what has long been understood as the necessary consequence of law, they have no action.
Id. at 812, 119 S.Ct. 2240 (internal quotation marks omitted). In response to this equation of “right” with “private remedy,” the majority noted, first, that the State remains obligated to obey the federal law even absent a mandatory judicial enforcement mechanism; and second, that the state's sovereign immunity would not protect it should the federal government choose to sue it on behalf of the petitioners to enforce the same right petitioners could not themselves enforce in a private action. See id. at 759-60, 119 S.Ct. 2240. That a judicial remedy for violation of the right conferred by the statute was neither available at the behest of the rights-holder nor assured did not, in the Court's view, alter the basic entitlement created by the federal statute.
Any analysis of the reach of § 1983 must therefore begin with, and not lose sight of, the unexceptional proposition that rights are entirely distinct from any private, affirmative, judicial remedy that may exist for violation or deprivation of those rights. The authorities above demonstrate that a person can possess a meaningful right, and that right can have real-life consequences for the conduct of other persons, independent of a concomitant ability to sue for violation of that right.
It is this distinction that gives rise to the two lines of authority critical to this case: the implied right of action cases and the § 1983 cases. The implied right of action cases, from Cort v. Ash, 422 U.S. 66, 95 S.Ct. 2080, 45 L.Ed.2d 26 (1975) through Alexander v. Sandoval, 532 U.S. 275, 121 S.Ct. 1511, 149 L.Ed.2d 517 (2001), all involve attempts to bring a private civil action under a statute that does not explicitly provide a private remedy. See, e.g., Touche Ross & Co. v. Redington, 442 U.S. 560, 576, 99 S.Ct. 2479, 61 L.Ed.2d 82 (1979); Cannon v. University of Chicago, 441 U.S. 677, 717, 99 S.Ct. 1946, 60 L.Ed.2d 560 (1979).
In every implied right of action case, the reviewing court, however obliquely in some cases, has satisfied itself before permitting an action to proceed both that the statute confers a right and that permitting a private remedy to vindicate that right is consistent with Congress's intent in enacting the statute. Cort enunciated four factors to guide courts in this inquiry. The first is “does the statute create a federal right in favor of the plaintiff?” 422 U.S. at 78, 95 S.Ct. 2080. The remaining three factors all concern whether it would be appropriate for a court to permit a private remedy in federal court to vindicate that right. Although the Cort test has been modified, see Touche Ross, 442 U.S. at 575, 99 S.Ct. 2479, the inquiry retains this essential dichotomy between right and remedy. Failure of either part is sufficient to deny a private right of action. See Touche Ross, 442 U.S. at 576, 99 S.Ct. 2479.
Section 1983, in contrast, undisputably does create a right of action. Indeed, that is all it does: It “merely provides a mechanism for enforcing individual rights ‘secured’ elsewhere, i.e., rights independently ‘secured by the Constitution and laws' of the United States.” Gonzaga, 536 U.S. at 285, 122 S.Ct. 2268. Section 1983, therefore, is premised on, and only makes sense in light of, the idea that rights and remedies are distinct. Further, precisely because the only function of § 1983 is to provide a private cause of action for the enforcement of rights as to which Congress has not otherwise prescribed a private remedy, recognizing the availability of a § 1983 cause of action where the statute creating the right does not also create a private right of action is no anomaly.
Accordingly, the first step in a § 1983 action is to identify a federal right. Id. at 283, 122 S.Ct. 2268. Even still, demonstrating the existence of a federal right only creates a presumption of private enforceability. Id. at 284, 122 S.Ct. 2268. That presumption may be rebutted if the defendant can show that Congress explicitly or implicitly foreclosed a private remedy. Id. at 284 n. 4, 122 S.Ct. 2268; Middlesex County Sewerage Auth. v. National Sea Clammers Ass'n, 453 U.S. 1, 20, 101 S.Ct. 2615, 69 L.Ed.2d 435 (1981). For this reason as well, the very structure of a § 1983 case assumes that rights and private rights of action are distinct.
The majority's treatment of Sandoval offers a glimpse of its confusion on this fundamental point. The principal issue before the Court in Sandoval was whether it is permissible to imply a private right of action under a regulation when the statute the regulation implements does not provide for private redress. See id. at 291, 121 S.Ct. 1511. The Court declined to determine whether the regulation in question was independently rights-creating or not (or could be), and instead focused its analysis on the second half of the implied right of action inquiry: whether Congress specifically intended private remedies to enforce the rights created by regulation. Id.
Ultimately, the Court concluded that Congress did not so intend and thus that the agency regulations themselves could not support an implied right of action. Id. at 293, 121 S.Ct. 1511. The Court held that Congress had expressed no intent when it enacted the statute that the right at issue be enforceable by private actions, and thus the statute did not confer on the agency the authority to permit private enforcement of its regulations. See id. at 288-89, 121 S.Ct. 1511. It is in this context that the Court wrote that:
Language in a regulation may invoke a private right of action that Congress through statutory text created, but it may not create a right that Congress has not․ [I]t is most certainly incorrect to say that language in a regulation can conjure up a private cause of action that has not been authorized by Congress. Agencies may play the sorcerer's apprentice but not the sorcerer himself.
Id. at 291, 121 S.Ct. 1511.
Critically, the Court's conclusion was not based, as the majority believes, on a theory of rights creation. Instead, it rested on a separation of powers principle, namely, the Court's view of the proper function of the federal courts vis-a-vis Congress. “Raising up causes of action where a statute has not created them may be a proper function for common-law courts, but not for federal tribunals.” Id. at 287, 121 S.Ct. 1511 (quoting Lampf, Pleva, Lipkind, Prupis & Petigrow v. Gilbertson, 501 U.S. 350, 365, 111 S.Ct. 2773, 115 L.Ed.2d 321 (1991)). That federal tribunals, unlike common law courts, may ordinarily not create remedial schemes in turn reflects “a concern, grounded in separation of powers, that Congress rather than the courts controls the availability of remedies for violations of statutes.” Wilder v. Virginia Hospital Ass'n, 496 U.S. 498, 508 n. 9, 110 S.Ct. 2510, 110 L.Ed.2d 455 (1990). In other words, only Congress may provide access to the federal courts, and thus only Congressional intent is relevant in determining whether to imply a right of action.
The special separation of powers concerns underlying Sandoval do not apply in a § 1983 case. Congress explicitly granted access to the federal courts when it enacted § 1983. See Wilder, 496 U.S. at 508 n. 9, 110 S.Ct. 2510 (“Because § 1983 provides an alternative source of express congressional authorization of private suits, these separation-of-powers concerns are not present in a § 1983 case.”). So, when a cause of action that otherwise meets the requirements of § 1983 is at issue, our task is simply to determine whether the regulation in question is the type of legal prescript that Congress meant to be enforceable under § 1983. Our decision on this question does not implicate the separation of powers concerns that drove Sandoval. To the contrary, to fail to apply § 1983 to a set of legal rules that fits within its compass would flaunt the intent of the Congress that enacted § 1983.
The majority thus misinterprets Gonzaga as holding that “individual rights enforceable through § 1983 are similar to implied rights of action․ Since only Congress can create implied rights of action ․ only Congress can create rights enforceable through § 1983.” Ante at 938-939. As I have demonstrated, the concept of individual rights does differ fundamentally from the concept of a private cause of action. To maintain otherwise is simply a play on the word “right,” nothing more.5
Gonzaga in no way supposes otherwise. Instead, that case held only that the initial inquiry in § 1983 cases and in implied right of action cases is the same inquiry: whether the law at issue creates an individual right. Gonzaga did not merge the two lines of cases in their entirety, the majority's repeated suggestions to the contrary notwithstanding. Nor did Gonzaga “suggest[ ] that only Congress can create rights enforceable through § 1983.” Ante at 939. To the extent that Gonzaga stressed Congressional intent, see 536 U.S. at 283-84, 122 S.Ct. 2268, that was because the plaintiff there had brought suit under § 1983 to enforce a right he believed was secured directly by a statute. Gonzaga neither raised nor discussed the question at the heart of this case-whether a particular type of law can create a right. The answer to this question, as the next section demonstrates, involves quite a different set of considerations.
Aside from the supposed identity of rights and private causes of action, the majority rests its contrary categorical assertion that “agency regulations cannot independently create rights,” ante at 939, on its view of the role of administrative agencies. That view flies in the face of seventy years of administrative law jurisprudence. Applying contemporary administrative law principles rather than antiquated ones, I can see no reason why valid agency regulations cannot create individual rights and do so independently of specific Congressional intent regarding the rights created.
Agency rules, for present purposes, come in two stripes: interpretive rules and legislative regulations. See Richard J. Pierce, Jr., Administrative Law Treatise, § 6.4 at 325 (4th ed.2002). The principal differences between the two rest on both Congress's intent when it enacts a statute and the agency's intent when it promulgates a regulation to implement that statute.
An agency has the inherent power to promulgate interpretive rules, and thus may do so independently of any express grant of power from Congress and without following special procedures. See id. Legislative regulations, on the other hand, require an express delegation of rule-making authority from Congress and must be promulgated according to specific procedures. See id.; Batterton v. Francis, 432 U.S. 416, 425-25, 97 S.Ct. 2399, 53 L.Ed.2d 448 (1977); 5 U.S.C. § 553. Legislative regulations can “impose distinct obligations on members of the public in addition to those imposed by statute.” Pierce, supra, at 325 (emphasis added).
In practical terms, legislative regulations have all the relevant properties of statutes. Like statutes, agency regulations are prescriptive, forward-looking, and of general applicability. See 5 U.S.C. § 551(4) (defining a “rule” as “an agency statement of general or particular applicability and future effect designed to implement, interpret, or prescribe law or policy ․”); Prentis v. Atlantic Coast Line Co., 211 U.S. 210, 226, 29 S.Ct. 67, 53 L.Ed. 150 (1908) (“Legislation ․ looks to the future and changes existing conditions by making a new rule to be applied thereafter to all or some part of those subject to its power”) (Holmes, J.). Agency regulations, like statutes, often reflect a careful balance between competing interests and policy considerations. See Chevron U.S.A., Inc. v. Natural Res. Def. Council, 467 U.S. 837, 865, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984). And, like statutes, agency rules “affect[ ] individual rights and obligations”: they are binding on the individuals to whom they apply in the same way statutes are. See Chrysler Corp. v. Brown, 441 U.S. 281, 302, 99 S.Ct. 1705, 60 L.Ed.2d 208 (1979). They are also binding, within limits, on the judiciary as well as the executive. See United States v. Mead Corp., 533 U.S. 218, 227, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001); Chevron, 467 U.S. at 844, 104 S.Ct. 2778. Regulations thus have the same form and the same effect, and are based on the same types of considerations, as statutes. Like statutes, they can order the relationship between one individual and another, and they are backed by the coercive power of the government.
Indeed, to the extent that regulations are inferior to statutes, they are inferior in ways external to the relation-ordering function they perform. We accord regulations less deference than we do statutes. While we may declare a statute unconstitutional only if we find that Congress has exceeded its constitutional authority in enacting it, see United States v. Morrison, 529 U.S. 598, 607, 120 S.Ct. 1740, 146 L.Ed.2d 658 (2000), or violated the requirements of bicameralism and presentment, see I.N.S. v. Chadha, 462 U.S. 919, 959, 103 S.Ct. 2764, 77 L.Ed.2d 317 (1983), we may invalidate a regulation if it conflicts with the Constitution, see 5 U.S.C. § 706(2)(B), if it exceeds the statutory authority under which it was promulgated, see 5 U.S.C. § 706(2)(C), or if it represents an arbitrary or capricious exercise of that authority, see 5 U.S.C. § 706(2)(A). Yet, the deference we accord a regulation in a challenge to its validity says nothing about the regulation's force if valid.
Regulations are more numerous and specific than statutes. Numerosity, however, does not impact function, and specificity is an essential requirement of judicial enforcement of rights. See Blessing v. Freestone, 520 U.S. 329, 340-41, 117 S.Ct. 1353, 137 L.Ed.2d 569 (1997). If anything, the abundance and specificity of regulations would suggest that they could create many rights enforceable under § 1983, not render them impotent to do so.
Regulations may be more transient than statutes. Thus, allowing that regulations may create rights could mean that some rights are relatively short-lived. See Wright v. City of Roanoke Redevelopment & Housing Auth., 479 U.S. 418, 438, 107 S.Ct. 766, 93 L.Ed.2d 781 (1987) (O'Connor, J., dissenting). Accepting that this is true, it is not clear why a short-lived right is any less a right while in existence. More importantly, however, transience is a matter of degree. Statutes, which the majority concludes can create rights, can be repealed-and the concomitant rights they create extinguished-by a simple majority in both Houses of Congress and the assent of the President. Though there may be a greater institutional tendency to preserve the status quo in Congress than in an agency,6 the rights protected by statute are in any case more transient than those protected by the Constitution, for it is easier to repeal a law than it is to amend the Constitution. Thus, any rights created by statute are transient relative to rights protected by the Constitution, just as rights created by regulation are transient relative to rights protected by statutes.
The majority has provided no explanation rooted in any pertinent functional differences concerning why it chooses to draw a line between statutes and regulations. As the foregoing paragraphs demonstrate there is no functional difference between statutes and regulations that would justify the majority's holding that only the former may create rights.
It is not surprising, therefore, that we have in the past treated regulations as if they create rights. To take a fairly random example, in Long v. Coast Resorts, Inc., 267 F.3d 918, 923 (9th Cir.2001), we permitted a disabled plaintiff to sue to enforce regulations implementing the Americans with Disabilities Act (“ADA”). The statutory provision in question defined discrimination only as a failure to “design and construct facilities ․ that are readily accessible to and usable by individuals with disabilities․” 42 U.S.C. § 12183(a)(1). One of the many detailed regulations implementing that provision required that doorways in facilities covered by the ADA be at least thirty-two inches wide. Id. at 922; see 28 C.F.R. Appendix A to Part 36, Guideline 4.13.5. The plaintiff's suit alleged that the door widths in 819 of 839 rooms at a Las Vegas hotel were narrower than the regulation required. See id. at 921. We remanded the case with instructions that the lower court issue an injunction ordering the hotel to bring its doorways into compliance with the regulations. Id. at 923.
The ADA does not contain any door-width requirements. See 42 U.S.C. §§ 12181-12188. To the extent that the plaintiff had any “right” to have thirty-two inch doorways, it was a right he derived from the implementing regulation. Yet, in the eyes of this court, it was no less a right because it derived from a regulation rather than a statute. To the contrary, the right functioned exactly like a right created by statute: It entitled the plaintiff to demand certain conduct of another, the hotel owner, and to seek the coercive power of the government (through an injunction) to back up his demand.7
By ignoring regulation-enforcing cases such as Long, although they are legion, the majority avoids any discussion of the differences between regulations and statutes that might support its conclusion that only the latter can create rights. Its decision seems rather to be based on formalism. The majority writes that “it is an elementary principle of constitutional law that lawmaking is the province of Congress,” and again that “Congress, rather than the executive, is the lawmaker in our democracy.” Ante at 939. According to the majority, in other words, a law's capacity to create rights is due not to any properties of the law itself but instead to the law's source; only if a law comes from Congress can it create a right. The majority envisions a neat division between legislative and executive power and assumes that Congress may not delegate any of its power to an executive agency.
In support of its assertion, the majority quotes A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495, 529, 55 S.Ct. 837, 79 L.Ed. 1570 (1935): “The Congress is not permitted to abdicate or transfer to others the essential legislative functions with which it is thus vested.” It is telling that the majority chose this case and that quote to support its assertion. Schechter Poultry remains an important case in agency law for almost the opposite proposition: It is one of only two cases in history in which the Court has invalidated an Act of Congress as an overbroad delegation of legislative authority to an administrative agency. See Mistretta v. United States, 488 U.S. 361, 373, 109 S.Ct. 647, 102 L.Ed.2d 714 (1989). The other was Panama Refining Co. v. Amazon Petroleum Corp., 293 U.S. 388, 55 S.Ct. 241, 79 L.Ed. 446 (1935), decided the same year as Schechter Poultry. Far from suggesting that Congress may not delegate “legislative functions” to other bodies, Schechter Poultry now demarcates the boundaries of acceptable delegation. See Mistretta, 488 U.S. at 373, 109 S.Ct. 647.
Indeed, one would have thought Schechter Poultry 's dubious viability on this point obvious. Whitman v. American Trucking Ass'n, Inc., 531 U.S. 457, 121 S.Ct. 903, 149 L.Ed.2d 1 (2001), which the majority also cites, makes this plain. There, the Court reversed a lower court decision that had invalidated a federal statute as an overbroad delegation of legislative power to the Environmental Protection Agency (“EPA”). Id. at 475-76, 121 S.Ct. 903. The lower court found that the statute did not provide the agency an “intelligible principle” to guide its exercise of discretion. Id. at 463, 121 S.Ct. 903. The Supreme Court disagreed, noting that “[i]n the history of the Court, we have found the requisite ‘intelligible principle’ lacking in only two statutes․” Id. at 474, 121 S.Ct. 903 (citing Schechter Poultry and Panama Refining ). The reason for this reluctance, the Court explained, is that “we have ‘almost never felt qualified to second-guess Congress regarding the permissible degree of policy judgment that can be left to those executing or applying the law.’ ” Id. at 474-75, 121 S.Ct. 903 (quoting Mistretta, 488 U.S. at 416, 109 S.Ct. 647 (Scalia, J., dissenting)). Thus, while it is indisputably true that Congress may not delegate its legislative power to administrative agencies, in practice this limitation operates to prohibit only the broadest of delegations. It does not, as the majority believes, relegate agencies to the sole task of defining statutory provisions.
As Whitman indicates, the Supreme Court long ago rejected the type of formalist analysis of the relationship between Congress and administrative agencies upon which the majority relies. Compare Field v. Clark, 143 U.S. 649, 693-94, 12 S.Ct. 495, 36 L.Ed. 294 (1892) (“The true distinction ․ is between the delegation of power to make the law, which necessarily involves a discretion as to what it shall be, and conferring authority or discretion as to its execution, to be exercised under and in pursuance of the law. The first cannot be done; to the latter no valid objection can be made.”) (internal quotation marks omitted), and Schechter Poultry, 295 U.S. at 529-530, 55 S.Ct. 837; with Mistretta v. United States, 488 U.S. 361, 372, 109 S.Ct. 647, 102 L.Ed.2d 714 (1989) (“our jurisprudence has been driven by a practical understanding that in our increasingly complex society, replete with ever changing and more technical problems, Congress simply cannot do its job absent an ability to delegate power under broad general directives”), and Loving v. United States, 517 U.S. 748, 758, 116 S.Ct. 1737, 135 L.Ed.2d 36 (1996) (“This Court established long ago that Congress must be permitted to delegate to others at least some authority that it could exercise itself”).
Today's administrative law jurisprudence is therefore driven by a pragmatic view of the roles of Congress and the administrative agencies. That jurisprudence does not inquire whether Congress has delegated legislative power at all, but only whether Congress has placed appropriate limits on the agency's exercise of legislative authority. The majority's dogmatic assertion that “Congress, rather than the executive, is the lawmaker in our democracy,” is a truism, but it does not capture the nuances of our contemporary understanding of the relationship between Congress and the administrative agencies.
Nor must Congress intend-in whatever sense a collective body intends anything-each and every regulation an agency promulgates to implement a statute. See Mead, 533 U.S. at 229, 121 S.Ct. 2164; Chevron, 467 U.S. at 845, 104 S.Ct. 2778; Chrysler, 441 U.S. at 308, 99 S.Ct. 1705 (“This is not to say that any grant of legislative authority to a federal agency by Congress must be specific before regulations promulgated pursuant to it can be binding on courts in a manner akin to statutes”). To the contrary, Congress may choose not to legislate specifically in a particular area but instead leave it to the agency to fill out the area with regulations. See Chevron, 467 U.S. at 843-44, 104 S.Ct. 2778. In such instances, the agency performs much like a legislature, albeit only as to matters pre-designated by Congress. See Chrysler, 441 U.S. at 304-05, 99 S.Ct. 1705 (for regulations to be valid, there must be “a nexus between the regulations and some delegation of the requisite legislative authority by Congress”).
Chevron succinctly illustrates these ideas. The Clean Air Act (“Act”) required certain states to establish permit programs regulating new or modified “stationary sources” of air pollution. 467 U.S. at 839-40, 104 S.Ct. 2778. The Environmental Protection Agency (“EPA”) promulgated regulations to implement the Act, including one regulation (the “bubble rule”) that permitted states to adopt a “plant-wide” definition of “stationary source.” Id. at 840, 104 S.Ct. 2778. Thus, under the regulation, all pollution emitting devices within a single plant could be treated as part of the same “stationary source,” as opposed to treating each one (say, each smoke stack) as an individual “stationary source.” Id. The Court rejected a challenge to the regulation, holding that Congress did not have a specific intention regarding the “bubble rule” and the rule represented a reasonable policy choice that Congress had left the agency to make.8 Id. at 845, 104 S.Ct. 2778. In its conclusion, the Court noted:
Congress intended to accommodate both interests, but did not do so itself on the level of specificity presented by these cases. Perhaps that body consciously desired the Administrator to strike the balance at this level, thinking that those with great expertise and charged with responsibility for administering the provision would be in a better position to do so; perhaps it simply did not consider the question at this level; and perhaps Congress was unable to forge a coalition on either side of the question, and those on each side decided to take their chances with the scheme devised by the agency. For judicial purposes, it matters not which of these things occurred.
Id. 865, 104 S.Ct. 2778. In other words, Congress may empower an agency to perform many of the same functions that Congress itself performs. So long as Congress circumscribes the area in which the agency may exercise this authority, its delegation will be upheld and the regulations the agency promulgates will have the force and effect of law. See Mistretta, 488 U.S. at 372-73, 109 S.Ct. 647 (“this Court has deemed it constitutionally sufficient if Congress clearly delineates the general policy, the public agency which is to apply it, and the boundaries of this delegated authority”) (internal quotation marks omitted).
The majority's notion that regulations are valid only if they flesh out a specific statutory provision is therefore wrong. And, just as regulations may create new obligations, not specifically intended by Congress, within the sphere properly delegated to the promulgating agency, I can see nothing in the administrative law principles governing legislative regulations that precludes promulgation of the particular form of rules that we describe as creating “rights.” Such rights-creating regulations, like other valid legislative regulations, have the force and effect of laws and are binding on individuals regulated, the courts and the agency itself.
This line of reasoning explains our decision in Buckley v. City of Redding, 66 F.3d 188 (9th Cir.1995), and explains as well why the majority departs from stare decisis principles in refusing to follow Buckley. In Buckley we held that the Federal Aid in Sport Fish Restoration Act, codified at 16 U.S.C. § 777 et seq., created a right enforceable under § 1983 to equal river access for boats of “common horsepower ratings.” 66 F.3d at 192. In reaching this conclusion, we looked to the regulations promulgated to implement the Act. Id. Buckley thus necessarily stands for the proposition that legislative regulations, by themselves, may create rights enforceable under § 1983.
The majority distinguishes Buckley by saying that “[a]lthough we referred to the regulations, we were careful to emphasize that the regulations were interpretive.” (emphasis in original). The majority stresses the appellation “interpretive,” apparently, to suggest that in fact the regulations in Buckley were only “fleshing out” a right that the statute itself created.
The regulations in Buckley, however, were promulgated under an express and broad grant of authority from Congress, see 16 U.S.C. § 777i, and created a right found nowhere in the statute. They were, in other words, legislative regulations which validly created a binding obligation in addition to those enumerated in the statute itself. See Kissimmee River Valley Sportsman Ass'n v. City of Lakeland, 250 F.3d 1324, 1326-27 (11th Cir.2001) (describing the same regulation at issue in Buckley as one which, assuming its validity, “goes beyond explicating the specific content of the statutory provision and imposes distinct obligations in order to further the broad objectives underlying the statutory provision”). The majority's decision that such regulations cannot create rights is therefore in direct conflict with Buckley.9
The majority has assumed-as it must, in light of Guardians Ass'n v. Civil Serv. Comm'n of New York, 463 U.S. 582, 103 S.Ct. 3221, 77 L.Ed.2d 866 (1983)-that the DOT regulation at issue here is valid.10 The necessary corollary of this assumption is that the DOT was acting within its statutory grant of authority when it promulgated the disparate impact regulation. Accordingly, the regulation should have the same force and effect as a statute, and may (although I ultimately conclude that it does not) create a right to be free of disparate impact discrimination.
If agency regulations can create rights, then individuals should be able to vindicate those rights through private actions brought under § 1983. Section 1983 provides:
Every person who, under color of any statute, ordinance, regulation, custom or usage, of any State or Territory or the District of Columbia, subjects, or cause to be subjected, any citizen of the United States or other person within the jurisdiction thereof to the deprivation of any rights, privileges, or immunities secured by the Constitution and laws, shall be liable to the party injured in an action at law, suit in equity, or other proper proceeding for redress․
42 U.S.C. § 1983.
By its own terms, § 1983 permits enforcement of rights secured by the “Constitution and laws.” As I have demonstrated, binding regulations are in form and function equivalent to laws and are commonly said to have the force of law. “This doctrine is so well established that agency regulations implementing federal statutes have been held to pre-empt state law under the Supremacy Clause.” Chrysler, 441 U.S. at 295-96, 99 S.Ct. 1705. Chrysler, indeed, requires a “clear showing of contrary legislative intent” to rebut the presumption that regulations implementing one statute are “law” for purposes of interpreting another statute using the term “law.” 11 Id. at 296, 99 S.Ct. 1705. So, absent some reason for concluding otherwise, “laws” in § 1983 includes rights secured by regulations.
Far from indicating otherwise, the language and structure of § 1983 as a whole confirms that “laws” includes regulations. We generally assume that when Congress uses different words in a statute, it intends them to have different meanings. See S.E.C. v. McCarthy, 322 F.3d 650, 656 (9th Cir.2003). Here, Congress used the phrase “Constitution and laws” rather than “Constitution and statutes,” yet referred elsewhere in the same sentence to “any statute, ordinance, regulation, custom or usage․” See 42 U.S.C. § 1983. While today's large federal bureaucracy did not exist when § 1983 was enacted in 1874, the 1874 Congress was quite aware, as § 1983 itself indicates, that there are different sources of law, including regulations. In this context, the terms “laws” and “statutes” must have different meanings. Further, the term “laws” necessarily has a broader meaning than “statutes,” not an equivalent or narrower meaning.
Indeed, the Supreme Court has rejected narrow interpretations of the phrase “and laws” in the past: In Maine v. Thiboutot, 448 U.S. 1, 7, 100 S.Ct. 2502, 65 L.Ed.2d 555 (1980), the Court made clear that the phrase does not encompass only civil rights laws but includes rights secured by other federal laws as well.12 Id. at 10, 100 S.Ct. 2502. Applying the Chrysler presumption, “laws” in § 1983 includes regulations as well.
The final question is whether the regulation at issue here creates a federal right enforceable under § 1983. I conclude-with some reluctance-that it does not, but not for the reasons the majority relies upon.
Gonzaga now guides the inquiry whether a particular provision creates a right. See 536 U.S. at 283-84, 122 S.Ct. 2268. Although the Court had a statute before it in Gonzaga, the inquiry the Court there enunciated is no less appropriate for agency regulations. The only difference is that, in the context of agency regulations, there is a prior question of the particular regulation's validity.
Only those regulations that unambiguously create individual rights will support a § 1983 action in the context of a Spending Clause statute.13 Cf. Gonzaga, 536 U.S. at 283, 122 S.Ct. 2268. We look for “individually-focused,” “rights-creating language” in the regulation. Cf. id. at 287, 122 S.Ct. 2268. We also ask whether the regulation is aimed at individual instances of conduct or instead functions at the level of institutional policy and practice. Cf. id. at 288, 122 S.Ct. 2268. Contrary to SOV's assertion, there is simply no fair reading of the disparate impact regulation at issue here that meets these newly-announced criteria.
Section 21.5(a) of the DOT regulations provides that “no person in the United States shall, on the grounds of race, color or national origin, be excluded from participation in, be denied the benefits of, or be otherwise subjected to discrimination under any program” which receives federal funds. 49 C.F.R. § 21.5(a). Section 21.5(b)(2) states, in turn, that “[a] recipient, in determining the types of services, financial aid, or other benefits or facilities which will be provided under any such program ․ may not utilize criteria or methods of administration which have the effect of subjecting persons to discrimination” on account of race, color or national origin.14 SOV maintains that § 21.5(b)(2) defines what actions constitute discrimination within the meaning of the § 21.5(a) rights. And taken together, SOV argues, these two regulations create a right to be free of federally funded transportation programs that disparately impact members of a group based on their race. The problem, however, is that after Sandoval, the two regulations cannot be taken together, and taken alone, § 21.5(b)(2) does not meet the Gonzaga criteria.
Section 21.5(a), it is true, unambiguously prohibits “discrimination” and uses rights-creating language to do so. The statement that “[n]o person in the United States shall ․” suffer discrimination is identical to language previously held to create individual rights. See Cannon v. University of Chicago, 441 U.S. 677, 696, 99 S.Ct. 1946, 60 L.Ed.2d 560 (1979).
Sandoval, however, states in no uncertain terms that this very language, as used in § 601 of Title VI, “prohibits only intentional discrimination,” 532 U.S. at 280, 121 S.Ct. 1511, and that “the disparate impact regulations [therefore] do not simply apply § 601,” id. at 285, 121 S.Ct. 1511. So we are precluded from reading the disparate impact regulation, as plaintiffs would have us do, as “spelling out the meaning of the ‘general language’ ” of § 601, as repeated in the regulation. To do so would run head-long into Sandoval.15
The question, then, is whether the disparate impact regulation, § 21.5(b)(2), can instead be understood, standing on its own, as a valid, rights-creating legislative regulation, consistent with the authorizing statute and promulgated within the scope of a Congressional delegation to an agency to enunciate policy as to particular matters. My answer is that § 21.5(b)(2)-the majority's skepticism on this point notwithstanding-is a valid legislative regulation promulgated under § 602, but under Gonzaga, it does not create a separate right to be free of actions with a disparate impact.
The disparate impact regulation effectuates the prohibition on intentional discrimination by recipients of federal funds for DOT projects by requiring recipients to provide assurance that their programs will not disparately impact protected classes of people. To be sure, the regulation will result in denying funds to some projects that do not violate § 601 of the statute, because those projects were not adopted or administered with discriminatory intent. Discovering discriminatory intent, however, is a fact-intensive process, as the Supreme Court has recognized. See Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 266, 97 S.Ct. 555, 50 L.Ed.2d 450 (1977) (“Determining whether invidious discriminatory purpose was a motivating factor demands a sensitive inquiry into such circumstantial and direct evidence of intent as may be available”). A prohibition on practices with a discriminatory impact but no adequate justification can serve as an effective way of ferreting out actions undertaken with a discriminatory intent. See Washington v. Davis, 426 U.S. 229, 242, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976) (“It is also not infrequently true that the discriminatory impact ․ may for all practical purposes demonstrate unconstitutionality because in various circumstances the discrimination is very difficult to explain on nonracial grounds”). Thus, just as Congress may, in the exercise of its authority under Section 5 of the Fourteenth Amendment, prohibit “a somewhat broader swath of conduct” than the Fourteenth Amendment itself proscribes, see Nevada Dept. of Human Res. v. Hibbs, 538U.S. 721, ----, 123 S.Ct. 1972, 1977, 155 L.Ed.2d 953 (2003), so too may the DOT in its efforts to enforce Title VI's prohibition on intentional discrimination by recipients of federal funds proscribe methods of administration likely to cloak actions taken in violation of the intentional discrimination prohibition.16
That the disparate impact regulation is a valid means of implementing the protected right to be free of intentional discrimination is not, however, an adequate basis for concluding that it creates an enforceable right under the regulation. Rather, Gonzaga stresses that “[nothing] short of an unambiguously conferred right [will] support a cause of action brought under § 1983” in Spending Clause cases, 536 U.S. at 283, 122 S.Ct. 2268, and requires that we consider in detail the pertinent provision to see whether it meets this standard. Applying Gonzaga to § 21.5(b)(2), I conclude that the regulation does not create a separate right in the affected group of people.
For one thing, the language of § 21.5(b)(2), standing alone, does not meet the primary criterion established for the first time by Gonzaga, namely, the requirement that “for a [regulation] to create such private rights, its text must be phrased in terms of the persons benefitted.” 536 U.S. at 284, 122 S.Ct. 2268 (quoting Cannon, 441 U.S. at 692, n. 13, 99 S.Ct. 1946). Instead, § 21.5(b)(2) begins with a directive to the “recipient, in determining the types of services, financial aid or other benefits or facilities which will be provided under any such program” and goes on to prescribe the “criteria or methods of administration” that such recipient may use-namely those which do not “have the effect of subjecting persons to discrimination because of their race, color or natural origin or the effect of defeating or substantially impairing accomplishment of the objects of the program with respect to individuals of a particular race, color or natural origin.” Id. So the focus of the regulation is on the fund recipient and its methods of operating the funded program, not any individual affected thereby. As such, “the provision[ ] entirely lack[s] the sort of ‘rights-creating’ language,” Gonzaga, 536 U.S. at 287, 122 S.Ct. 2268, critical under Gonzaga. Instead, the disparate impact regulation is “ ‘written simply as a ban on discriminatory conduct by recipients of federal funds.’ ” Id. at 287, 122 S.Ct. 2268 (quoting Cannon, 441 U.S. at 690-93, 99 S.Ct. 1946).
Further, the disparate impact regulation also shares a second characteristic flagged in Gonzaga as inconsistent with an intention to create individual rights: It has an “ ‘aggregate focus,’ ” id. at 288, 122 S.Ct. 2268 (quoting Blessing, 520 U.S. at 343, 117 S.Ct. 1353), as it speaks “in terms of institutional policy and practice,” id. Section 21.5(b)(2) refers to the utilization of “criteria or methods of administration” on “persons” and “individuals,” not of the right of any single person to be free of actions that have a disparate impact. Compare § 21.5(a)(i)-(v) (proscribing various discriminatory actions, each as they affect “a person”).17
Neither of these two shortcomings would have disqualified § 21.5(b)(2) as a “rights-creating” under the Court's pre-Gonzaga standard. That standard, described in Blessing, 520 U.S. at 340-41, 117 S.Ct. 1353, required only that a provision: (1) evince an intent to benefit the plaintiff; (2) not be so vague and amorphous as to preclude judicial enforcement; and (3) be couched in mandatory, not precatory terms. Section 21.5(b)(2) satisfies these requirements: It benefits certain “persons” and “individuals,” imposes a standard long-enforced in federal court, see 42 U.S.C. § 2000e-2, and is stated in mandatory, not precatory terms. So, but for Gonzaga 's more rigid test, I would have only one hesitation to finding that § 21.5(b)(2) creates an individually enforceable right.
One sentence in Sandoval can be read as indicating that any such regulation would be beyond the authority Congress delegated to the DOT under § 602 of Title VI. As the majority notes, Sandoval states that “ § 602 limits agencies to ‘effectuating’ rights already created by § 601.” 532 U.S. at 289, 121 S.Ct. 1511. Although that statutory limitation is not apparent to me from the language of § 602, the quoted language from Sandoval may be delineating the reach of Congress's authorization to promulgate legislative regulations under § 602.18
For the reasons already indicated, any such limitation on agencies' authority to promulgate regulations in no way implicates the overall validity of the disparate impact regulation: One can effectuate rights by prescribing rules designed to assure respect for the rights that are protected, without at the same time creating new rights. Yet, I would be reluctant to conclude that in a single sentence, in the midst of a discussion on a different matter, Sandoval purported to describe the limits of the agencies' authority to promulgate regulations under § 602. I therefore do not rely on this reading of Sandoval, but only on Gonzaga 's instructions regarding the description of private rights in Spending Clause cases.
For this reason alone, I concur in the majority's result.
1. SOV also claimed that Sound Transit's plan would violate the National Environmental Policy Act (NEPA), 42 U.S.C. § 4321 et seq., the Fair Housing Act, 42 U.S.C. §§ 3601-3631, and Title VI of the Civil Rights Act, which prohibits recipients of federal funds from administering programs in an intentionally discriminatory manner. The district court dismissed SOV's NEPA claim, holding that Sound Transit had studied the feasibility of the Rainier Valley Tunnel alternative (at the request of SOV members) and reasonably concluded that alternative was not feasible. It granted summary judgment to Sound Transit on the Fair Housing Act claim on the ground that Sound Transit was not engaged in housing-related activities. The court held that SOV's claims based on § 601 of Title VI could proceed to the extent that SOV can prove that any discrimination was intentional. None of these decisions has been appealed, so we limit our review to whether SOV can sue under § 1983 to enforce the disparate-impact regulations promulgated pursuant to § 602 of Title VI.
2. The disparate-impact regulation at issue was promulgated by the Department of Transportation based on authority granted it by Congress in Title VI of the Civil Rights Act of 1964. In § 601 of that Title, Congress created a right to be free from intentional discrimination based on race. In § 602, Congress authorized federal agencies to “effectuate the provisions of [§ 601] ․ by issuing rules, regulations, or orders of general applicability.” 42 U.S.C. § 2000d-1. Pursuant to § 602, the Department of Transportation promulgated the disparate-impact regulation, which prohibits funding recipients from undertaking activities that have racially discriminatory effects. The regulation forbids recipients of federal funds to “utilize criteria or methods of administration which have the effect of subjecting persons to discrimination because of their race, color, or national origin, or have the effect of defeating or substantially impairing accomplishment of the objectives of the program with respect to individuals of a particular race, color, or national origin.” 49 C.F.R. § 21.5(b)(2).The regulation goes further than the statute it implements, proscribing activities that have disparate effects on racial groups, even though such activities are permissible under § 601. The Supreme Court has recognized that there is “considerable tension” between § 601 and the disparate-impact regulation. Alexander v. Sandoval, 532 U.S. 275, 282, 121 S.Ct. 1511, 149 L.Ed.2d 517 (2001). For purposes of deciding this case, as the Supreme Court did in Sandoval, id., that the regulation may validly proscribe activities that have a disparate impact on racial groups. No party has challenged the validity of the regulation, and we need not decide whether the regulation is invalid.
3. Compared with the Third and Eleventh Circuits' analysis, the Fourth Circuit's was superficial. The Fourth Circuit simply noted the absence of Supreme Court precedent and that Supreme Court justices had expressed doubt (in dissent) that “administrative regulations alone could create such a right.” Smith, 821 F.2d at 984. It therefore held that “[a]n administrative regulation ․ cannot create an enforceable § 1983 interest not already implicit in the enforcing statute.” Id.
4. Building on an interesting survey of sources from Immanuel Kant to Roscoe Pound, the partial dissent argues at length that rights enforceable through § 1983 are different from implied rights of action. We do not disagree with this proposition. As the partial dissent correctly explains, rights enforceable through § 1983 are in the nature of a substantive entitlement; implied rights of action are in the nature of a remedy.The partial dissent does not argue, and could not argue persuasively, that rights enforceable through § 1983 differ from implied rights of action in a manner relevant to this appeal. Specifically, the partial dissent does not explain why executive agencies cannot create implied rights of action (the Supreme Court's Sandoval holding) but can in its view create rights enforceable through § 1983 (the partial dissent's approach). The reason for the partial dissent's omission is clear: the partial dissent's approach cannot be reconciled with Sandoval. Rights enforceable through § 1983, no less than implied rights of action, are creatures of substantive federal law. As such, they cannot be created by executive agencies. See Sandoval, 532 U.S. at 286, 121 S.Ct. 1511 (“Like substantive federal law itself, private rights of action to enforce federal law must be created by Congress.”).
5. SOV and the amici curiae rely on our 1984 opinion in Keaukaha-Panaewa Comm. Ass'n. v. Hawaiian Homes Comm'n., 739 F.2d 1467. In that decision we stated:The lack of an implied private right of action under a federal act ․ does not by itself dispose of the issue of Congressional intent to foreclose private actions under section 1983 ․ As we have stated expressly, “there could well be federal rights enforceable under section 1983 which are not enforceable by means of a private right of action under the statute creating them.”Id. at 1470 (quoting Boatowners and Tenants Ass'n, Inc. v. Port of Seattle, 716 F.2d 669, 674 (9th Cir.1983)). Even if the above-quoted language remains an accurate statement of our law, the Supreme Court in Gonzaga added another consideration: “[I]mplied right of action cases should guide the determination of whether a statute confers rights enforceable under § 1983.” 536 U.S. at 283, 122 S.Ct. 2268.
6. For other cases in the courts of appeals dealing with causes of action relying at least in part on a regulation, see Farley v. Philadelphia Hous. Auth., 102 F.3d 697 (3d Cir.1996); Buckley v. City of Redding, Cal., 66 F.3d 188 (9th Cir.1995); Albiston v. Maine Comm'r of Human Services, 7 F.3d 258 (1st Cir.1993); Howe v. Ellenbecker, 8 F.3d 1258 (8th Cir.1993), cert. denied, 511 U.S. 1005, 114 S.Ct. 1373, 128 L.Ed.2d 49 (1994); and Samuels v. District of Columbia, 770 F.2d 184 (D.C.Cir.1985). The Eleventh Circuit observed that underlying at least some of these cases is the understanding that only statutes can create rights and that regulations merely define the contours of rights created by Congress. See Harris, 127 F.3d at 1007 n. 18 (citing Farley, 102 F.3d at 699 (“[The] cause of action arises strictly under [the statute]. Regulation § 966.57(b) merely interprets that section.”)).
7. SOV also refers to what it calls “the Supreme Court's established practice of adjudicating section 1983 cases to enforce regulations under the Social Security Act, which itself contains no private cause of action.” Even if it is true that the Supreme Court has permitted litigants to enforce Social Security Act regulations, it does not necessarily follow that the regulations-rather than the Act-created the individual right being enforced.
8. “[W]e must first determine whether the statute and its interpretive regulations create an enforceable federal statutory right.” Buckley, 66 F.3d at 190.
9. For this reason, SOV's lengthy argument that regulations are “laws” is not helpful.
10. This normally would involve an application of the Supreme Court's three-prong Blessing test:First, Congress must have intended that the provision in question benefit the plaintiff. Second, the plaintiff must demonstrate that the right assertedly protected by the statute is not so “vague and amorphous” that its enforcement would strain judicial competence. Third, the statute must unambiguously impose a binding obligation on the states. In other words, the provision giving rise to the asserted right must be couched in mandatory, rather than precatory terms.Blessing v. Freestone, 520 U.S. 329, 340-41, 117 S.Ct. 1353, 137 L.Ed.2d 569 (1997). Here, however, we need not apply the three-prong Blessing test because the Supreme Court already has told us that Title VI does not create the claimed right.
11. The partial dissent refers to “[t]he majority's notion that regulations are valid only if they flesh out a specific statutory provision.” See infra at 9288. We express no such notion. Indeed, we express no opinion as to the validity of any regulations. Rather, we hold that the disparate-impact regulation does not create a right enforceable through § 1983, particularly in light of the Supreme Court's Sandoval and Gonzaga decisions.
12. The district court's award of costs to Sound Transit as the prevailing party is reviewed for an abuse of discretion. Sea Coast Foods, Inc. v. Lu-Mar Lobster and Shrimp, Inc., 260 F.3d 1054, 1058 (9th Cir.2001).
1. This definition is schematic, and does not account for variants in which (1) as here, one of the “individuals” is another level of government; or (2) the “state” is embodied in the Constitution or another binding document, regulating government officials in their relationships with individuals.
2. The framers of the Bill of Rights, like the largely overlapping group of Founders who subscribed to the ideas expressed in the Declaration of Independence, believed that many of the rights expressly recognized therein were natural rights, that is, they inhered in individuals independent of the existence of civil society. The Bill of Rights was an attempt to create civil analogues to these natural rights, and thereby to safeguard those natural rights from government encroachment. See Erwin Chemerinsky, Constitutional Law Principles and Policies § 6.4.2, at 390 (1997).
3. See e.g. Akhil Reed Amar and Renée B. Lettow, Fifth Amendment First Principles: The Self-Incrimination Clause, 93 Mich. L.Rev. 857 (1995); Eben Moglen, Taking the Fifth: Reconsidering the Origins of the Constitutional Privilege Against Self-Incrimination, 92 Mich. L.Rev. 1086, 1121-23 (1994); R. Carter Pittman, The Colonial and Constitutional History of the Privilege Against Self-Incrimination in America, 21 Va. L.Rev. 763, 788-89 (1935).
4. Section 7 of that Act provides:Employees shall have the right to self-organization, to form, join, or assist labor organizations, to bargain collectively through representatives of their own choosing, and to engage in other concerted activities for the purpose of collective bargaining or other mutual aid or protection, and shall also have the right to refrain from any or all of such activities except to the extent that such right may be affected by an agreement requiring membership in a labor organization as a condition of employment as authorized in section 158(a)(3) of this title.29 U.S.C. § 157.
5. “Right of action” is an odd locution that signifies a concept at the intersection of statutory standing and pleading: An individual has a “right of action” if he or she can affirmatively prosecute in court a cognizable cause of action. See Black's Law Dictionary (7th ed.1999) (defining “cause of action” as “group of operative facts giving rise to one or more bases for suing;” and defining “right of action” as “the right to bring a specific case to court”); 1 Am.Jur.2d Actions § 2 (1962). As such, the term “right of action” is a technical, jurisprudential one, signifying only the courts' authority to address certain complaints at the behest of certain individuals.
6. This is not necessarily the case. An agency may not simply repeal a rule (and thereby extinguish a concomitant right). To the contrary, it must go through the same notice and comment rule-making required to promulgate a rule in the first place. See Motor Vehicle Mfrs. Ass'n of the United States v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 41, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983). This process can take years.
7. Long was not, of course, a § 1983 case but one brought under the ADA. See id. at 920.
8. Long, discussed above, illustrates the same point. See 267 F.3d at 922-23. It is simply not the case that when Congress drafted the ADA, it intended that hotel doorways be thirty-two inches wide. To the contrary, Congress prohibited disability discrimination in general terms, see 42 U.S.C. §§ 12182(a) and 12182(b)(2)(A)(iv), and then directed the Attorney General to promulgate regulations implementing the ADA's policies, 42 U.S.C. § 12186(b). Congress, in other words, gave the Attorney General a guiding policy but then left him free to determine how best to implement that policy.
9. It should be apparent that while I agree with the Eleventh Circuit's characterization of the regulation at issue in Buckley, I do not agree with its understanding that regulations so related to their enabling statutes cannot create rights enforceable under § 1983.
10. As Sandoval noted, “five Justices [in Guardians] ․ voted to uphold the disparate-impact regulations․” 532 U.S. at 283, 121 S.Ct. 1511. We are bound by the majority view unless and until the Court reconsiders the issue and by majority vote decides otherwise. See also Alexander v. Choate, 469 U.S. 287, 293, 105 S.Ct. 712, 83 L.Ed.2d 661 (1985) (characterizing Guardians as having held that “actions having an unjustifiable disparate impact on minorities could be redressed through agency regulations designed to implement the purposes of Title VI”).
11. The statute at issue in Chrysler, the Trade Secrets Act, prohibited any “officer or employee of the United States or of any department or agency thereof,” from publishing, divulging, disclosing or making known “in any manner or to any extent not authorized by law any information coming to him in the course of his employment or official duties․” 441 U.S. at 294, 99 S.Ct. 1705 (quoting 18 U.S.C. § 1905). The regulations, on the other hand, were promulgated in response to Executive Orders 11246 and 11375, and required government contractors to supply the Department of Labor's Office of Federal Contract Compliance Programs with information about the contractors' affirmative action programs in order to ensure that the contractors were providing equal employment opportunities. Chrysler, 441 U.S. at 286, 99 S.Ct. 1705. Chrysler began with the recognition that the regulations could be “laws” as that term was used in the Trade Secrets Act, although it ultimately concluded that the particular regulations at issue did not have the force and effect of law.
12. It is worth, perhaps, correcting the historical record in one regard; Thiboutot was not the first case in which the Court held that § 1983 applied to violations of federal statutes. Instead, Thiboutot was in the main a stare decisis decision, listing at some length the many prior cases which had permitted § 1983 actions to vindicate rights secured by statutes. See 448 U.S. at 5-6, 100 S.Ct. 2502 (collecting cases).
13. Gonzaga seems to be confined to Spending Clause statutes. Compare 536 U.S. at 280, 122 S.Ct. 2268 and Pennhurst State School and Hosp. v. Halderman, 451 U.S. 1, 28, 101 S.Ct. 1531, 67 L.Ed.2d 694 (1981), with Livadas v. Bradshaw, 512 U.S. 107, 117, 114 S.Ct. 2068 (1994) and Golden State Transit Corp. v. City of Los Angeles, 475 U.S. 608, 618, 106 S.Ct. 1395, 89 L.Ed.2d 616 (1986), both holding that a non-Spending Clause statute can create rights by structural implication.
14. In its entirety, § 21.5(b)(2) reads as follows:A recipient, in determining the types of services, financial aid, or other benefits, or facilities which will be provided under any such program, or the class of person to whom, or the situations in which, such services, financial aid, other benefits, or facilities will be provided under any such program, or the class of persons to be afforded an opportunity to participate in any such program; may not, directly or through contractual or other arrangements, utilize criteria or methods of administration which have the effect of subjecting persons to discrimination because of their race, color, or national origin, or have the effect of defeating or substantially impairing accomplishment of the objectives of the program with respect to individuals of a particular race, color, or national origin.
15. Plaintiffs do not suggest that we can construe the regulation's repetition of the very same language that appears in the governing statute as having a different meaning than those same words as they appear in the statute, and I do not see how we could.
16. I recognize that Justice O'Connor's opinion in Guardians, 463 U.S. at 612-15, 103 S.Ct. 3221, alluded to by the majority in Sandoval, 532 U.S. at 283, 121 S.Ct. 1511, disapproved the analogy between Congress' authority under Section 5 and the discretion accorded an administrative agency in promulgating regulations. Three members of the Court in Guardians, however, were of the opposite view. See Guardians, 463 U.S. at 643-45, 103 S.Ct. 3221 (Stevens, J., dissenting). Further, Justice O'Connor's concurrence in Guardians did not specifically consider whether a disparate impact regulation could be valid as an administratively effective way of precluding funding in circumstances in which a discriminatory intent is operative, given both the factual difficulties of determining such intent and the consideration that the government should have considerable discretion in adopting methods of assuring compliance with the rules governing the distribution of federal funds.
17. I agree with the plaintiffs that in other circumstances-under Title VII, for example-the prohibition upon actions taken with a disparate impact does run to individuals and create a right in each person in a protected class to be free from the impact of such actions. The problem is that § 21.5(b)(2) simply does not so state.
18. This does not mean, however, that private parties like SOV are without recourse. As the analysis above indicates, the DOT provides administrative remedies for individuals wronged by a recipient's conduct. In addition, inasmuch as Title VI and the DOT's implementing regulations establish the terms of a contract between the federal government and a recipient of federal funds, private parties like SOV may seek redress as third-party beneficiaries in a state cause of action. See Guardians, 463 U.S. at 599, 103 S.Ct. 3221 (opinion of White, J.); id. at 633, 103 S.Ct. 3221 (Marshall, J., dissenting).
Opinion by Judge GOULD; Dissent by Judge BERZON.