GERARD E. LYNCH, Circuit Judge:

Plaintiffs-Appellants the Building Industry Electrical Contractors Association ("BIECA") and United Electrical Contractors Association ("UECA") appeal the dismissal of their complaint challenging a number of agreements entered into by the City of New York with respect to labor conditions on certain City construction projects. Appellants argue that the agreements regulate the labor market and are therefore preempted by the National Labor Relations Act. We find the project labor agreements in this case materially indistinguishable from agreements the Supreme Court found permissible under the market participant exception to preemption in Building and Construction Trades Council of Metropolitan District v. Associated Builders and Contractors of Massachusetts/Rhode Island Inc., 507 U.S. 218, 113 S.Ct. 1190, 122 L.Ed.2d 565 (1993) ("Boston Harbor"). Because the City acted as a market participant and not a regulator in entering the agreements, its actions fall outside the scope of NLRA preemption. We therefore affirm the judgment of the district court.

BACKGROUND

The contracts at issue in this case are project labor agreements ("PLAs"), large-scale contracts common in the construction industry. PLAs typically select a union to represent workers on a project and are often signed before construction begins. Although pre-hire agreements normally violate the National Labor Relations Act's requirement that a collective bargaining representative be selected by a majority vote of workers, see 29 U.S.C. § 159(a), Congress created an exception for the construction industry because of the unique conditions of that industry. Construction is characterized by "seasonal work, jobs of brief duration, and employees who typically work for many employers and for none of them continuously." Building and Construction Trades Department, AFL-CIO Br. at 6 (internal quotation marks and citation omitted). These conditions convinced Congress, as part of the 1959 Landrum-Griffin Act, Pub.L. No. 86-257, to enact special provisions governing the construction industry, which codified and expressly legalized contemporary practice in that industry. See Woelke & Romero Framing, Inc. v. NLRB, 456 U.S. 645, 652-60, 102 S.Ct. 2071, 72 L.Ed.2d 398 (1982); see also Boston Harbor, 507 U.S. at 230, 113 S.Ct. 1190.

Although public employers such as the City are excluded from the NLRA's coverage, see 29 U.S.C. § 152(2), two key provisions of the Landrum-Griffin Act provide important background on the use of prehire agreements in the construction industry. First, the Act added to the NLRA Section 8(f), codified at 29 U.S.C. § 158(f), which permits unions and employers in the construction industry to enter into collective bargaining agreements before the union establishes majority status by vote. Second, the Act amended NLRA Section 8(e), codified at 29 U.S.C. § 158(e), to except the construction industry from the usual prohibition on agreements between an employer and union to refrain from doing business with a third party. The amended Section 8(e) allows unions and employers to agree that only contractors who sign particular contracts, such as a project's PLA, will be permitted to work on a site. Together, Sections 8(e) and 8(f) allow the manager of a construction project to enter into a comprehensive agreement that sets common employment terms to govern the many different trades involved in a construction project. The two provisions help solve the problems otherwise created by the specific conditions in the construction industry, including "the short-term nature of employment which makes posthire collective bargaining difficult, the contractor's need for predictable costs and a steady supply of skilled labor, and a longstanding custom of prehire bargaining in the industry." Boston Harbor, 507 U.S. at 231, 113 S.Ct. 1190.

Though large in scope and dollar amount, the PLAs in this case ("City PLAs") contain terms typical of PLAs. Initially announced in November 2009, they have been estimated to cover about half of the City's construction projects over the five years between 2009 and 2014, and provide that the covered projects will be serviced by contractors recognizing the Building and Construction Trades Council of Greater New York and Vicinity ("BCTC") and its affiliates as the sole bargaining representatives for all construction workers on PLA-covered projects. The BCTC is affiliated with Local Union No. 3, International Brotherhood of Electrical Workers, AFL-CIO ("Local 3"), and the PLAs incorporate favorable terms for members of Local 3, which will provide the lion's share of the electrical labor on projects covered by the City PLAs. In addition to the requirement that contractors on PLA-covered projects recognize BCTC affiliates as the collective bargaining representatives for project employees, relevant common terms of the City PLAs include: union security clauses, which require employees on the PLAs to pay dues, or their equivalent, whether or not they join the relevant BCTC-affiliated union; a requirement for any signatory contractor to secure at least 88% of its labor through BCTC affiliates' hiring halls; prohibitions on unions' discriminating in referrals based on union affiliation; requirements that contractors contribute to affiliated union fringe benefit funds; standard work rules and hours; and no-strike clauses and dispute resolution systems. It is undisputed that these terms are not materially different from those in other private and public PLAs. A contractor wishing to obtain work under one of the City PLAs must sign a letter of assent which binds the contractor to the PLA's terms and specifies that where the PLA conflicts with the contractor's collective bargaining agreement ("CBA"), the PLA will govern.

BIECA has a collective bargaining agreement with a different, non-BCTC-affiliated union—Local 363, United Service Workers Union—under which Local 363 has the right to provide labor on BIECA's contracts. This and other terms in BIECA's collective bargaining agreement will make it difficult for BIECA to seek work under the City PLAs. As discussed more fully below, BIECA argues that the PLAs effectively punish BIECA for its bargaining agreement with Local 363. Similarly, UECA has been engaged in an ongoing labor dispute with Local 3, but the PLAs will set the terms under which UECA members can employ Local 3 members on PLA projects. This, UECA argues, improperly interferes with UECA's right to freely negotiate its collective bargaining agreements.1

Plaintiffs BIECA and UECA brought suit in the United States District Court for the Southern District of New York (Robert P. Patterson, J.). They argued principally that the PLAs are preempted by the NLRA, and also asserted related causes of action under federal and state law. Defendants moved to dismiss pursuant to Federal Rules of Civil Procedure 12(b)(6) and (1). After holding that the PLAs are not preempted, the district court found that the remaining federal claims lacked merit and dismissed them; it then declined to exercise supplemental jurisdiction over the state claims. Building Industry Elec. Contractors Ass'n v. City of New York, No. 10 Civ. 8002, 2011 WL 3427138 (S.D.N.Y. Aug. 5, 2011). This appeal followed.

DISCUSSION

We review de novo the district court's dismissal of an action for failure to state a claim under Rule 12(b)(6), e.g., In re Citigroup ERISA Litig., 662 F.3d 128, 135 (2d Cir.2011), or for lack of jurisdiction under Rule 12(b)(1), see Moore v. PaineWebber, Inc., 189 F.3d 165, 169 n. 3 (2d Cir.1999). Dismissal is appropriate if the complaint fails to state a claim "plausible on its face." Ashcroft v. Iqbal, 556 U.S. 662, 678, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009). In assessing the legal sufficiency of the claim, we must accept factual allegations in the complaint as true and draw all reasonable inferences in favor of the nonmoving party. DiFolco v. MSNBC Cable L.L.C., 622 F.3d 104, 111 (2d Cir.2010); see also Citigroup, 662 F.3d at 135. We will consider only the complaint and any documents attached thereto or incorporated by reference and "documents upon which the complaint `relies heavily.'" Citigroup, 662 F.3d at 135, quoting DiFolco, 622 F.3d at 111.

Although the complaint asserted five causes of action and the district court dismissed them all, BIECA concedes that the threshold and potentially dispositive question before this court is whether the district court correctly held that the PLAs are not preempted, since all of the district court's rulings rest on that conclusion.

BIECA argues that the PLAs are invalid because they are preempted by the elaborate regulatory scheme set out in the National Labor Relations Act. "States may not regulate activity that the NLRA protects, prohibits, or arguably protects or prohibits," Healthcare Ass'n of N.Y. State, Inc. v. Pataki, 471 F.3d 87, 95 (2d Cir. 2006) (internal quotation marks omitted), nor may they "interfere with the open space created by the NLRA for the free play of economic forces," id. at 107 (internal quotation marks omitted).

But "pre-emption doctrines apply only to state regulation." Boston Harbor, 507 U.S. at 227, 113 S.Ct. 1190. Congress did not intend states' decisions about how to spend their own money as participants in the labor market to be subject to the same scrutiny as state regulation of the private labor market. "When a State owns and manages property," i.e., when it is a proprietor, "it must interact with private participants in the marketplace." Id. Such marketplace interactions are not regulation and so are not normally subject to preemption analysis at all. See Healthcare Ass'n, 471 F.3d at 108 ("A major limitation on the labor law preemption doctrines is the principle that state conduct will not be preempted if the state's actions are proprietary, rather than regulatory.").

The clearest illustration of the market participant exception is a case with facts nearly identical to this one: the Supreme Court's Boston Harbor decision. In Boston Harbor, contractors who did not use union labor challenged a project labor agreement covering public works and environmental projects to clean up Boston Harbor worth $6.1 billion over ten years. 507 U.S. at 221-23, 113 S.Ct. 1190. The PLA designated the Building and Construction Trades Council as the bargaining agent for all employees, designated BCTC-affiliated hiring halls as the primary source of labor, required all employees to join a BCTC-affiliated union within seven days of employment or pay union agency fees, and required contractors to be bound by the PLA. Id. at 221-22, 113 S.Ct. 1190.

The nonunion contractor plaintiffs argued that the PLA was preempted by, among other statutes, the NLRA. However, the 1959 amendments to the NLRA, Sections 8(e) and 8(f), indisputably permitted private parties to enter into prehire agreements with the same restrictions and effects on nonunion contractors. Id. at 230, 113 S.Ct. 1190. The Court noted that the amendments authorizing PLAs were enacted to accommodate the specific needs of the construction industry. Id. at 231, 113 S.Ct. 1190. The Court found "no reason to expect [the] defining features of the construction industry to depend upon the public or private nature of the entity purchasing contracting services." Id. Therefore, the Court concluded that Congress could not have intended to forbid public entities from entering precisely the same agreements that a private developer could. Id. at 231-32, 113 S.Ct. 1190. Because the state agency was acting precisely as a private developer would to efficiently obtain services for its construction projects, the Court held that preemption was simply inapplicable—the state was purchasing, not regulating. See id. at 232-33, 113 S.Ct. 1190 ("There is no question but that [the state agency] was attempting to ensure an efficient project that would be completed as quickly and effectively as possible at the lowest cost.").

Boston Harbor appears to decide this case. BIECA makes two principal efforts to distinguish it. First, it argues that because its union contractor members are bound by CBAs, the City PLAs in this case have such dramatic extracontractual effect that they are tantamount to regulation and thus preempted. Second, it argues that these agreements fall outside the market participant exception because the City was motivated to enter them not by efficiency, but by political cronyism. Neither argument is persuasive.

I. Extracontractual Effect on Union Contractors

First, BIECA argues that as an association of union contractors, as opposed to the nonunion contractor plaintiffs in Boston Harbor, the City PLAs interfere with its right to collectively bargain with union workers. When a PLA requires a nonunion contractor to adopt employment terms contrary to its normal practices, it may simply adopt those terms; by contrast, a union contractor is bound by a collective bargaining agreement that cannot be changed unilaterally or at a moment's notice. The PLA therefore has an "extracontractual" effect on the union contractor: it must alter the CBA that governs all of its projects in order to do work under the PLAs.

In support of this argument, BIECA correctly notes that the market participant exception does not immunize from scrutiny any choice a state makes about expending state funds on state projects. A state cannot use its spending power to regulate labor. For example, in Wisconsin Department of Industry, Labor and Human Relations v. Gould Inc., 475 U.S. 282, 287-88, 106 S.Ct. 1057, 89 L.Ed.2d 223 (1986), the Supreme Court held that the NLRA preempted a state law forbidding the state from doing business with repeat NLRA violators, because the uncontroverted purpose of the law was to deter NLRA violations—the unique province of the federal government. Similarly, in Chamber of Commerce of United States v. Brown, 554 U.S. 60, 62, 69, 128 S.Ct. 2408, 171 L.Ed.2d 264 (2008), the Court struck down a state law prohibiting recipients of state funds from using that money to "to assist, promote, or deter union organizing." The law did not fall within the market participant exception because it was "neither `specifically tailored to one particular job' nor a `legitimate response to state procurement constraints or to local economic needs,'" id. at 70, 128 S.Ct. 2408, quoting Boston Harbor, 507 U.S. at 232, 113 S.Ct. 1190, and its legislative purpose was "not the efficient procurement of goods and services, but the furtherance of a labor policy," id.

BIECA may also be correct that a state law or contract with profound effects outside of the state's market interest in the transaction would be preempted. See Healthcare Ass'n, 471 F.3d at 102 ("[A] State cannot leverage its money to affect the contractor's protected activity beyond the contractor's dealings with the State."); Cardinal Towing & Auto Repair, Inc. v. City of Bedford, Tex., 180 F.3d 686, 693 (5th Cir.1999) (holding that a state's action is more likely to be proprietary than regulatory when the "narrow scope of the challenged action defeat[s] an inference that its primary goal was to encourage a general policy rather than address a specific proprietary problem").2 But the supposed extracontractual effect of the City PLAs that Appellants point to in this case—the "requirement" that they renegotiate their CBAs—is entirely self-inflicted. BIECA is free to renegotiate its CBA with Local 363 and its other signatory unions. BIECA is equally free to decline work on City PLA projects and continue to work under its existing CBAs on other private projects, or on City projects not covered by the PLAs.

BIECA resists this conclusion by analogizing its CBA to the NLRA-violator status of contractors in Gould. It argues that just as the Gould law affected extracontractual NLRA violations, these PLAs affect extracontractual union affiliation decisions. But this argument misses the point of Gould, which centered on regulatory purpose rather than effect. The Gould law did not "essentially reflect the entity's own interest in its efficient procurement of needed goods and services." Cardinal Towing, 180 F.3d at 693. Instead, the law represented a policy of discouraging NLRA violations. See Gould, 475 U.S. at 287-88, 106 S.Ct. 1057 ("The State concedes, as we think it must, that the point of the statute is to deter labor law violations and to reward fidelity to the law." (internal quotation marks omitted)). The Supreme Court reasonably found that Congress intended such policy decisions to be an exclusively federal matter. Id. at 291, 106 S.Ct. 1057.

Moreover, BIECA's extracontractual effect argument is plainly foreclosed by Boston Harbor, decided several years after Gould and factually indistinguishable from this case. The Boston Harbor PLA surely put pressure on the plaintiff contractors' extracontractual decision to employ nonunion workers. But preemption is a matter of Congress's intent, and the Court found it implausible that Congress intended to foreclose that type of so-called extracontractual effect in public sector construction, given that it had expressly ratified identical effects in the private sector by enacting Sections 8(e) and (f). Boston Harbor, 507 U.S. at 230-32, 113 S.Ct. 1190.

In sum, we see no reason to distinguish Boston Harbor simply because it dealt with nonunion contractors. We recognize that, as compared to nonunion contractors, it may be more difficult for BIECA's members to comply with the PLAs' terms where those terms differ from their usual practice. But this difference does not alter the market participant analysis. The effects that BIECA complains of are entirely ordinary consequences of PLAs, in private as well as public contracts. Project labor agreements create winners and losers among contractors and labor unions. Congress foresaw and weighed these consequences when it expressly legalized construction industry PLAs. See generally Woelke & Romero Framing, 456 U.S. at 654-60, 102 S.Ct. 2071 (discussing legislative history of § 8(e)'s exception for the construction industry). Or, as explained by then-Chief Judge Breyer of the First Circuit, dissenting from the en banc ruling that the Supreme Court reversed in Boston Harbor,

935 F.2d 345, 361 (1st Cir.1991) (Breyer, C.J., dissenting), quoted in Boston Harbor, 507 U.S. at 233, 113 S.Ct. 1190.3

II. Alleged Favoritism

BIECA next argues that the PLAs cannot fall within the market participant exception, since the City's purpose in agreeing to them was not to achieve cost reductions, but to further the interests of the BCTC and its affiliates, which were politically aligned with City officials' interests. At a minimum, BIECA argues, the dispute over the City's motive is an issue of material fact that rendered dismissal inappropriate. This argument misapprehends both the means for determining governmental purpose and the bounds of the market exception itself.

First, when a court assesses whether a governmental policy has a regulatory purpose, it looks primarily to the objective purpose clear on the face of the enactment, not to allegations about individual officials' motivations in adopting the policy. We will not search for an impermissible motive where a permissible purpose is apparent, because "[f]ederal preemption doctrine evaluates what legislation does, not why legislators voted for it or what political coalition led to its enactment." N. Ill. Chapter of Associated Builders & Contractors, Inc. v. Lavin, 431 F.3d 1004, 1007 (7th Cir.2005). This principle is true across many fields of substantive law. See, e.g., Mueller v. Allen, 463 U.S. 388, 394-95, 103 S.Ct. 3062, 77 L.Ed.2d 721 (1983) (noting that for Establishment Clause analysis, the Court is "reluctan[t] to attribute unconstitutional motives to the States, particularly when a plausible secular purpose for the state's program may be discerned from the face of the statute"). Our reluctance is motivated in no small part by separation of powers concerns: "Judicial inquiries into legislative or executive motivation represent a substantial intrusion into the workings of other branches of government. Placing a decisionmaker on the stand is therefore usually to be avoided." Vill. of Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252, 268, 97 S.Ct. 555, 50 L.Ed.2d 450 (1977) (internal quotation marks and citations omitted). Other circuits have expressed reluctance to examine official motives in the PLA context. See Colfax Corp. v. Ill. State Toll Highway Auth., 79 F.3d 631, 635 (7th Cir.1996) (declining to "go behind the contract to determine whether the ... real, but secret motive was to regulate labor"); Johnson v. Rancho Santiago Cmty. Coll. Dist., 623 F.3d 1011, 1026-27 (9th Cir.2010). Moreover, it is difficult to discern officials' motives given the complex workings of government. See Bush v. Vera, 517 U.S. 952, 1012-14 & n. 9, 116 S.Ct. 1941, 135 L.Ed.2d 248 (1996) (Stevens, J., dissenting) (discussing difficulty of determining legislature's "predominant" motive for passing a law, and collecting cases). We decline BIECA's suggestion that its bare allegation of cronyism should permit an elaborate inquiry into individual governmental officials' motivations. See Iqbal, 556 U.S. at 678-79, 129 S.Ct. 1937 ("Rule 8... does not unlock the doors of discovery for a plaintiff armed with nothing more than conclusions.").

Second, even if we were more willing to scrutinize the City's motives for entering into the challenged PLAs, BIECA's assertion that the City's interest is noneconomic rests unpersuasively on the claim that the PLAs do not achieve the lowest possible costs for the City, because doing business with BIECA and UECA would in some instances achieve even greater savings than the PLAs' terms provide. Appellants misunderstand the purpose that satisfies the market participant exception. It cannot be correct that to qualify for the exception, the City must show that its contracts are maximally efficient. "Acting like a proprietor" does not mean acting exclusively with the narrow goal of minimizing costs regardless of consequences. Private proprietors are entitled to, and sometimes do, prefer working with familiar faces or contracting with larger entities that can consistently and simply provide all their requirements. These practices may reflect long-run economic rationality. But even if they do not, accepting BIECA's restrictive theory of economic rationality would expose every public PLA to challenge based on speculative ex post evaluations of whether the PLAs in fact proved economically prudent. Cf. Rancho Santiago, 623 F.3d at 1025 (noting that a state interest in "efficient procurement" satisfying market participant exception "does not necessarily mean `cheap' procurement, but rather `procurement that serves the state's purposes'"), quoting Engine Mfrs. Ass'n v. S. Coast Air Quality Mgmt. Dist., 498 F.3d 1031, 1046 (9th Cir. 2007).

Similarly, BIECA's allegation that the City relied on poor or even biased expert data to justify the PLAs, if true, might raise concerns about the efficacy of the City's contracting process. But, as BIECA concedes, "[a]n inept proprietor is still a proprietor." Appellant Br. at 46. The allegation, even if true, does not demonstrate that the City was regulating. We find persuasive the Ninth Circuit's reasoning in Rancho Santiago, a similar case in which nonunion contractors challenged a pre-hire construction labor agreement:

623 F.3d at 1025.

It bears repeating that BIECA's theory of the case is preemption: that the PLAs are not contracts but regulations and therefore are preempted by the NLRA. It is hard to see why, even if political favoritism was a motivating factor in the City's decision to contract with particular contractors or unions, the PLAs would be thereby transformed from contracts into regulations. See Rancho Santiago, 623 F.3d at 1026 ("Plaintiffs contend that the [PLA]'s primary purpose was to reward the unions that supported the Measure E campaign.... [W]e are quite certain that Congress did not intend for the NLRA's or ERISA's preemptive scope to turn on state officials' subjective reasons for adopting a regulation or agreement.").4

CONCLUSION

"In the absence of any express or implied indication by Congress that a State may not manage its own property when it pursues its purely proprietary interests, and where analogous private conduct would be permitted, this Court will not infer such a restriction." Boston Harbor, 507 U.S. at 231-32, 113 S.Ct. 1190. The PLAs challenged here represent the City's permissible proprietary choice; the City has behaved just as any other major landowner or developer might to secure labor for many of its construction projects. Because the PLAs are market activity and not regulation, the preemption claim must fail.

The judgment of the district court is therefore AFFIRMED.