Iowa League of Cities v. Environmental Protection Agency

GRUENDER, Circuit Judge. The Iowa League of Cities (“League”) seeks direct appellate review of two letters sent by the Environmental Protection Agency (“EPA”) to Senator Charles Grass-ley. The League argues that these letters effectively set forth new regulatory requirements with respect to water treatment processes at municipally owned sewer systems. According to the League, the EPA not only lacks the statutory authority to impose these regulations, but it violated the Administrative Procedures Act (“APA”), 5 U.S.C. § 500 et seq., by implementing them without first proceeding through the notice and comment procedures for agency rulemaking. We find that we have subject matter jurisdiction over the claims, and we vacate under APA section 706(2)(C), (D). I. Background The League previously sought our review in 2010 of six EPA documents, consisting of letters, internal memoranda, and a Federal Register notice, that allegedly constituted new regulatory requirements for water treatment processes. The EPA moved to dismiss, arguing that judicial review was premature because the documents were part of an ongoing agency decisionmaking process. An administrative panel of this court granted the EPA’s motion to dismiss for lack of subject matter jurisdiction. The League continued to perceive a conflict between the agency’s official written policies and the expectations it was transmitting to the state entities that served as liaisons between the EPA and municipal wastewater treatment facilities. Consequently, the League enlisted the assistance of Senator Charles Grassley to obtain clarification from the EPA. The EPA sent two letters (“June 2011 letter” and “September 2011 letter”) in response to Senator Grassley’s inquiries. According to the EPA, these guidance letters merely discuss existing regulatory requirements. The League disagrees, viewing the letters as contradicting both the Clean Water Act (“CWA”), 33 U.S.C. § 1251 et seq., and the EPA’s lawfully promulgated regulations. As it did in 2010, the EPA moved to dismiss for lack of subject matter jurisdiction. This time an administrative panel denied the motion but requested that the parties address the merits of all relevant jurisdictional and substantive arguments. 1 *855 The APA “empowers federal courts to ‘hold unlawful and set aside agency action, findings, and conclusions’ if they fail to conform with any of six specified standards.” Marsh v. Or. Natural Res. Council, 490 U.S. 360, 375, 109 S.Ct. 1851, 104 L.Ed.2d 377 (1989) (quoting 5 U.S.C. § 706(2)). Inter alia, a reviewing court may set aside agency action that has failed to observe those “procedure[s] required by law.” § 706(2)(D). Agencies must conduct “rule making” in accord with the APA’s notice and comment procedures. 5 U.S.C. § 553(b), (c). However, only new “legislative” rules are required to be created pursuant to notice and comment rule-making. See id.; see also Minnesota v. Ctrs. for Medicare & Medicaid Servs., 495 F.3d 991, 996 (8th Cir.2007). “Interpretative rules” 2 and “general statements of policy” are statutorily exempt from the procedural requirements applicable to “rule making.” See § 553(b)(3)(A); see also Shalala v. Guernsey Mem’l Hosp., 514 U.S. 87, 99, 115 S.Ct. 1232, 131 L.Ed.2d 106 (1995). The crux of the League’s procedural claim is that the EPA’s letters announced new legislative rules for water treatment processes at municipally owned sewer systems, thereby modifying the EPA’s existing legislative rules. The EPA admits it did not engage in notice and comment procedures, but it insists there has been no procedural impropriety because the letters should be considered general policy statements or, at most, interpretative rules. The League asks us to find not only that the EPA’s actions are procedurally invalid but also to go one step further and set aside the rules as imposing regulatory requirements that surpass the EPA’s statutory authority. See § 706(2)(C) (authorizing federal courts to set aside agency action that is “in excess of statutory jurisdiction, authority, or limitations, or short of statutory right”). The two areas of regulation addressed in the challenged EPA letters are “mixing zones” and “blending.” Our analysis first requires a discussion of the CWA’s regulatory scheme and the water treatment processes at issue. A. The Clean Water Act The CWA forbids the “discharge of any pollutant” — defined as the “addition of any pollutant to navigable waters from any point source” 3 — unless executed in compliance with the Act’s provisions. 33 U.S.C. §§ 1311(a), 1362(12). A permit program called the National Pollution Discharge Elimination System (“NPDES”) plays a central role in federal authorization of permissible discharges. See 33 U.S.C. § 1342. The EPA may issue an NPDES permit, but states also are authorized to administer their own NPDES programs. § 1342(b). The vast majority elect to do so. 4 If a state chooses to operate its own permit program, it first must obtain EPA permission and then ensure that it issues discharge permits in accord with the same *856 federal rules that govern permits issued by the EPA. § 1342(a); 40 C.F.R. § 122.41. Many of these rules are in the form of “effluent limitations,” which “restrict the quantities, rates, and concentrations of specified substances which are discharged from point sources.” Arkansas v. Oklahoma, 503 U.S. 91, 101, 112 S.Ct. 1046, 117 L.Ed.2d 239 (1992) (citing §§ 1311, 1314). The NPDES permit system “serves to transform generally applicable effluent limitations ... into the obligations ... of the individual discharger.” EPA v. California ex rel. State Water Res. Control Bd., 426 U.S. 200, 205, 96 S.Ct. 2022, 48 L.Ed.2d 578 (1976). The EPA applies effluent limitations at the point of discharge into navigable waters, known as “end-of-the-pipe,” unless monitoring at the discharge point would be “impractical or infeasible.” 40 C.F.R. § 122.45(a), (h). The baseline effluent limitations are “technology-based,” § 1311(b); 40 C.F.R. § 125.3(a), in that they set “a minimum level of effluent quality that is attainable using demonstrated technologies.” EPA, NPDES Permit Writers’ Manual 5-1 (2010). 5 The EPA has interpreted this regime as “preclud[ing] [it] from imposing any particular technology on a discharger.” In re Borden, Inc., Decision of the General Counsel on Matters of Law Pursuant to 40 C.F.R. § 125.36(m), No. 78 (Feb. 19, 1980), at *2; see also NPDES Permit Writers’ Manual 5-14, 5-15 (“Therefore, each facility has the discretion to select any technology design and process changes necessary to meet the performance-based discharge limitations and standards specified by the effluent guidelines.”). The technology-based effluent limitations applicable to publicly-owned treatment works (“POTWs”), 6 such as municipal sewer authorities, are based on a special set of rules known as the “secondary treatment” regulations. § 1311(b)(1)(B); 40 C.F.R. § 125.3(a)(1); see generally 40 C.F.R. § 133.102 (describing average monthly and weekly “minimum level[s] of effluent quality attainable by secondary treatment”). The secondary treatment regulations also do not mandate the use of any specific type of technology to achieve their requisite levels of effluent quality. See 48 Fed. Reg. 52,258, 52,259 (Nov. 16, 1983). When technology-based effluent limitations would fall short of achieving desired water quality levels, the EPA is authorized to devise additional, more stringent water quality-based effluent limitations for those particular point sources. 33 U.S.C. § 1312(a). Thus, the CWA is a program of state and federal cooperation, but state discretion is exercised against a backdrop of significant EPA authority over state-run NPDES programs. The EPA dictates the effluent limitations applicable to all permits, while states are in charge of categorizing their waterways in terms of designated uses and setting forth “water quality standard[s]” for each type of waterway. 33 U.S.C. § 1313(c)(2). These standards supplement effluent limitations to ensure that overall water quality remains at an acceptable level. Arkansas, 503 U.S. at 101, 112 S.Ct. 1046. A major component of a state’s water quality standards is “the set of water quality criteria sufficient to support the designated uses of each water-body.” 7 NPDES Permit Writers’ Manual *857 6^4. At least every three years, states must submit their water quality standards to the EPA for approval. § 1313(c)(1). The EPA must approve the standards within sixty days or disapprove them within ninety days. 66 Fed.Reg. 11,202, 11,215 (Feb. 22, 2001). States are also required to forward a copy of each permit application they receive to the EPA, which is afforded an opportunity to block the issuance of the permit. § 1342(d); 40 C.F.R. § 123.29. In sum, states evaluate discharge permit applications under a mixture of federal regulations and their own water quality standards, crafted subject to federal approval. B. Bacteria Mixing Zones One element of state water quality standards are policies regarding “mixing zones.” The EPA has defined mixing zones as “[a] limited area or volume of water where initial dilution of a discharge takes place and where numeric water quality criteria can be exceeded.” EPA, Water Quality Handbook Ch. 5.1 (1994) (“Handbook”); see also NPDES Permit Writers’ Manual 6-15. In effect, a mixing zone allows the permit holder to create a higher concentration of pollutants in navigable waters near the immediate point of discharge, as long as the discharge is sufficiently diffused as it moves through the larger body of water. The requisite water quality criteria, then, need not be met at the end of the pipe. It is undisputed that in at least some instances, states are allowed to approve discharge permit applications that incorporate mixing zones. See 40 C.F.R. § 131.13 (“States may, at their discretion, include in their State standards, policies generally affecting their application and implementation, such as mixing zones.... ”). But as one of its water quality standards, a state’s policy on mixing zones remains subject to the triennial review of the EPA. See § 1313(c)(1). In addition, the EPA has the authority to veto any permit application incorporating what it views as an inappropriate mixing zone. See § 1342(d)(2). Mixing zones are addressed in one of the EPA’s regulations, 40 C.F.R. § 122.44(d)(1). Subparagraph (ii) of that regulation describes the procedures a state should apply when determining whether a discharge would cause — or contribute to causing — a body of water to deviate from the state’s water quality criteria, thereby necessitating the imposition of water-quality based effluent limitations on that discharge (in addition to the default technology-based effluent limitations already in effect). See 54 Fed.Reg. 23,868, 23,872 (June 2, 1989). In particular, state permitting authorities should consider “any dilution of the effluent in the receiving water, after considering mixing zones if applicable.” Id. Although some commentators responded to the proposal of subparagraph (ii) by requesting that the EPA prohibit mixing zones, the EPA subsequently reiterated that the “use of mixing zones raises issues that are more appropriately addressed in the state water quality standards adoption process,” and therefore it would retain “the reference to mixing zones in paragraph (d)(l)(ii).” Id. The League portrays 40 C.F.R. § 122.44(d)(1)(h) as channeling any federal objections to mixing zones, including mixing zones for bacterial effluents (“bacteria mixing zones”), through the EPA’s process of approving or rejecting state water quality standards. The June 2011 letter admits that, pursuant to 40 C.F.R. § 131.13, states “may, at *858 their discretion, include mixing zone policies in their state water quality standards.” Citing a 2008 memorandum from the Director of the EPA’s Office of Science and Technology to a regional EPA director (“the King memorandum”), however, the June 2011 letter then recites “the EPA’s long-standing policy” that all bacteria mixing zones in waters designated for “primary contact recreation” carry potential health risks and flatly states that they “should not be permitted.” The letter further acknowledges that the EPA “does not have additional regulations specific to mixing zones,” but it then refers the reader to the additional “recommendations regarding the use of mixing zones” in policy guidance such as the Handbook. The Handbook encourages states to incorporate a “definitive statement” into their water quality standards regarding “whether or not mixing zones are allowed” and, if they are, to “utilize a holistic approach to determine whether a mixing zone is tolerable.” Ch. 5.1, 5.1.1. The Handbook cautions, however, that mixing zones must be utilized in ways that “ensure ... there are no significant health risks, considering likely pathways of exposure.” Ch. 5.1. Additionally, mixing zones “should not be permitted where they may endanger critical areas,” such as “recreational areas.” Id. From the League’s perspective, states are able to approve bacteria mixing zones, even in waters designated as “primary contact recreation,” so long as site-specific factors create scenarios in which there are no health risks and recreational areas are not endangered. The EPA argues that the June 2011 letter is consistent with the Handbook, which explicitly envisioned limitations on mixing zones in recreational areas. C. Blending The second contested regulatory area involves “blending.” POTWs typically move incoming flows through a primary treatment process and then through a secondary treatment process. Most secondary treatment processes are biological-based, but the secondary treatment regulations do not “specify the type of treatment process to be used to meet secondary treatment requirements nor do they preclude the use of non-biological facilities.” 8 68 Fed.Reg. 63,042, 63,046 (Nov. 7, 2003). At many POTWs, primary treatment capacity exceeds secondary treatment capacity. Biological-based processes in particular are sensitive to deviations in volume of flow and pollutant level. Correspondingly, during periods of rain and snow, large influxes of stormwater can overwhelm a facility’s standard biological secondary treatment processes, potentially rendering them inoperable. Id. Blending can prevent this, by channeling a portion of “peak wet weather flows” around biological secondary treatment units and through non-biological units, recombining that flow with its counterpart that traveled through the biological units, and then discharging the combined stream. Id. at 63,045. Just like non-blended streams, the combined output must still comply with all applicable effluent limitations, including the water quality levels specified in the secondary treatment regulations. Id. at 63,047. Some members of the League wish to incorporate a method of treatment called ACTIFLO into the secondary treatment procedures at their wastewater treatment facilities. ACTIFLO units employ non-biological processes and are used as auxiliary secondary treatment units for peak wet weather flows. 9 The parties disagree *859 on the circumstances in which the CWA and EPA regulations permit the use of ACTIFLO. The League views ACTIFLO as a permissible technology within a POTW facility, as long as the overall output from the secondary treatment phase meets the effluent limitations imposed by the secondary treatment regulations. The EPA, on the other hand, views ACTIFLO as an impermissible “diversion” from traditional biological secondary treatment facilities. All issued permits must comply with federal regulations regarding “bypass,” which is the “intentional diversion of waste streams from any portion of a treatment facility.” 40 C.F.R. § 122.41(m)(l). Bypass is generally prohibited unless there are “no feasible alternatives.” § 122.41(m)(4). The bypass rule “is not itself an effluent standard,” but instead it “merely ‘piggybacks’ existing requirements.” 53 Fed.Reg. 40,562, 40,609 (Oct. 17,1988). The rule’s purpose is to “ensure that users properly operate and maintain their treatment facilities ... [pursuant to applicable] underlying technology-based standards,” by requiring incoming flows to move through the facility as it was designed to be operated. Id. Like the more general secondary treatment regulations, the bypass rule does not require the use of any particular treatment method or technology. Id.; see also NRDC v. EPA, 822 F.2d 104, 123 (D.C.Cir.1987). In 2003, the EPA offered “a proposed interpretation of the bypass provision (40 CFR [§ ] 122.41 (m)) as it applies to ... blending.” 68 Fed.Reg. at 63,049. Prior to this proposal, the EPA had “not established a national policy (either through rulemaking or through non-binding guid-anee to assist in the interpretation of the bypass regulation) regarding whether and under what circumstances wet weather blending at a POTW plant would not constitute a bypass.” Id. at 63,052. The 2003 proposed policy would have “provide[d] guidance to EPA Regional and State permitting authorities ... on how EPA intends to exercise its discretion in implementing the statutory and regulatory provisions related to discharges from POTWs where peak wet weather flow is routed around biological treatment units and then blended with the effluent from the biological units prior to discharge.” Id. at 63,051. Going forward, blending “would not be a prohibited bypass and could be authorized in an NPDES permit” so long as certain enumerated conditions were met. Id. at 63,049-50. These conditions primarily focused on ensuring that the discharge met all applicable effluent limitations and water quality standards, that it passed through a primary treatment unit prior to discharge, and that a “portion of the flow [w]ould only be routed around a biological or advanced treatment unit when the capacity of the treatment unit is being fully utilized.” Id. The EPA posted the proposed policy on its website and declared its consistency with the CWA. Implicitly, the 2003 policy seemed to view the secondary treatment phase as encompassing both traditional biological secondary treatment units and auxiliary non-biological treatments for peak wet weather flows, such as ACTIFLO. Accordingly, flows sent through ACTIFLO were not being intentionally “diverted” from a process they should have gone through; instead, these excess flows were simply designated to receive a different *860 type of secondary treatment. The focus was on whether the water quality of the resulting combined discharge at the end of the secondary treatment phase met all applicable effluent limitations. Two years later, the EPA abandoned the 2003 proposal. 70 Fed.Reg. 76,013, 76,015 (Dec. 22, 2005). The EPA acknowledged recent “confusion regarding the regulatory status of peak wet weather flow diversions around secondary treatment units at POTW treatment plants” and observed that they were treated only intermittently as bypasses. Id. at 76,015. The 2005 policy announced that this type of “diversion” was now considered a bypass and would be allowed only if there were “no feasible alternatives.” Id. at 76,016. As of the creation of the EPA letters in 2011, the 2005 policy had not been finalized or otherwise officially adopted. As late as June 2010, the EPA continued to solicit input on the 2005 policy through notices in the Federal Register. See 75 Fed.Reg. 30,395, 30,401 (June 1, 2010). During the spring of 2011, the League asked the EPA whether it could use “physical/chemical treatment processes, such as Aetiflo ... to augment biological treatment and recombine the treatment streams prior to discharge, without triggering application of [the bypass rule].” The June 2011 letter responded by summarizing the EPA’s 2005 proposed policy without specifically addressing how the application of that policy would impact the use of ACTIFLO or similar processes. The League sought additional clarification on whether this response meant that AC-TIFLO could be used only if there were no feasible alternatives, which the September 2011 letter answered in the affirmative. According to the EPA, ACTIFLO units fail to “provide treatment necessary to meet the minimum requirements provided in the secondary treatment regulations at 40 CFR 133.” Because ACTIFLO by itself is not considered a satisfactory secondary treatment unit, the EPA views the practice of intentionally routing flows away from a facility’s traditional biological secondary treatment units and through AC-TIFLO as a bypass that would only be allowed upon a showing of no feasible alternatives. The League argues that by prohibiting the use of ACTIFLO internally, as one element of a facility’s secondary treatment procedures, the EPA is effectively dictating treatment design, despite the agency’s acknowledgment that the bypass rule and secondary treatment regulations do not allow for such determinations at the federal level. The League also claims that the EPA is effectively applying secondary treatment effluent limitations within a treatment facility; that is, it is applying effluent limitations to the individual streams exiting peak flow treatment units, instead of at the end of the pipe. The EPA responds that using ACTIFLO to process peak wet weather flows diverts water from biological secondary treatment units, and therefore subjecting its use to a no-feasible-alternatives analysis comports with the plain language of the bypass rule. II. Jurisdiction A. Direct appellate review The League challenges the EPA’s positions on bacteria mixing zones and blending, as set forth in the two letters, as new rules promulgated in violation of APA notice and comment requirements and as in conflict with the CWA. The APA waives sovereign immunity for suits seeking judicial review of an “[ajgency action made reviewable by statute.” 10 5 U.S.C. *861 § 704. “The CWA establishes a bifurcated jurisdictional scheme whereby courts of appeals have jurisdiction over some categories of challenges to EPA action, and the district courts retain jurisdiction over other types of complaints.” Nat'l Pork Producers Council v. EPA, 685 F.3d 738, 755 (5th Cir.2011). The League invokes CWA section 509(b)(1)(E), which vests the courts of appeals with exclusive jurisdiction to review the EPA’s “action ... in approving or promulgating any effluent limitation or other limitation under section 1311, 1312, 1316, or 1345.” 33 U.S.C. § 1369(b)(1)(E). The EPA counters that we have no jurisdiction to review guidance letters and that, in any event, its positions are consistent with the CWA. “The existence of subject-matter jurisdiction is a question of law that this court reviews de novo.” ABF Freight Sys., Inc. v. Int’l Bhd. of Teamsters, 645 F.3d 954, 958 (8th Cir.2011). In order to be timely filed, interested parties must file for review within 120 days from the date of the promulgation. § 1369(b)(1). The 120-day window to challenge promulgations begins two weeks after a document is signed. 40 C.F.R. § 23.2. Here, the letters were signed on June 30, 2011, and September 14, 2011, and therefore the time period to challenge the letters — should they be found to be promulgations — began on July 14, 2011, and September 28, 2011, respectively. The League filed for review on November 4, 2011, and thus its petition is timely. We must consider, then, whether the act of sending the letters constituted an action “promulgating any effluent limitation or other limitation.” 11 The EPA urges us to dismiss the case for lack of subject matter jurisdiction, disputing the factual basis for the League’s characterization of the letters. Because the EPA raises a factual challenge to our jurisdiction under Federal Rule of Civil Procedure 12(b)(1), “no presumptive truthfulness attaches to the [League’s] allegations, and the existence of disputed material facts will not preclude [us] from evaluating ... the merits of the jurisdictional claims.” Osborn v. United States, 918 F.2d 724, 729-30 & n. 6 (8th Cir.1990) (quoting Mortensen v. First Fed. Sav. & Loan Ass’n, 549 F.2d 884, 891 (3d Cir.1977)). 1. “[Promulgating” Neither the Supreme Court nor this court has defined the circumstances in which an agency action can be considered a promulgation. Black’s Law Dictionary defines “promulgate” as “(Of an administrative agency) to carry out the formal process of rulemaking by publishing the proposed regulation, inviting public comments, and approving or rejecting the proposal.” (8th ed. 2004). This narrow interpretation would allow direct appellate review only of rules formally promulgated through notice and comment procedures. Yet, the Supreme Court has recognized a preference for direct appellate review of agency action pursuant to the APA. See, e.g., Fla. Power & Light Co. v. Lorion, 470 U.S. 729, 745, 105 S.Ct. 1598, 84 L.Ed.2d 643 (1985) (“Absent a firm indication that Congress intended to locate initial APA review of agency action in the district courts, we will not presume that Congress intended to depart from the sound policy of placing initial APA review in the courts of appeals.”); see also Nat’l Auto. Dealers *862 Ass’n v. FTC, 670 F.3d 268, 270 (D.C.Cir.2012); Jaunich v. Commodity Futures Trading Comm’n, 50 F.3d 518, 521 (8th Cir.1995). Moreover, the Supreme Court has interpreted broadly the direct appellate review provision in CWA section 509(b)(1)(F), which authorizes review of agency “action ... in issuing or denying a permit.” The Court viewed an EPA veto of a state — issued permit to be “functionally similar” to a direct denial of a permit application by the EPA itself, and therefore held that the petitioner could bring his challenge directly to a court of appeals under section 509(b)(1)(F). Crown Simpson Pulp Co. v. Costle, 445 U.S. 193, 196, 100 S.Ct. 1093, 63 L.Ed.2d 312 (1980) (per curiam). By analogy, we are persuaded that it would be more appropriate to interpret “promulgating” to include agency actions that are “functionally similar” to a formal promulgation. See Modine Mfg. Corp. v. Kay, 791 F.2d 267, 271 (3d Cir.1986) (finding jurisdiction to review directly “the agency’s interpretation of pretreatment standards applicable to indirect dischargers” because they constituted an action “promulgating any effluent ... pretreatment standard” under CWA section 509(b)(1)(C)); see also NRDC v. EPA, 673 F.2d 400, 405 (D.C.Cir.1982) (“Our decision ... follows the lead of the Supreme Court in according section 509(b)(1) a practical rather than a cramped construction.”). In considering jurisdictional statutes similar to section 509(b)(1)(E), our colleagues on the District of Columbia Circuit have adopted a practical conception of whether an agency action constitutes a promulgation. That court has explained, “To determine whether a regulatory action constitutes promulgation of a regulation, we look to three factors: (1) the Agency’s own characterization of the action; (2) whether the action was published in the Federal Register ....; and (3) whether the action has binding effects on private parties or on the agency.” Molycorp, Inc. v. EPA 197 F.3d 543, 545 (D.C.Cir.1999) (internal citation omitted). Molycorp identifies the third factor as the “ultimate focus” of this test, and we agree that whether an agency announcement is binding on regulated entities or the agency should be the touchstone of our analysis. To place any great weight on the first two Molycorp factors potentially could permit an agency to disguise its promulgations through superficial formality, regardless of the brute force of reality. See also Cement Kiln Recycling Coal. v. EPA 493 F.3d 207, 227-28 (D.C.Cir.2007) (holding that it lacked jurisdiction to consider a purported agency “promulgation” because the document was not binding). “[A]n agency pronouncement will be considered binding as a practical matter if it either appears on its face to be binding or is applied by the agency in a way that indicates it is binding.” Gen. Elec. Co. v. EPA 290 F.3d 377, 383 (D.C.Cir.2002) (citations omitted). Thus, our functional analysis of whether an agency action constitutes a promulgation encompasses those words and deeds that bind legally or as a practical matter. Cf. South Dakota v. Ubbelohde, 330 F.3d 1014, 1028 (8th Cir.2003) (“Agency statements can be binding upon the agency absent notice-and-comment rulemaking in certain circumstances.”); Appalachian Power Co. v. EPA 208 F.3d 1015, 1021 (D.C.Cir.2000) (“[W]e have also recognized that an agency’s other pronouncements can, as a practical matter, have a binding effect.”). This includes statements prospectively restricting the agency’s discretion, see Am. Mining Cong. v. Mine Safety & Health Admin., 995 F.2d 1106, 1111 (D.C.Cir.1993), or having a “present-day binding effect” on regulated entities, thereby “conclusively disposing of certain issues,” see McLouth *863 Steel Prods. Corp. v. Thomas, 838 F.2d 1317, 1321 (D.C.Cir.1988). 12 Here, the letters can be considered “promulgations” for the purposes of establishing our jurisdiction under section 509(b)(1)(E) because they have a binding effect on regulated entities. “If an agency acts as if a document issued at headquarters is controlling in the field, if it treats the document in the same manner as it treats a legislative rule, if it bases enforcement actions on the policies or interpretations formulated in the document, if it leads private parties or State permitting authorities to believe that it will declare permits invalid unless they comply with the terms of the document, then the agency’s document is for all practical purposes ‘binding.’ ” Appalachian Power Co., 208 F.3d at 1021. In particular, the court in Appalachian Power found that the contested agency guidance before it was binding because it reflected “a position [the EPA] plans to follow in reviewing State-issued permits, a position it will insist State and local authorities comply with in settling the terms and conditions of permits issued to petitioners, a position EPA officials in the field are bound to apply.” Id. at 1022. This reasoning persuades us that the June 2011 and September 2011 letters are binding as well. First, the June 2011 letter reflects a binding policy with respect to bacteria mixing zones. In response to the League’s 2010 challenge to the EPA’s policy on mixing zones, the EPA submitted to this court a motion to dismiss, which described the King memorandum as nothing but “one office director’s view of a regula *864 tory requirement.” But in the June 2011 letter to Senator Grassley, the EPA characterized the King memorandum as reflecting “the EPA’s position.” Although the EPA coyly continues to insist that the letter is the “consummation of nothing,” something apparently was consummated between 2010 and June 2011. Furthermore, the language used to express “the EPA’s position” — “should not be permitted” — is the type of language we have viewed as binding because it “speaks in mandatory terms.” Ubbelohde, 380 F.3d at 1028; see also Gen. Elec. Co., 290 F.3d at 383 (“[T]he mandatory language of a document alone can be sufficient to render it binding ....”); cf. Catawba Cnty., N.C. v. EPA, 571 F.3d 20, 34 (D.C.Cir.2009) (per curiam) (finding that an agency memo was not binding because it “ ‘encouraged’ states to address all nine factors EPA identified, but did not require them to do so”). The League’s appendix includes several affidavits from representatives of municipal wastewater treatment facilities and the Iowa Department of Natural Resources, the state permitting authority. 13 These individuals averred that they indeed have taken the June 2011 letter at face value, interpreting it as establishing a new prohibition on bacteria mixing zones, one by which they must abide in the permit application process. We agree that private parties have “reasonably [been] led to believe that failure to conform will bring adverse consequences,” which tends to make the document binding as a practical matter. See Gen. Elec. Co., 290 F.3d at 383 (quoting Robert A. Anthony, Interpretive Rules, Policy Statements, Guidances, Manuals and the Like — Should Federal Agencies Use Them to Bind the Public?, 41 Duke L.J. 1311,1328 (1992)). The EPA asks us to believe that the June 2011 letter did not flatly prohibit the use of bacteria mixing zones in waters designated for primary contact recreation because although it intoned that states “should not” permit bacteria mixing zones in primary contact recreation areas, it nonetheless mentioned that ' under 40 C.F.R. § 131.13, states “may, at their discretion, include mixing zone policies in their state water quality standards.” With respect to bacteria mixing zones in primary contact recreation areas, we struggle to spot the surviving state discretion. The *865 letter instructs state permitting authorities to reject certain permit applications, regardless of the state’s water quality standards. The EPA’s protestations to the contrary are particularly unavailing where, as here, Iowa’s water permitting authority has received communications from the EPA indicating that it would object to any permits that were inconsistent with the policy outlined in the EPA letters. In effect, the EPA asks us to agree that when it couches an interdiction within a pro for-ma reference to state discretion, the prohibition is somehow transformed into something less than a prohibition. We decline to accept such Orwellian Newspeak. Second, the September 2011 letter presents a binding policy on blending. Although the June 2011 letter describes the “2005 draft Policy” on blending as merely “a viable path forward” that “has not been finalized,” the September 2011 letter applies the 2005 policy to the League’s proposed use of ACTIFLO. 14 In requiring ACTIFLO to pass a no-feasible-alternatives analysis, the EPA made clear that it “plans to follow [the 2005 policy] in reviewing State-issued permits,” and “it will insist State and local authorities comply with [the 2005 policy] in settling the terms and conditions of permits issued to petitioners.” See Appalachian Power Co., 208 F.3d at 1022. Just as it did in Appalachian Power, the EPA dissembles by describing the contested policy as subject to change. See id. at 1022-23. Yet, all regulations are susceptible to alteration. Hedging a concrete application of a policy within a disclaimer about hypothetical future contingencies does not insulate regulated entities from the binding nature of the obligations and similarly cannot serve to innoculate the agency from judicial review. Accordingly, we hold that the June 2011 and September 2011 letters were promulgations for the purposes of CWA section 509(b)(1)(E). 2. “[A]ny effluent limitation or other limitation” The CWA defines effluent limitations as “any restriction established by a State or the [EPA] on quantities, rates, and concentrations of chemical, physical, biological, and other constituents which are discharged from point sources into navigable waters.” 33 U.S.C. § 1362(11). *866 The Supreme Court has referred to effluent limitations as “direct restrictions on discharges.” EPA v. California ex rel. State Water Res. Control Bd., 426 U.S. 200, 204, 96 S.Ct. 2022, 48 L.Ed.2d 578 (1976). Other circuits have held that the expansiveness of the phrase “any restriction” encompasses both numerical and non-numerical effluent limitations. See, e.g., Waterkeeper Alliance, Inc. v. EPA, 399 F.3d 486, 502 (2d Cir.2005) (“[W]e believe that the terms of the nutrient management plans constitute effluent limitations”); NRDC v. EPA, 656 F.2d 768, 775 (D.C.Cir.1981) (finding an effluent limitation where, “[a]s a practical matter,” agency action “restricts] the discharge of sewage by limiting the availability of a variance to a class of applicants”). The phrase “other limitation” leaves much to the imagination. The Fourth Circuit explained that it “constructs] that term as a restriction on the untrammeled discretion of the industry ... [as it existed] prior to the passage of the [CWA].” Va. Elec. & Power Co. (VEPCO) v. Costle, 566 F.2d 446, 450 (4th Cir.1977). VEPCO found jurisdiction under section 509(b)(1)(E) because although the challenged regulations involved “structures,” rather than “discharges of pollutants into the water,” and therefore were not “effluent limitations,” they were nonetheless “other limitations” because they “referred] to information that must be considered in determining the type of intake structures that individual point sources may employ.” Id. at 449-50. Many of our sister circuits have adopted the VEPCO approach. See, e.g., Friends of the Everglades v. EPA 699 F.3d 1280, 1287 (11th Cir.2012) (finding no jurisdiction under section 509(b)(1)(E) because challenged rule did the opposite of restricting industry discretion, by “freeing] the industry from the constraints of the permit process”); Nw. Envtl. Advocates v. EPA 537 F.3d 1006, 1015-16 (9th Cir.2008) (finding no jurisdiction under section 509(b)(1)(E) because the challenged regulations created “categorical and permanent exemptions” from “any limit imposed by” CWA permit requirements); NRDC, 673 F.2d at 402, 405 (finding jurisdiction under section 509(b)(1)(E) to review “a complex set of procedures for issuing or denying NPDES permits” that restricted industry discretion). We agree that an agency action is a “limitation” within the meaning of section 509(b)(1)(E) if entities subject to the CWA’s permit requirements face new restrictions on their discretion with respect to discharges or discharge-related processes. Applying this definition, we find that the contested letters involve “effluent or other limitations.” The EPA’s position that bacteria mixing zones in waters “designated for primary contact recreation ... should not be permitted” is a restriction that directly affects the concentration of discharge from a point source and therefore is an effluent limitation. See Am. Iron & Steel Inst. v. EPA 115 F.3d 979, 986 (D.C.Cir.1997) (per curiam) (finding jurisdiction under CWA section 509(b)(1)(E) to review “the prohibition in Guidance Procedure 3.C against using mixing zones for new and existing BCC discharges”). The rule regarding the use of blending is an “other limitation” because, as in VEPCO, it restricts the discretion of municipal sewer treatment plants in structuring their facilities. As a result, both requirements for direct appellate review are satisfied here. 15 *867 B. Ripeness The judicially created doctrine of ripeness “flows from both the Article III ‘cases’ and ‘controversies’ limitations and also from prudential considerations for refusing to exercise jurisdiction.” Neb. Pub. Power Dist. v. MidAm. Energy Co., 234 F.3d 1032, 1037 (8th Cir.2000) (citing Reno v. Catholic Soc. Servs., Inc., 509 U.S. 43, 57 n. 18, 113 S.Ct. 2485, 125 L.Ed.2d 38 (1993)). “ ‘Ripeness is peculiarly a question of timing’ and is governed by the situation at the time of review, rather than the situation at the time of the events under review.” Id. at 1039 (quoting Anderson v. Green, 513 U.S. 557, 559, 115 S.Ct. 1059, 130 L.Ed.2d 1050 (1995) (per curiam)). A party seeking review must show both “the fitness of the issues for judicial decision and the hardship to the parties of withholding court consideration.” Pub. Water Supply Dist. No. 10 of Cass Cnty. v. City of Peculiar, 345 F.3d 570, 572-73 (8th Cir.2003) (quoting Abbott Labs. v. Gardner, 387 U.S. 136, 149, 87 S.Ct. 1507, 18 L.Ed.2d 681 (1967)). Both of these factors are weighed on a sliding scale, but each must be satisfied “to at least a minimal degree.” Neb. Pub. Power Dist., 234 F.3d at 1039. Fitness rests primarily on whether a case would “benefit from further factual development,” and therefore cases presenting purely legal questions are more likely to be fit for judicial review. Pub. Water Supply, 345 F.3d at 573. The hardship factor looks to the harm parties would suffer, both financially and as a result of uncertainty-induced behavior modification in the absence of judicial review. Neb. Pub. Power Dist., 234 F.3d at 1038. We do not require parties to operate beneath the sword of Damocles until the threatened harm actually befalls them, but the injury must be “certainly impending.” Pub. Water Supply, 345 F.3d at 573 (quoting Paraquad, Inc. v. St. Louis Hous. Auth., 259 F.3d 956, 958-59 (8th Cir.2001)). “The immediacy and the size of the threats ened harm” will also affect the interplay of these factors. Neb. Pub. Power Dist., 234 F.3d at 1038. This case hinges upon whether the EPA’s letters constitute legislative rules. We agree with our colleagues who have commented that “whether [a] Guidance Document is a legislative rule is largely a legal, not a factual, question, turning ... primarily upon the text of the Document.” Gen. Elec. Co. v. EPA, 290 F.3d at 380; see also Warder v. Shalala, 149 F.3d 73, 79 (1st Cir.1998); Chief Probation Officers of Cal. v. Shalala, 118 F.3d 1327, 1330 (9th Cir.1997). As primarily legal questions, such challenges tend to present questions fit for judicial review. On the other hand, postponing a procedural challenge to an agency guidance document may be appropriate where further factual development regarding the agency’s application of the document would aid our decision. Nat’l Park Hospitality Ass’n v. Dep’t of Interior, 538 U.S. 803, 812, 123 S.Ct. 2026, 155 L.Ed.2d 1017 (2003). This is so because the purpose of the ripeness doctrine is to prevent courts “from entangling themselves in abstract disagreements over administrative policies.” Abbott Labs., 387 U.S. at 148, 87 S.Ct. 1507. In this case, we are not wading into the abstract because the disagreements before us are quite concrete. Nothing about the proclamation that “the EPA’s position, as stated in the [King] memorandum, is that [bacteria mixing *868 zones in primary contact recreation waters] should not be permitted” indicates that the EPA’s posture will vary based on each applicant’s specific factual circumstances. Similarly, when asked if the use of “peak flow treatment processes” such as ACTIFLO would be subject to a “no feasible alternatives” demonstration, the EPA responded “Yes.” 16 The question is whether the statements are simply reminders of preexisting regulatory requirements or whether they create new regulatory obligations. Because such inquiries do not implicate contingent factual circumstances, this controversy is ripe for our review. See CropLife Am. v. EPA, 329 F.3d 876, 884 (D.C.Cir.2003) (finding that petitioners presented a “purely legal question” that was ripe for review where “the EPA directive states unequivocally that the agency will not consider any third-party human studies”). The second ripeness factor, hardship to parties, is also present. Although the EPA portrays the harm as lurking, if at all, on the distant horizon, the threatened harm is more immediate, and it is certainly not speculative. League members must either immediately alter their behavior or play an expensive game of Russian roulette with taxpayer money, investing significant resources in designing and utilizing processes that — if these letters are in effect new legislative rules— were viable before the publication of the letters but will be rejected when the letters are applied as written. See Neb. Pub. Power Dist., 234 F.3d at 1039 (“Delayed judicial resolution would only increase the parties’ uncertainty, and would require [petitioners] to gamble millions of dollars on an uncertain legal foundation.”). Postponing our review until the EPA has denied a permit application in accord with the letters renders a hardship on municipal water authorities, who already would have invested irretrievable funds into their applications. Cf. Toilet Goods Ass’n, Inc. v. Gardner, 387 U.S. 158, 164, 87 S.Ct. 1520, 18 L.Ed.2d 697 (1967) (finding a challenged agency action not ripe for review where “no irremediable adverse consequences [would] flow from requiring a later challenge to this regulation”). Therefore, we find that denying judicial review would be a hardship to the parties and that this case evinces the requisite degree of ripeness. See Abbott Labs., 387 U.S. at 153, 87 S.Ct. 1507 (“Where the legal issue presented is fit for judicial resolution, and where a regulation requires an immediate and significant change in the plaintiffs’ conduct of their affairs with serious penalties attached to noncompliance, access to the courts under the Administrative Procedure Act ... must be permitted, absent a statutory bar or some other unusual circumstance .... ”); Pac. Gas & Elec. Co. v. State Energy Res. Conservation & Dev. Comm’n, 461 U.S. 190, 201-02, 103 S.Ct. 1713, 75 L.Ed.2d 752 (1983) (finding a challenge to an as-yet unimplemented statute ripe because “requir[ing] the industry to proceed without knowing whether the moratorium is valid would impose a palpable and considerable hardship”); see also *869 Sackett v. EPA, 566 U.S.-, 132 S.Ct. 1367, 1374, 182 L.Ed.2d 367 (2012) (“[T]here is no reason to think that the Clean Water Act was uniquely designed to enable the strong-arming of regulated parties into ‘voluntary compliance’ without the opportunity for judicial review....”). C. Article III Standing If a litigant lacks Article III standing to bring his claim, then we have no subject matter jurisdiction over the suit. Miller v. Redwood Toxicology Lab., Inc., 688 F.3d 928, 934 (8th Cir.2012). “To show standing under Article III of the U.S. Constitution, a plaintiff must demonstrate (1) injury in fact, (2) a causal connection between that injury and the challenged conduct, and (3) the likelihood that a favorable decision by the court will redress the alleged injury.” Young Am. Corp. v. Affiliated Computer Servs. (ACS), Inc., 424 F.3d 840, 843 (8th Cir.2005) (citing Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-61, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992)). Because the League, rather than an individual permit applicant, is filing suit, it also must prove associational standing. “An association has standing to bring suit on behalf of its members when its members would otherwise have standing to sue in their own right, the interests at stake are germane to the organization’s purpose, and neither the claim asserted nor the relief requested requires the participation of individual members in the lawsuit.” Friends of the Earth, Inc. v. Laidlaw Envtl. Servs., Inc., 528 U.S. 167, 181, 120 S.Ct. 693, 145 L.Ed.2d 610 (2000). The League need not establish that all of its members would have standing to sue individually so long as it can show that “any one of them” would have standing. See Warth v. Seldin, 422 U.S. 490, 511, 95 S.Ct. 2197, 45 L.Ed.2d 343 (1975). The EPA concedes that the League meets the second and third elements of the associational standing test, and we agree. The only remaining element is whether any individual member would have standing to sue in its own right, which requires any League member to satisfy the three components that encompass the “irreducible constitutional minimum of standing.” See Am. Library Ass’n v. FCC, 406 F.3d 689, 696 (D.C.Cir.2005) (quoting Lujan, 504 U.S. at 560, 112 S.Ct. 2130). “[Standing is to be determined as of the commencement of the suit.” Lujan, 504 U.S. at 570 n. 5, 112 S.Ct. 2130. The party seeking judicial review bears the burden of persuasion and must support each element “with the manner and degree of evidence required at the successive stages of litigation.” Id. at 561, 112 S.Ct. 2130. Therefore, at the pleading stage a petitioner can move forward with “general factual allegations of injury,” whereas to survive a summary judgment motion, he “must set forth by affidavit or other evidence specific facts.” City of Clarkson Valley v. Mineta, 495 F.3d 567, 569 (8th Cir.2007) (quoting Lujan, 504 U.S. at 561, 112 S.Ct. 2130). The Supreme Court has not addressed “the manner and degree of evidence required” when a petitioner is seeking appellate review of an administrative action, nor has this circuit addressed the matter. The District of Columbia Circuit has equated such a petition with a motion for summary judgment, in that both request a final judgment on the merits. Sierra Club v. EPA, 292 F.3d 895, 899 (D.C.Cir.2002). Accordingly, parties seeking direct appellate review of an agency action must prove each element of standing as if they were moving for summary judgment in a district court. Id. Our colleagues on the Seventh Circuit have also taken this approach. See Citizens Against Ruining the Env’t v. EPA, 535 F.3d 670, 675 (7th Cir.2008). This reasoning is sound; because parties in the League’s position seek the type of relief *870 available on a motion for summary judgment, they correspondingly should bear the responsibility of meeting the same burden of production, namely “specific facts” supported by “affidavit or other evidence.” See Lujan, 504 U.S. at 561, 112 S.Ct. 2130. The EPA raises a factual challenge to our subject matter jurisdiction by attacking the facts asserted by the League with respect to standing, and therefore the League must establish standing “without the benefit of any inferences in [its] favor.” Defenders of Wildlife, Friends of Animals & Their Env’t v. Lujan, 911 F.2d 117, 120 (8th Cir.1990), rev’d on other grounds, 504 U.S. 555, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992). Parties seeking to litigate in federal court “have the burden of establishing jurisdiction,” including standing, “by a preponderance of the evidence.” Yeldell v. Tutt, 913 F.2d 533, 537 (8th Cir.1990). But see Sierra Club, 292 F.3d at 899 (imposing a burden of proof to establish elements of standing to a “substantial probability” (quoting Am. Petroleum Inst. v. EPA 216 F.3d 50, 63 (D.C.Cir.2000))). The League seeks to assert both a procedural and a substantive challenge to the letters. We address separately its standing to make each claim. See Int’l Bhd. of Teamsters v. Pena, 17 F.3d 1478, 1483-84 (D.C.Cir.1994). With respect to the substantive challenges, as the foregoing discussion regarding hardship has indicated, the League members’ affidavits evince the type of “concrete” and “actual or imminent” harm necessary to establish an injury in fact. See Thomas v. Anchorage Equal Rights Comm’n, 220 F.3d 1134, 1138-39 (9th Cir.1999) (en banc) (“[I]n many cases, ripeness coincides squarely with standing’s injury in fact prong.”). At least some members are currently operating under permits that allow them to utilize blending and bacteria mixing zones in circumstances inconsistent with the EPA letters, which they must imminently rectify. Cf. CropLife Am., 329 F.3d at 884 (“The disputed directive concretely injures petitioners, because it unambiguously precludes the agency’s consideration of ... studies that petitioners previously have been permitted to use to verify the safety of their products.”). Moving into compliance will be costly. The League has therefore articulated an injury in fact. See City of Waukesha v. EPA 320 F.3d 228, 234 (D.C.Cir.2003) (per curiam) (“The administrative record shows that the City of Waukesha would face substantial costs if it was required to comply with the ... regulations. EPA has not disputed that record evidence. This is sufficient for injury-in-fact.”). Causation for standing purposes requires that the harm asserted be “fairly traceable to the challenged action of the defendant.” Braden v. Wal-Mart Stores, Inc., 588 F.3d 585, 591 (8th Cir.2009) (quoting Lujan, 504 U.S. at 560, 112 S.Ct. 2130). The EPA disputes causation because it argues that the letters are not binding. Because we have ruled otherwise, we find that the League has established causation. Finally, the League has shown that it is “ ‘likely,’ as opposed to merely ‘speculative,’ that the injury will be ‘redressed by a favorable decision.’ ” Lujan, 504 U.S. at 561, 112 S.Ct. 2130 (quoting Simon v. E. Ky. Welfare Rights Org., 426 U.S. 26, 38, 43, 96 S.Ct. 1917, 48 L.Ed.2d 450 (1976)). If the rules were vacated as substantively unlawful, it is indeed likely that the members’ injuries would be redressed. With respect to the procedural challenge, namely that the EPA dodged the APA’s notice and comment procedures and de facto implemented new legislative rules regulating members’ activities under the CWA, the violation of a procedural right can constitute an injury in fact “so long as the procedures in question are *871 designed to protect some threatened concrete interest of [the petitioner] that is the ultimate basis of his standing.” Lujan, 504 U.S. at 573 n. 8, 112 S.Ct. 2130; see also Sierra Club v. Glickman, 156 F.3d 606, 616 (5th Cir.1998). The League’s members have a concrete interest not only in being able to meet their regulatory responsibilities but in avoiding regulatory obligations above and beyond those that can be statutorily imposed upon them. Notice and comment procedures for EPA rulemaking under the CWA were undoubtedly designed to protect the concrete interests of such regulated entities by ensuring that they are treated with fairness and transparency after due consideration and industry participation. See, e.g., Chrysler Corp. v. Brown, 441 U.S. 281, 316, 99 S.Ct. 1705, 60 L.Ed.2d 208 (1979) (“In enacting the APA, Congress made a judgment that notions of fairness and informed administrative decisionmaking require that agency decisions be made only after affording interested persons notice and an opportunity to comment.”). Thus, the League has established an injury in fact related to the EPA’s purported procedural deficiencies. Causation and redressability, and therefore standing to assert this procedural challenge, follow from these conclusions. Where a challenger is the subject of agency action, “there is ordinarily little question that the action ... has caused him injury, and that a judgment preventing ... the action will redress it.” Lujan, 504 U.S. at 561-62, 112 S.Ct. 2130. This is particularly true for individuals asserting violations of procedural rights. Id. at 572 n. 7, 112 S.Ct. 2130 (“The person who has been accorded a procedural right to protect his concrete interests can assert that right without meeting all the normal standards for redressability and immediacy.”). If a petitioner “is vested with a procedural right, that litigant has standing if there is some possibility that the requested relief will prompt the injury-eausing party to reconsider the decision that allegedly harmed the litigant.” Massachusetts v. EPA, 549 U.S. 497, 518, 127 S.Ct. 1438, 167 L.Ed.2d 248 (2007); see also Sierra Club v. EPA, 699 F.3d 530, 533 (D.C.Cir.2012) (“Having shown its members’ redressable concrete interest, [a petitioner association] can assert violation of the APA’s notice-and-comment requirements, as those procedures are plainly designed to protect the sort of interest alleged. As to such requirements, [the petitioner association] enjoys some slack in showing a causal relation between its members’ injury and the legal violation claimed.”). Correspondingly, redressability in this context does not require petitioners to show that the agency would alter its rules upon following the proper procedures. Sugar Cane Growers Coop. of Fla. v. Veneman, 289 F.3d 89, 95 (D.C.Cir.2002) (“If a party claiming the deprivation of a right to notice-and-comment rulemaking under the APA had to show that its comment would have altered the agency’s rule, section 553 would be a dead letter.”); see also Minard Run Oil Co. v. U.S. Forest Serv., 670 F.3d 236, 247 n. 4 (3d Cir.2011) (“Even if the [U.S. Forest Service is correct on the merits], the Agreement nevertheless establishes — in violation of appellees’ notice and comment rights — a new substantive rule.... This suffices for standing purposes.”); Pye v. United States, 269 F.3d 459, 471 & n. 7 (4th Cir.2001). The League’s remaining burden as to standing is met because “there is some possibility that the requested relief,” namely remanding to the EPA for application of notice and comment procedures, would “prompt the [EPA] to reconsider the decision that allegedly harmed” League members. See Massachusetts, 549 U.S. at 518, 127 S.Ct. 1438. We conclude that the League has standing to assert its claims. Having resolved *872 all jurisdictional questions, we now turn to the merits of the League’s petition for review. III. Merits of Procedural Challenge A. Standard of Review The parties disagree over the appropriate standard of review to be applied where, as here, an appellate court reviews challenges to agency procedural compliance under § 706(2)(D). The League urges us to follow the Ninth Circuit, which “reviews de novo the agency’s decision not to follow the APA’s notice and comment procedures because complying with the notice and comment provisions when required by the APA ‘is not a matter of agency choice.’ ” Reno-Sparks Indian Colony v. EPA, 336 F.3d 899, 909 n. 11 (9th Cir.2003) (quoting Sequoia Orange Co. v. Yeutter, 973 F.2d 752, 757 n. 4 (9th Cir.1992)). The EPA argues its characterization of the letters is entitled to a deferential abuse of discretion review. Our pri- or decisions have not clearly announced a standard of review, other than to note that the agency’s characterization of its rule as legislative or interpretative, “while not dis-positive, is entitled to deference.” Drake v. Honeywell, Inc., 797 F.2d 603, 607 (8th Cir.1986). But see United States v. Hacker, 565 F.3d 522, 524 (8th Cir.2009) (stating in dicta that challenges to procedural compliance under the APA present “a question of law, which we review de novo”), abrogated on other grounds by Bond v. United States, 564 U.S.-, 131 S.Ct. 2355, 180 L.Ed.2d 269 (2011); South Dakota v. Ubbelohde, 330 F.3d 1014, 1028 (8th Cir.2003) (“Where a policy statement purports to create substantive requirements, it can be a legislative rule regardless of the agency’s characterization.”). We agree with our colleagues on the Ninth Circuit that much of the rationale for granting deference to administrative decisions is simply not applicable where the topic of our review — compliance with APA procedural requirements — is not a matter that Congress has committed to the agency’s discretion. In other words, whether and when an agency must follow the law is not an area uniquely falling within its own expertise, and thus the agency’s decision is less deserving of deference. Cf. Campanale & Sons, Inc. v. Evans, 311 F.3d 109, 120 n. 14 (1st Cir.2002) (“We are unaware of any line of cases that allows an agency to make a binding determination that it has complied with specific requirements of the law.... As to the so-called ‘specialized experience’ of the agency, it would appear that it is the courts that qualify for such a title on an issue of legislative interpretation.”). Furthermore, because the categorization of an agency’s action as a legislative or interpretative rule is largely a question of law, a de novo standard of review is consistent with the standard of review we generally apply to questions of law in similar contexts. See Qwest Corp. v. Minn. Pub. Utils. Comm’n, 427 F.3d 1061, 1064 (8th Cir.2005). At least two circuits in addition to the Ninth Circuit have expressly announced a de novo standard of review when distinguishing between legislative rules and other types of agency action. See Meister v. Dep’t of Agric., 623 F.3d 363, 370 (6th Cir.2010); Warder, 149 F.3d at 79. We adopt a de novo standard of review as well. This is not to say that the agency’s label is to be ignored. As discussed above, an agency’s characterization of its rule is a relevant component of our review and is a factor entitled to some deference. Our posture in this regard mirrors similar comments made by other courts of appeals. See Gen. Motors Corp. v. Ruckelshaus, 742 F.2d 1561, 1565 (D.C.Cir.1984) (“[T]he agency’s own label, while relevant, is not dispositive.”) (en *873 banc); accord Prof'ls & Patients for Customized Care v. Shalala, 56 F.3d 592, 595 (5th Cir.1995); La Casa Del Convaleciente v. Sullivan, 965 F.2d 1175, 1178 (1st Cir.1992); Metro. Sch. Dist. of Wayne v. Davila, 969 F.2d 485, 489 (7th Cir.1992); Friedrich v. HHS, 894 F.2d 829, 884-35 (6th Cir.1990); Lewis-Mota v. Sec’y of Labor, 469 F.2d 478, 481-82 (2d Cir.1972). The critical distinction between legislative and interpretative rules is that, whereas interpretative rules “simply state what the administrative agency thinks the statute means, and only ‘remind’ affected parties of existing duties,” a legislative rule “imposes new rights or duties.” Nw. Nat’l Bank v. U.S. Dep’t of the Treasury, 917 F.2d 1111, 1117 (8th Cir.1990) (quoting Jerri’s Ceramic Arts, Inc. v. Consumer Prod. Safety Comm’n, 874 F.2d 205, 207 (4th Cir.1989)). When an agency creates a new “legal norm based on the agency’s own authority ” to engage in supplementary lawmaking, as delegated from Congress, the agency creates a legislative rule. Syncor Int’l Corp. v. Shalala, 127 F.3d 90, 95 (D.C.Cir.1997). Expanding the footprint of a regulation by imposing new requirements, rather than simply interpreting the legal norms Congress or the agency itself has previously created, is the hallmark of legislative rules. See Ubbelohde, 330 F.3d at 1028; Martin v. Gerlinski, 133 F.3d 1076, 1079 (8th Cir.1998); Syncor Int’l Corp., 127 F.3d at 94-95. It follows from this distinction that interpretative rules do not have “the force of law.” 17 Shalala v. St. Paul-Ramsey Med. Ctr., 50 F.3d 522, 527 n. 4 (8th Cir.1995). Whether or not a binding pronouncement is in effect a legislative rule that should have been subjected to notice and comment procedures thus depends on whether it substantively amends or adds to, versus simply interpreting the contours of, a preexisting rule. See U.S. Telecom Ass’n v. FCC, 400 F.3d 29, 34-35 (D.C.Cir.2005). Identifying where a contested rule lies on the sometimes murky spectrum between legislative rules and interpretative rules can be a difficult task, but it is not just an exercise in hair-splitting formalism. As agencies expand on the often broad language of their enabling statutes by issuing layer upon layer of guidance documents and interpretive memoranda, formerly flexible strata may ossify into rule-like rigidity. An agency potentially can avoid judicial review through the tyranny of small decisions. Notice and comment procedures secure the values of government transparency and public participation, compelling us to agree with the suggestion that “[t]he APA’s notice and comment exemptions must be narrowly construed.” Prof'ls & Patients for Customized Care, 56 F.3d at 596 (quoting United States v. Picciotto, 875 F.2d 345, 347 (D.C.Cir.1989)); see also City of New York v. Permanent Mission of India to United Nations, 618 F.3d 172, 201 (2d Cir.2010). B. Bacteria Mixing Zones Since at least 1994, the EPA’s long-standing policy toward bacteria mixing zones has been that states should exercise their “discretion” — as set forth in 40 C.F.R. § 131.13 — to adopt a “definitive statement” in their water quality standards “on whether or not mixing- zones are *874 allowed.” Handbook Ch. 5.1, 5.1.1. States are authorized to consider mixing zones in determining the types of standards necessary to preserve water quality. 40 C.F.R. § 122.44(d)(1)(h). States do not enjoy complete discretion in creating a mixing zone policy because they operate within the shadow of EPA-crafted effluent limitations. The Handbook interprets certain instances of mixing zones as inconsistent with EPA regulations: states should not draft water quality standards that allow point source dischargers to utilize mixing zones in ways that “may endanger critical areas,” such as recreational areas, or pose “significant health risks.” Ch. 5.1. Notably, no preexisting regulation establishes that all bacteria mixing zones in recreational areas necessarily “may endanger critical areas” or create “significant health risks.” 18 In fact, under the Handbook, whether a mixing zone causes such a state of affairs was to be determined based on a “holistic approach.” Id. Yet, when now asked if a state “[m]ay ... approve a bacteria mixing zone for waters designated for body contact recreation,” the EPA flatly proclaims that such mixing zones “should not be permitted.” The June 2011 letter tells state permitting authorities that mixing zones in primary contact recreation areas are necessarily inconsistent with achieving the water quality levels required by federal regulations. The EPA eviscerates state discretion to incorporate mixing zones into their water quality standards with respect to this type of body of water. In effect, the EPA has created a new effluent limitation: state permitting authorities no longer have discretion to craft policies regarding bacteria mixing zones in primary .contact recreation areas. Instead, such mixing zones are governed by an effluent limitation that categorically forbids them. To be sure, in 1994 the EPA stated that as its “understanding of pollutant impacts on ecological systems evolves, cases could be identified where no mixing zone is appropriate.” Handbook Ch. 5.1.1. It seems that the EPA’s understanding of pollutant impacts has so evolved, and it has now identified an entire class of cases “where no mixing zone is appropriate.” However, the effect of the EPA applying its more developed understanding of pollutant impacts is to promulgate a new effluent limitation that state permitting authorities must follow. See Nat’l Family Planning & Reprod. Health Ass’n v. Sullivan, 979 F.2d 227, 235 (D.C.Cir.1992) (“Thus, a rule is legislative if it attempts ‘to supplement [a statute,] not simply to construe it.’ ”) (alteration in original). In short, the June 2011 letter creates a new legal norm for bacteria mixing zones based on the EPA’s authority to promulgate effluent limitations. The hallmark of an interpretative rule or policy statement is that they cannot be independently legally enforced. It is the underlying legislative rules that drive compliance, and thus when an agency applies a newly announced interpretative rule or policy statement, there must be some external legal basis supporting its implementation. See St. Paul-Ramsey Med. Ctr., 50 F.3d at 528 n. 4; Prof'ls & *875 Patients for Customized Care, 56 F.3d at 596. The EPA has not cited any preexisting effluent limitation or lawfully promulgated legislative rule that supplies the basis for the prohibition on bacteria mixing zones in primary contact recreation areas. This reinforces our conclusion that this new legal norm is a legislative rule and that the EPA violated the APA when it bypassed notice and comment procedures. Accordingly, we vacate the EPA’s new rule banning bacteria mixing zones in all waters designated for primary contact recreation as promulgated “without observance of procedure required by law.” 5 U.S.C. § 706(2)(D). C. Blending The EPA contends that the letters simply reflect an interpretation of the bypass rule, which it has been considering since 2005. See 70 Fed.Reg. at 76,-015 (describing the 2005 policy as “the Agency’s interpretation” of the bypass rule). To be sure, a legislative rule is not created simply because an agency “supplies crisper and more detailed lines than the authority being interpreted.” Am. Mining Cong. v. Mine Safety & Health Admin., 995 F.2d 1106, 1112 (D.C.Cir.1993). Nevertheless, the EPA’s new blending rule is a legislative rule because it is irreconcilable with both the secondary treatment rule and the bypass rule. See Nat’l Family Planning & Reprod. Health Ass’n, 979 F.2d at 235 (“If a second rule repudiates or is irreconcilable with [a prior legislative rule], the second rule must be an amendment of the first; and, of course, an amendment to a legislative rule must itself be legislative.” (alteration in original) (quoting Michael Asimow, Nonlegislative Rulemaking and Regulatory Reform, 1985 Duke L.J. 381, 396 (1985))). The September 2011 letter simply applies the 2005 draft Policy to the proposed use of ACTIFLO as if the 2005 draft were an existing obligation of regulated entities. However, the record indicates that prior to 2005, the EPA had not viewed the use of a process such as ACTIFLO as an inevitable trigger of a no-feasible-alternatives requirement. The 2005 draft Policy characterized itself as “significantly different” from the EPA’s 2003 proposal on blending. 70 Fed.Reg. at 76,014. The 2003 proposal, in turn, corresponds to what the record indicates is the reality on the ground: widespread use by POTWs of blending peak wet weather flows. The 2005 draft Policy acknowledges that blending previously had been “permitted at [POTWs] without consideration of the bypass regulation criteria.” 70 Fed.Reg. at 76,015. In a response to a 2002 Freedom of Information Act (“FOIA”) request, the EPA admitted to “the use of federal funds under the Construction Grants Program to build facilities that were designed to blend effluent from primary treatment processes with effluent from biological treatment processes during peak wet weather events.” 19 In a 2004 report to Congress, the EPA praised the use of blending processes like ACTIFLO to deal with peak wet weather flows with no reference to a no-feasible-alternatives requirement. Various Iowa municipal water authorities have averred that the Iowa Department of Natural Resources has approved permits— with no objection from the EPA and no imposition of a no-feasible-alternatives requirement — allowing cities to construct facilities utilizing non-biological peak flow secondary treatment processes. Municipalities chose to use ACTIFLO and analogous blending methods as an ex *876 ercise of their discretion under the bypass rule, see 53 Fed.Reg. at 40,609, and secondary treatment rule, see 48 Fed.Reg. at 52,259, to select the particular technologies they deemed best suited to achieving the applicable secondary treatment requirements. However, the September 2011 letter severely restricts the use of “ACTIFLO systems that do not include a biological component” because the EPA does not “consider! ] [them to be] secondary treatment units.” The effect of this letter is a new legislative rule mandating certain technologies as part of the secondary treatment phase. If a POTW designs a secondary treatment process that routes a portion of the incoming flow through a unit that uses non-biological technology disfavored by the EPA, then this will be viewed as a prohibited bypass, regardless of whether the end of pipe output ultimately meets the secondary treatment regulations. The EPA’s new blending rule further conflicts with the secondary treatment regulations because the EPA has made clear that effluent limitations apply at the end of the pipe unless it would be impractical to do so. 40 C.F.R. § 122.45(h). There is no indication that the secondary treatment regulations established situations in which it would be impractical to apply effluent limitations at the end of the pipe or otherwise altered the application of this default rule. See 40 C.F.R. § 133.100-102. But the blending rule applies effluent limitations within facilities’ secondary treatment processes. The September 2011 letter rejected the use of ACTIFLO because these units “do not provide treatment necessary to meet the minimum requirements provided in the secondary treatment regulations at 40 CFR 133.” If streams move around traditional biological secondary treatment processes and through a non-biological unit that “is itself a secondary treatment unit,” then the system would not need to meet the restrictive no-feasible-alternatives requirement. In other words, under the September 2011 blending rule, if POTWs separate incoming flows into different streams during the secondary treatment phase, the EPA will apply the effluent limitations of the secondary treatment regulations to each individual stream, rather than at the end of the pipe where the streams are recombined and discharged. Because the September 2011 letter had the effect of announcing a legislative rule with respect to blending peak wet weather flows, the EPA violated the APA’s procedural requirements by not using notice and comment procedures. We also vacate this new rule because it is “without observance of procedure required by law.” 5 U.S.C. § 706(2)(D). IV. Merits of substantive challenge Even if the EPA’s legislative rules had been promulgated through the proper procedural channels, the League argues they nonetheless should be “set aside ... [as] in excess of statutory jurisdiction, authority, ... or short of statutory right.” 5 U.S.C. § 706(2X0. This subsection of the APA authorizes courts to strike down as ultra vires agency rules promulgated without valid statutory authority. United States ex rel. O’Keefe v. McDonnell Douglas Corp., 132 F.3d 1252, 1257 (8th Cir.1998). The League urges us to find that the EPA exceeded its statutory authority under the CWA by prohibiting mixing zones outside the state water quality standard adoption process and by using the blending prohibition to dictate facility treatment design and apply effluent limitations internally, rather than at the end of the pipe. Appellate review under APA section 706(2)(C) proceeds under the familiar Chevron framework. See Clark v. U.S. Dep’t of Agric., 537 F.3d 934, 939 (8th Cir.2008). We first “conduct an independent review of the statute and of its legislative history.” Ark. AFL-CIO v. *877 FCC, 11 F.3d 1430, 1441 n. 9 (8th Cir.1993) (en banc). “Deference to the agency is appropriate only when a court finds the statute to be ambiguous.” Id.; see also Chevron, U.S.A, Inc. v. NRDC, 467 U.S. 837, 843 n. 9, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984) (“[T]he judiciary is the final authority on issues of statutory construction and must reject administrative constructions which are contrary to clear congressional intent.”). If confronted with an ambiguous statute, we look to whether the agency’s construction of the statute is reasonable. Ark. AFL-CIO, 11 F.3d at 1441. Agency rules will survive ultra vires allegations so long as we can “reasonably conclude that the grants of authority in the statutory provisions cited by the government contemplate the issuance.” O’Keefe, 132 F.3d at 1257. We find our circuit in the same position as the District of Columbia Circuit, which recently observed that its “case law provides little direction on whether, having determined to vacate on procedural grounds, we should nonetheless address substantive claims.” NRDC v. EPA, 643 F.3d 311, 321 (D.C.Cir.2011); cf. U.S. Steel Corp. v. EPA, 649 F.2d 572, 577 (8th Cir.1981). The decision implicates competing tensions, both compelling. If we choose to vacate solely on procedural grounds, regulated entities who have already spent considerable time crossing the hot shoals of regulatory uncertainty must continue to do so. On the other hand, should we move to the merits of whether the EPA’s legislative rules reflect an arbitrary and capricious interpretation of the CWA, we short-circuit the APA’s notice and comment procedures and preclude interested parties from participating in the agency’s analytic process. Cf. Smiley v. Citibank, N.A., 517 U.S. 735, 741, 116 S.Ct. 1730, 135 L.Ed.2d 25 (1996) (“[T]he notice-and-comment procedures of the Administrative Procedure Act [are] designed to assure due deliberation.”). In a recent case, the District of Columbia Circuit found the “interest in preserving the integrity of the notice and comment process” outweighed “concern[s] about delay” where the EPA’s rule was not “obviously precludefd]” by the relevant enabling act. See NRDC, 643 F.3d at 321. Here, too, we conclude that the EPA’s new mixing zone rule is not obviously precluded by the plain meaning of any applicable CWA provisions. Therefore, should the EPA wish to institute this rule, it may seek to do so using the appropriate procedures. However, the blending rule clearly exceeds the EPA’s statutory authority and little would be gained by postponing a decision on the merits. As discussed above, the September 2011 letter applies effluent limitations to a facility’s internal secondary treatment processes, rather than at the end of the pipe. The CWA permits the EPA to set “effluent limitations based upon secondary treatment.” 33 U.S.C. § 1311(b)(1)(B). But effluent limitations are restricted to regulations governing “discharges from point sources into navigable waters.” 33 U.S.C. § 1362(11). The EPA is authorized to administer more stringent “water quality related effluent limitations,” but the CWA is clear that the object of these limitations is still the “discharges of pollutants from a point source.” 33 U.S.C. § 1312(a). In turn, “discharge of pollutant” refers to the “addition of any pollutant to navigable waters.” § 1362(11). The EPA would like to apply effluent limitations to the discharge of flows from one internal treatment unit to another. We cannot reasonably conclude that it has the statutory authority to do so. See also Am. Iron & Steel Inst. v. EPA, 115 F.3d 979, 996 (D.C.Cir.1997) (“The statute is clear: The EPA may regulate the pollutant levels in a waste stream that is discharged directly into the navigable waters of the United States through a *878 ‘point source’; it is not authorized to regulate the pollutant levels in a facility’s internal waste stream.”). Therefore, insofar as the blending rule imposes secondary treatment regulations on flows within facilities, we vacate it as exceeding the EPA’s statutory authority. V. Conclusion For the foregoing reasons, we deny the EPA’s motion to dismiss and grant the League’s petition for review. We vacate both the mixing zone rule in the June 2011 letter and the blending rule in the September 2011 letter as procedurally invalid. Further, we vacate the blending rule as in excess of statutory authority insofar as it would impose the effluent limitations of the secondary treatment regulations internally, rather than at the point of discharge into navigable waters. We remand to the EPA for further consideration. 20 Footnotes [1] Our ability to make a final decision on jurisdiction is unaffected by the rulings of either *855 this administrative panel or the 2010 administrative panel. See In re Rodriquez, 258 F.3d 757, 758-59 (8th Cir.2001) (per curiam). [2] Some courts also use the phrase "interpretive” rules interchangeably with "interpretative” rules. [3] A “point source” is "any discernible, confined and discrete conveyance, including but not limited to any pipe, ditch, channel, tunnel, conduit, well, discrete fissure, container, rolling stock, concentrated animal feeding operation, or vessel or other floating craft, from which pollutants are or may be discharged.” 33 U.S.C. § 1362(14). This case involves municipal wastewater treatment facilities, which both parties agree are point sources. [4] Iowa is one of forty-six states approved to administer an NPDES program. EPA, State Program Status, http://cfpubl.epa.gov/npdes/ statestats.cfm (last visited Feb. 14, 2013). [5] Available at http://www.epa.gov/npdes/pubs/ pwm_2010.pdf. [6] POTWs are "any devices and systems used in the storage, treatment, recycling and reclamation of municipal sewage or industrial wastes of a liquid nature” that are "owned by a State or municipality.” 40 C.F.R. § 403.3(q). [7] "Water quality criteria are the threshold values against which ambient concentrations are compared to determine whether a water-body exceeds the water quality standard.... NPDES permits must establish limits on any *857 pollutant, where necessary to attain and maintain applicable water quality standards." 54 Fed.Reg. 23,868, 23,872 (June 2, 1989). [8] Biological-based systems use microorganisms to treat incoming flows. A facility can be designed to use non-biological treatment processes, such as chemical additives or physical filtration equipment, instead of or in conjunction with biological facilities. [9] ACTIFLO is a physical/chemical process that uses ballasted flocculation. "In ballasted *859 flocculation or sedimentation, a metal salt coagulant is added to the excess wet weather flows to aggregate suspended solids. Then, fine-grained sand, or ballast, is added along with a polymer. The polymer acts like glue which bonds the aggregated solids and sand. The process increases the particles’ size and mass which allows them to settle faster.” EPA, Report to Congress: Impacts and Control of CSOs and SSOs 2 (2004). [10] The APA does not create federal subject matter jurisdiction. Preferred Risk Mut. Ins. Co. v. United States, 86 F.3d 789, 792 (8th Cir.1996). Rather, a federal court has federal question jurisdiction under 28 U.S.C. § 1331 over challenges to federal agency action. Ochoa v. Holder, 604 F.3d 546, 549 (8th Cir. *861 2010); see also Reno v. Catholic Soc. Servs., Inc., 509 U.S. 43, 56, 113 S.Ct. 2485, 125 L.Ed.2d 38 (1993). [11] The League did not contend that the EPA's letters were “actions ... approving” effluent or other limitations, rather than promulgating them, and therefore we did not consider the matter. ■ . [12] The EPA argues that no federal court has jurisdiction over this claim because these letters are not “final agency actions.” Even if there were an implicit finality requirement applicable to "[ajgency actions made reviewable by statute," this would not affect federal jurisdiction; the APA's requirements are part of a party's cause of action and are not jurisdictional. Air Courier Conference v. Am. Postal Workers Union, 498 U.S. 517, 523 n. 3, 111 S.Ct. 913, 112 L.Ed.2d 1125 (1991) ("The judicial review provisions of the APA are not jurisdictional."); see also Ochoa, 604 F.3d at 549 (8th Cir.2010); Trudeau v. FTC, 456 F.3d 178, 183-84 (D.C.Cir.2006). In this case, analyzing whether an agency pronouncement is binding evokes considerations of finality. However, they arise not from the APA, but rather from the conditions placed on the CWA’s grant of direct appellate jurisdiction. The APA allows judicial review in two situations: "Agency action made reviewable by statute and final agency action for which there is no other adequate remedy in a court....” 5 U.S.C. § 704. The word “final” modifies the second use of "agency action,” but not the first. While some courts have interpreted the phrase "[ajgency action made reviewable by statute” as including an implied finality requirement, see, e.g., Appalachian Energy Grp. v. EPA, 33 F.3d 319, 322 (4th Cir.1994); Carter/Mondale Presidential Comm., Inc. v. Fed. Election Comm’n, 711 F.2d 279, 285 n. 9 (D.C.Cir.1983), we decline to conjure up a finality requirement for "[ajgency actions made reviewable by statute” where none is located in the text of the APA, particularly where the Supreme Court has implied that the two phrases incorporate distinct requirements, see Lujan v. Nat’l Wildlife Fed’n, 497 U.S. 871, 882, 110 S.Ct. 3177, 111 L.Ed.2d 695 (1990) ("When, as here, review is sought not pursuant to specific authorization in the substantive statute, but only under the general review provisions of the APA, the ‘agency action’ in question must be ‘final agency action.' ”); id. at 891, 110 S.Ct. 3177 (“Some statutes permit broad regulations to serve as the ‘agency action,' and thus to be the object of judicial review directly, even before the concrete effects normally required for APA review are felt.”); see also Yankton Sioux Tribe v. Podhradsky, 606 F.3d 994, 1012 (8th Cir.2010) ("[Tjhe 'cardinal canon' of statutory interpretation is 'that a legislature says in a statute what it means and means in a statute what it says there.’ ” (quoting Conn. Nat'l Bank v. Germain, 503 U.S. 249, 253-54, 112 S.Ct. 1146, 117 L.Ed.2d 391 (1992))). The CWA expressly makes specified agency actions reviewable, and our task therefore is to determine whether the asserted agency action falls within the statutory terms. [13] The League provided these affidavits in an unopposed appendix supplementing the EPA's administrative record. After oral argument, the League filed a motion to further supplement the record with additional affidavits from the Iowa and Kansas water permitting authorities. The EPA objects to the League's attempt to further supplement the record at this stage. The Supreme Court has explained that when applying the arbitrary and capricious standard of review under APA section 706(2)(A), "the focal point for judicial review should be the administrative record already in existence.” Camp v. Pitts, 411 U.S. 138, 142, 93 S.Ct. 1241, 36 L.Ed.2d 106 (1973). Therefore, if "there is a contemporaneous administrative record and no need for additional explanation of the agency decision,” we will permit supplementation of the administrative record only where there is a "strong showing of bad faith or improper behavior.” Newton Cnty. Wildlife Ass’n v. Rogers, 141 F.3d 803, 807 (8th Cir.1998) (quoting Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 420, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971)). The rationale for this rule is that judicial review of the reasonableness of an agency's actions should concentrate upon the evidence available to the agency when making its decision. See Robinette v. Comm’r, 439 F.3d 455, 459 (8th Cir.2006). But where, as here, rulemaking masquerading as explication is alleged, the informality of the agency's decisionmaking process makes the possibility of a sparse "contemporaneous administrative record” more likely. While we question whether the Camp standard would necessarily apply to such challenges under APA section 706(2)(D), we need not decide the matter because we reached our conclusions without resort to the League’s proposed supplementary materials. Therefore, we deny the League's motion to supplement the record. [14] League Question: "Is the permitted use of ACTIFLO or other similar peak flow treatment processes to augment biological treatment subject to a 'no feasible alternatives' demonstration?” EPA Response: "Yes.” The EPA insists that this challenge is time-barred because the proper time to raise the challenge was in 2005. We find this contention unpersuasive because prior to the September 2011 letter, the EPA never indicated that the 2005 policy became final. For example, the June 1, 2010 Federal Register notice explained that the EPA was continuing to "solicit[] input from the general public concerning the impact of the proposed rule.” 75 Fed.Reg. 30,-395, 30,401 (June 1, 2010). Even the June 2011 letter explained that the agency was “continu[ing] to consider whether the 2005 Policy should be finalized or incorporated into the EPA’s other potential wet weather rulemaking effort announced June 1, 2010 in the Federal Register.” In contrast, the September 2011 letter simply applies the 2005 Policy to the regulated entities as if it had already been finalized. The EPA’s approach to the period for seeking appellate review would eviscerate the direct appellate review provisions of the CWA by enabling an agency to announce consideration of a proposal and then wait 121 days before treating the proposal as binding. Cf. CropLife Am. v. EPA, 329 F.3d 876, 884 (D.C.Cir.2003) (refusing to find that the petitioners' claim was time-barred "because the new rule clearly represents the first time that the agency has adopted an unequivocal, wholesale ban”). The time to seek direct appellate review begins to run not when the agency first floats its proposal to the public, but rather when the agency promulgates that announcement — in other words, when they make its substance binding. [15] The EPA insists that as a result of finding its conduct here reviewable, there will be a chilling effect on the informal channels of communication between agencies and regulated entities. We acknowledge the great value in such modes of communication and encourage agencies to continue to utilize them. However, when agencies veer from merely *867 advisory statements or interpretations into binding proclamations, they become susceptible to judicial review. [16] The September 2011 letter acknowledged that if ACTIFLO independently met secondary treatment requirements, then flows moving through ACTIFLO units instead of the facility’s biological secondary treatment units would not be considered a bypass. However, the letter also stated that ACTIFLO failed to meet these requirements and that the EPA would "continue to explore in what circumstances use of [ACTIFLO-type] technologies is consistent with a determination that there are 'no feasible alternatives.’ ” During oral argument, counsel for EPA informed us that the use of newer, modified versions of ACTIFLO units "may well satisfy the secondary treatment regulations.” This type of belated backpedaling is insufficient to render these challenges so intertwined with hypothetical future conditions that they are unripe for review. [17] The EPA insists the letters are neither legislative nor interpretative rules but rather constitute policy statements. Policy statements are not binding, either as a legal or practical matter. See NRDC v. EPA, 643 F.3d 311, 321 (D.C.Cir.2011) (“To begin with, because the Guidance binds EPA regional directors, it cannot, as EPA claims, be considered a mere statement of policy; it is a rule.”). Because we have determined that the letters evince binding rules regarding bacteria mixing zones and blending, neither can be characterized as a policy statement. [18] The EPA’s own guidance also belies any interpretation of its preexisting legislative rules as categorically prohibiting the use of mixing zones in waters designated for primary recreational contact. See EPA, Guidance: Coordinating CSO Long-Term Planning with Water Quality Standards Reviews 5 (2001) (describing how states may alter their water quality standards to apply bacteria water quality criteria "at the beach or at the point of contact rather than at the end-of-pipe or at the edge of the mixing zones”); EPA, Guidance on Application of State Mixing Zone Policies in EPA-Issued NPDES Permits 1 (1996) ("Thus, individual state law and policy determine whether or not a mixing zone is permitted.”). [19] FOIA request submitted by John Hall to the EPA on October 25, 2001; response dated April 5, 2002, No. HQ-RIN-00459-02. [20] The League also requested attorneys' fees under CWA section 509(b)(3), which authorizes courts, "whenever ... appropriate,” to award litigation costs to any "prevailing or substantially prevailing party.” To be a prevailing party entitled to attorneys' fees, a plaintiff must achieve at least some relief on the merits that effectuates a "material alteration of the legal relationship of the parties.” Buckhannon Bd. & Care Home, Inc. v. West Va. Dep’t of Health & Human Res., 532 U.S. 598, 604, 121 S.Ct. 1835, 149 L.Ed.2d 855 (2001) (quoting Tex. State Teachers Ass’n v. Garland Indep. Sch. Dist., 489 U.S. 782, 792-93, 109 S.Ct. 1486, 103 L.Ed.2d 866 (1989)); see also Sierra Club v. City of Little Rock, 351 F.3d 840, 845 (8th Cir.2003) (applying Buck-hannon to a claim for attorneys' fees under the CWA). The League is clearly a prevailing party, even on the basis of its procedural challenge alone. See Chem. Mfrs. Ass’n v. EPA, 885 F.2d 1276, 1279 (5th Cir.1989) (describing "substantive significance” of a remand on procedural grounds). An award of litigation costs under section 509(b)(3) must also be "appropriate.” Statutory provisions authorizing an award of litigation costs often serve to incentivize the achievement of statutory objectives, and therefore "an award is usually 'appropriate' when a party has advanced the goals of the statute invoked in the litigation.” Id.; see also Saint John's Organic Farm v. Gem Cnty. Mosquito Abatement Dist., 574 F.3d 1054, 1061 (9th Cir.2009); NRDC v. EPA, 512 F.2d 1351, 1357 (D.C.Cir.1975). The CWA's goals, involve the restoration and maintenance of the "chemical, physical, and biological integrity of the Nation’s waters.” 33 U.S.C. § 1251(a). The League, however, was largely vindicating its own rights, rather than the purposes of the CWA, and it has neglected to brief us on why an award of attorneys' fees would otherwise be "appropriate.” Therefore, we decline to award litigation costs under CWA section 509(b)(3).