Opinion by Judge CALLAHAN; Dissent by Judge IKUTA.

OPINION

These thirteen petitions for review challenge the Department of Energy’s (“DOE”) implementation of the Energy Policy Act of 2005 (“EPAct”), which added a new section 216 to the Federal Power Act (“FPA”), codified as 16 U.S.C. § 824p (sometimes referred to as “§ 216”). Petitioners offer three distinct challenges to DOE’s actions: (1) DOE failed to consult with the affected States in undertaking the Congestion Study as required by § 824p(a)(l); (2) DOE failed to properly consider the potential environmental consequences of its designation of national interest electric transmission corridors (“NIETCs”); and (3) DOE’s actual designations of the Mid-Atlantic Area National Corridor and the Southwest Area National Corridor are arbitrary, capricious, and not supported by the evidence. We determine that DOE failed to properly consult with the affected States in conducting the Congestion Study and failed to undertake any environmental study for its NIETC Designation as required by the National Environmental Policy Act (“NEPA”), 42 U.S.C. § 4332 (C). We also determine that these failings were not harmless errors. Accordingly, we vacate the Congestion Study and NIETC designation and remand the cases to the DOE for further proceedings. Because we vacate the NIETC designation, we do not consider the merits of petitioners’ challenges to the specific na *1080 tional corridors other than as necessary to determine that DOE’s failures to consult and to undertake an environmental study were not harmless errors.

I. BACKGROUND

A. The Critical Statute, 16 U.S.C. § 824

In response to a number of electrical brown-outs and black-outs, Congress passed the EPAct, Pub.L. No. 109-85, 119 Stat. 594 (2005). The EPAct added a new section 216 to the FPA. The first provisions of the section read:

16 U.S.C. § 824p (emphasis added).

The designation of an area as a “national interest electric transmission corridor” (variously referred to as a “National Corridor,” “NIET Corridor” or “NIETC”) makes available a fast-track approval process to utilities seeking permits for transmission lines within the corridor. See 16 U.S.C. §§ 824p(b)-(h). In particular, the Federal Energy Regulatory Commission (“FERC”) is empowered to grant a permit for a transmission line within the corridor if, among other conditions, a state agency fails to approve the permit application within a year. 16 U.S.C. § 824p(b). 1 In addition, the EPAct, in providing for the issuance of a permit, gives the applicant the right to acquire rights-of-way through eminent domain. 16 U.S.C. § 824p(e).

Moreover, 16 U.S.C. § 824p(j)(l) states, “[E]xcept as specifically provided, nothing in this section affects any requirement of an environmental law of the United States, including the National Environmental Policy Act of 1969 ( 42 U.S.C. §§ 4321 et seq.).”

B. The Congestion Study

Following the enactment of § 216, DOE gave presentations at a number of conferences regarding the new law. Its first request for comments or assistance from others was a February 2, 2006 “Notice of *1081 inquiry requesting comment and providing notice of a technical conference” (the “February 2 Notice”). 71 Fed.Reg. 5660-64 (Feb. 2, 2006). The notice sought “comment and information from the public concerning its plans for an electricity transmission congestion study and possible designation of [NIET Corridors].” 2 Id. at 5660. The February 2 Notice stated that work on the Congestion Study was “well underway” and that DOE intended to publish the study by August 8, 2006. Id. at 5661. The Notice also stated that a technical conference would be held in Chicago, Illinois, on March 29, 2006. Id. at 5660.

The technical conference was held in March 2006, and a number of State entities attended and some participated in various panels. A separate invitation-only meeting was held in May 2006 to “review and evaluate the congestion analyses performed by DOE’s contractors,” but no states were invited. 3

DOE asserts that it reached out to affected States through meetings with the National Association of Regulatory Utility Commissioners (“NARUC”) and through other meetings and correspondence with individual State entities.

DOE issued its Congestion Study in August 2006. The notice in the Federal Register requested comments “on the study and on the possible designation of national interest electric transmission corridors.” 71 Fed.Reg. 45,047 (Aug. 8, 2006). DOE received over 400 comments on the Congestion Study. On May 7, 2007, DOE responded to the comments and sought additional comments on “draft National Corridor designations for the two Critical Congestion Areas identified in the Congestion Study: the draft Mid-Atlantic Area National Corridor; and the draft Southwest Area National Corridor.” 72 Fed. Reg. 25,840 (May 7, 2007) (the “May 7 Notice”).

A major objection set forth in the comments was the assertion that DOE had failed to consult with affected States. Id. at 25,850 . DOE responded that it was “committed to fulfilling its obligation to consult with States” but asserted that “there are practical difficulties in conducting the level of consultation that some may prefer in the context of a study of this magnitude,” and that it “is difficult to know which States are ‘affected’ until the conclusions of the congestion study are *1082 known.” Id. DOE claimed to have met its obligation because it: (1) had “provided States with numerous opportunities for input and [] held meetings with officials representing individual States and groups of States;” (2) had made the Congestion Study available on August 8, 2006; and (3) had, “in addition to [having made] the draft National Corridor designations described in this notice available for comment, ... simultaneously contacted] the Governors of each State in which the draft National Corridors would be located to arrange consultation meetings.” Id.

The May 7 Notice also described a number of other comments that DOE received that may be divided into four groups. A first group of comments are objections to DOE’s interpretation of the scope of its authority to designate National Corridors, its definitions of “congestion” and “constraint,” and the need for the Congestion Study to accommodate state laws and policies on renewable portfolio standards. Id. at 25,842 . In responding to, and rejecting these objections, DOE stressed its discretion under § 216. Id. at 25,843. It noted that “there is no generally accepted understanding of what constitutes ‘constraints or congestion that adversely affects consumers’ ” and defended the definition adopted in the Congestion Study. Id. at 25,843-45.

A second group of comments concerned the relationship between regional planning processes and the designation of national corridors. Id. at 25,846. For example, NARUC argued that DOE should grant deference to the results of adequate regional planning processes, and the Public Utilities Commission of the State of California (“CPUC”) argued that “designation is unwarranted unless there is evidence that State and regional processes are not addressing the problem in a timely manner.” Id. DOE indicated that it “supports and encourages regional planning efforts” but nonetheless had decided not to defer to regional planning processes. Id. at 25,-846-47.

A third group raised comments concerning the drawing of boundaries. Some comments advocated the use of specific transmission projects to define national corridor boundaries, others suggested that boundaries should be tailored to aid in the construction of specific viable transmission projects, and “numerous commenters” argued that DOE “should draw National Corridor boundaries to exclude parks and other environmentally protected areas.” Id. at 25,847. In response, DOE first noted that the “statute provided little direction on how the Department should draw the boundaries of a National Corridor.” Id. at 25,848. DOE observed that the selection of source areas “will necessarily involve discretion and is not suited to a formulaic approach.” Id. In declining to make any changes in response to the comments, DOE noted:

Id. at 25,849.

The fourth group of comments asserted that DOE was required to prepare a Programmatic Environmental Impact Statement pursuant to NEPA before designating any National Corridor. Id. at 25,850. *1083 DOE responded that although NEPA requires environmental impact statements for major Federal actions, “[t]he designation of a National Corridor ... does not significantly affect the quality of the human environment,” and accordingly a “National Corridor designation is not a ‘proposal for a major Federal action significantly affecting the quality of the human environment’ that falls within the purview of NEPA.” 4 Id. at 25,851.

C. The Designation Order

On October 5, 2007, DOE issued its order formally designating two NIETCs, the Mid-Atlantic Area National Interest Electric Transmission Corridor (the “Mid-Atlantic Corridor”) and the Southwest Area National Interest Electric Transmission Corridor (the “Southwest Corridor”). 72 Fed.Reg. 56,992 (Oct. 5, 2007). DOE explained its perspective and its rejection of all comments recommending different approaches. 5 DOE reiterated that no environmental study was necessary, explaining:

72 Fed.Reg. at 57,022.

Finally, in March 2008, DOE issued an order denying rehearing of its Designation Order. 73 Fed.Reg. 12,959 (Mar. 11, 2008). DOE reiterated its determination that no environmental study was needed and further noted that “it would be highly speculative for the Department to make assumptions about whether, when, or where FERC might permit transmission facilities.” Id. at 12,969.

D. The Judicial Proceedings

The first petition to review DOE’s actions was filed by The Wilderness Society, et al. in this court on March 14, 2008. Twelve other petitions for review were timely filed in other Circuits. Pursuant to stipulations, all thirteen petitions were consolidated in the Ninth Circuit and are before this panel for consideration.

II. THE REQUIREMENT FOR CONSULTATION

A. Standards of Review

In reviewing DOE’s actions, we are guided by the standard of review established by Chevron USA, Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 , 104 S.Ct. 2778 , 81 L.Ed.2d 694 (1984), and its progeny. The Court’s seminal statement is:

Id. at 842-43 , 104 S.Ct. 2778 (footnotes omitted). The Court, however, also noted that “[t]he judiciary is the final authority on issues of statutory construction and must reject administrative constructions which are contrary to clear congressional intent,” and that “[i]f a court, employing traditional tools of statutory construction, ascertains that Congress had an intention on the precise question at issue, that intention is the law and must be given effect.” Id. at 843 n. 9, 104 S.Ct. 2778 .

Both prongs of the Chevron standard are in play in this case. On the one hand, Congress clearly directed DOE to engage in “consultation with affected States.” On the other hand, it did not explicitly define “consultation.” DOE urges that accordingly, Congress’s intent is not clear and that the highly deferential standard of review for agency action under the Administrative Procedure Act applies. See 5 U.S.C. § 706 . 6

We have recognized that this standard is “highly deferential, presuming the agency action to be valid” and that we may not substitute our judgment for that of the agency. Nw. Ecosystem Alliance v. U.S. Fish and Wildlife Serv., 475 F.3d 1136 , 1140 (9th Cir.2007). We therefore will “affirm!] the agency action if a reasonable basis exists for its decision.” Id. “Our task is simply to ensure that the agency considered the relevant factors and articulated a rational connection between the facts found and the choices made.” Id. (internal quotation marks and citation omitted).

*1085 B. Congressional Intent

Congress could hardly have been more explicit in directing DOE to consult with “affected States.” Section 824p(a)(l) specifically directs DOE to conduct a study of electric transmission congestion “in consultation with affected States.” Furthermore, § 824p(a)(2) directs DOE to issue a report which may designate National Interest corridors, “[a]fter considering alternatives and recommendations from interested parties (including an opportunity for comment from affected States).”

“Our first step in interpreting a statute is to determine whether the language at issue has a plain and unambiguous meaning....” Robinson v. Shell Oil Co., 519 U.S. 337, 340 , 117 S.Ct. 843 , 136 L.Ed.2d 808 (1997). This is “determined by reference to the language itself, the specific context in which that language is used, and the broader context of the statute as a whole.” Id. at 341 , 117 S.Ct. 843 . See also United States v. Morton, 467 U.S. 822, 828 , 104 S.Ct. 2769 , 81 L.Ed.2d 680 (1984) (“We do not, however, construe statutory phrases in isolation; we read statutes as a whole.”).

Here, § 824’s direction to the DOE to undertake the Consultation Study “in consultation with affected States” in context clearly means that DOE should have greater interaction with the States in preparing the Congestion Study than it need have when preparing a NIETC report, when it need only provide “an opportunity for comment from affected States.” Indeed, DOE does not really deny that it was required to “consult” with “affected States.” See 72 Fed.Reg. at 25,838 (stating that “FPA section 216(a)(1) requires the Secretary to consult with ‘affected States’ ”); 72 Fed.Reg. at 56,993 (stating that “FPA section 216(a)(1) states that the Department shall conduct the congestion study in consultation with affected States”).

Accordingly, the issue is whether DOE, while undertaking the Congestion Study, consulted with the affected States as mandated by Congress. This inquiry requires that we review DOE’s efforts to involve the affected States in the preparation of the Congestion Study and then evaluate whether those efforts amount, as DOE contends, to consultation. Finally, if we determine that DOE’s actions did not amount to consultation, we must determine whether any such shortcoming constitutes harmless error.

III. DOE FAILED TO CONSULT

A. DOE’s interaction with the “affected States”

The record shows that DOE sought the affected States’ input in three ways but also excluded them from participating in several significant respects. First, on February 2, 2006, DOE invited the public, including the “affected States,” to provide comments for its ongoing Congestion Study. Second, DOE informed the affected States of a technical conference that would be held in Chicago in March 2006. Third, when it issued the Congestion Study in August 2006, it invited comments on the designation of NIETCs. 71 Fed. Reg. at 45,047.

On the other hand, the record shows that DOE did not extend an invitation to potentially affected States to attend an “invitation-only” workshop on the Congestion Study that was held in May 2006. Also, DOE did not disclose to the affected States the congestion modeling data it used to conduct the Congestion Study. Furthermore, DOE never extended any invitation to the affected States or their *1086 Governors to “consult” on the preparation of the Congestion Study. 7

DOE also cites its meetings with NARUC and meetings and conferences with other State entities as evidence that it met its obligation to consult. There is little, however, to suggest that these events provided meaningful opportunities for dialogues between the States and DOE. NARUC, of course, is not a state or even a state entity. Rather, it is a “quasi-governmental organization that includes representatives of all fifty states....” NARUC v. FCC, 746 F.2d 1492 , 1497 n. 2 (D.C.Cir.1984). Moreover, NARUC itself advised DOE that meeting with it was not the same as consulting with the affected States. 8 The record also fails to support DOE’s assertion that those meetings that were held with certain State entities allowed for a meaningful exchange of information. 9

In sum, DOE’s claim that it met its obligation to consult with the affected States is based on the argument that it had the discretion to determine what “consultation” required, that it met its obligation by inviting comments from the public (including the affected States) while it was preparing the Congestion Study, that it subsequently considered all objections to the Congestion Study raised by the affected States, and that any failures in this process of “consultation” were harmless.

B. DOE’s interactions do not amount to consultation

DOE claims that “§ 216 does not require more than notice-and-comment proceedings.” DOE asserts that because Congress did not define what it meant by “consultation,” we must defer to DOE’s interpretation of the term. However, we do not read the statute as encompassing DOE’s proffered definition, and such a definition is contrary to the applicable rules of statutory interpretation as it would render Congress’s choice of language meaningless. Moreover, we find no support for DOE’s position in the relevant case law.

*1087 1. The definition of “consultation”

An ordinary meaning of the word consult is to “seek information or advice from (someone with expertise in a particular area)” or to “have discussions or confer with (someone), typically before undertaking a course of action.” The New Oxford Dictionary 369 (2001) (emphasis added). We conclude that this is the definition that Congress intended when it directed DOE to prepare the Congestion Study “in consultation with the affected States.” Thus, DOE was to confer with the affected States before it completed the study.

This conclusion is supported by all the applicable rules of statutory construction. It is required by the statutory context as the juxtaposition of the two sections indicates that Congress intended consultation to be more than responding to comments. See Robinson v. Shell Oil Co., 519 U.S. 337, 341 , 117 S.Ct. 843 , 136 L.Ed.2d 808 (1997) (noting that the plain meaning of statutory language is determined by reference to the specific context in which the language is used and the broader context of the statute as a whole). The definition gives meaning to every word in the statute. See Duncan v. Walker, 533 U.S. 167, 174 , 121 S.Ct. 2120 , 150 L.Ed.2d 251 (2001) (noting that it is a court’s “duty to give effect, if possible, to every clause and word of a statute.”). Moreover, DOE’s interpretation of “consult” to mean no more than notice-and-comment would render part of the statute superfluous. If “consultation” means no more than “an opportunity for comment,” there was no reason for Congress to use distinct language in § 824(a)(1) and § 824(a)(2). We have been directed to avoid such an interpretation. See Knight v. Comm’r, 552 U.S. 181, 190 , 128 S.Ct. 782 , 169 L.Ed.2d 652 (2008) (commenting that “accepting [a particular] approach would render part of the statute entirely superfluous, something that we are loath to do”) (quoting Cooper Indus. v. Aviall Servs., 543 U.S. 157, 166 , 125 S.Ct. 577 , 160 L.Ed.2d 548 (2004)).

Moreover, requiring DOE to actually confer with the affected States is consistent with the purpose of the EPAct. In reaction to black-outs and brown-outs, Congress sought to give the federal government a greater role in the development of transmission lines and to circumscribe somewhat the States’ traditional authority over the placement and construction of power lines. In recognition of this impact on the States’ traditional authority, Congress intended that affected States would participate in a study that might ultimately result in some limitation of their traditional powers. Indeed, the Supreme Court has been sensitive to these concerns. See Solid Waste Agency of Northern Cook County v. U.S. Army Corps of Eng’rs, 531 U.S. 159, 173 , 121 S.Ct. 675 , 148 L.Ed.2d 576 (2001). A recognition of the sensitivity of these issues supports our determination that where, as here, Congress has directed an agency to consult with States before taking action that may curtail traditional State powers, we must require that the agency heed Congress’s direction.

2. Case law defining consultation

In addition, our conclusion that the ordinary meaning of consult involves conferring with an entity before taking action is amply supported, if not compelled, by our relevant precedent. In Environmental Defense Center v. EPA, 344 F.3d 832 (9th Cir.2003), we considered a provision that required the EPA to conduct certain studies “in consultation with the States,” and to issue regulations based on these studies “in consultation with State and local officials.” Id. at 863. EPA asserted that it had met its obligation by consulting extensively with States and localities before is *1088 suing its regulations. Id. at 864. We agreed, noting:

Id. None of the efforts noted in Environmental Defense Center are present here. No draft was circulated to the States, no committee was created that included representatives from the States, and the affected States were not given access to the supporting data. Thus, DOE’s efforts here fall far short of the efforts that were determined to meet the requirement for consultation in Environmental Defense Center.

In Confederated Tribes & Bands of Yakima Indian Nation v. FERC, 746 F.2d 466 (9th Cir.1984), we held that the FERC violated its duty of consultation. We noted that it was not enough to give notice to the agencies and Indian tribes, as the “consultation obligation is an affirmative duty.” Id. at 475. We noted that “the respective fishery agencies believed the consultation process would take place in the preparation of [a fish and wildlife report],” but that the agency “issued the [report] before the exhibit was submitted.” Id. Here too, the agency had an affirmative duty to consult and the affected States reasonably believed the consultation process would take place, but DOE issued the Congestion Study without engaging in any meaningful consultation with the States.

Our perspective is also consistent with the opinion of the United States Court of International Trade in U.S. Steel Corp. v. United States, 29 C.I.T. 33 , 362 F.Supp.2d 1336 (Ct. Int’l Trade 2005). Addressing notice, comment, and consultation requirements, the court held that “it is not enough to prove that the agency solicited and received comments from the Domestic Producers before executing the Suspension Agreement.... The agency must also give those comments meaningful consideration” and “must engage the Domestic Producers in good faith consultations, in a timely fashion.” Id. at 40 . The court found that:

Id. at 40 n. 14. We think that DOE pursued a similarly erroneous course here, attempting to conflate its obligation to consult with the affected States while preparing the Congestion Study with its obli *1089 gation to provide the States an opportunity to comment on its NIETC report.

3. DOE’s failure to provide the affected States with modeling data interfered with their ability to consult with DOE

We note that, by failing to provide the affected States with the modeling data on which it based the Congestion Study, DOE prevented the affected States from providing informed criticism and comments. There can be no doubt that the modeling data was critical to DOE’s study. 10 Moreover, DOE recognizes that even under a notice-and-comment requirement, an agency has a duty to “identify and make available technical studies and data that it has employed in reaching the decisions to propose particular rules.” Kern County Farm Bureau v. Allen, 450 F.3d 1072, 1076 (9th Cir.2006). DOE, however, defends its failure to disclose any of the modeling data to the affected States prior to the issuance of the Congestion Study by arguing that it disclosed the technical studies and data after it issued the Congestion Study and that the information did not need to be disclosed because it was proprietary. 11 We have already held that DOE’s duty to consult in preparing the Congestion Study is separate from, and requires greater interaction with the affected States, than DOE’s obligation to the States when preparing the NIETC report. Accordingly, its post-study release of the information does not excuse its failure to consult with the affected States in preparing the Congestion Study.

Moreover, DOE’s argument that the proprietary interests in the modeling data justified their retention is not well taken. First, as noted by the States, there is no factual or legal basis for DOE’s unstated assumption that the States would not, or could not, respect any legitimate proprietary interests in the modeling data. Second, and more importantly, the case cited by DOE, American Radio Relay League, Inc. v. Federal Communications Commission, 524 F.3d 227 (D.C.Cir.2008), specifically states that under the Administrative Procedure Act (“APA”) an agency must disclose the technical studies and data upon which it bases a ruling. 12 Third, *1090 there is no doubt that the modeling information was critical to DOE’s preparation of the Congestion Study. 13 The ability to “consult” on a matter is severely compromised when an entity is denied access to the basis of the decision.

C. The Failure to Consult Was Not Harmless Error

More importantly, we cannot conclude that DOE’s failure to consult was harmless error. Certainly, American Radio Relay League notes that any failure to disclose information for public comment is subject to the rule of prejudicial error. Id. at 237 . We also have held that when reviewing agency action under the APA, we must take “due account” of the harmless error rule. Paulsen v. Daniels, 413 F.3d 999, 1006 (9th Cir.2005). We have stressed, however, that a court “must exercise great caution in applying the harmless error rule in the administrative rulemaking context.” Id. In Riverbend Farms, Inc. v. Madigan, 958 F.2d 1479 (9th Cir.1992), which involved a failure of an agency to fulfill the notice-and-comment procedures of the APA, we stated:

Id. at 1487.

We have applied this definition of harmless error — “clearly had no bearing on the procedure used or the substance of decision reached” — in a number of cases over the last eighteen years. For example, see Paulsen, 413 F.3d at 1006-08 (adopting the “no bearing” standard, noting that the agency’s “mistake clearly had a bearing on the procedure used,” and commenting that the fact that “petitioners had an opportunity to protest an alreadyeffeetive rule prior to the time it was applied to each of them does not render the APA violation harmless”); City of Sausalito v. O’Neill, 386 F.3d 1186 , 1220 (9th Cir.2004) (holding that “in the rulemaking context, we exercise great caution in ap *1091 plying the harmless error rule, holding that failure to provide notice and comment is harmless only where the agency’s mistake clearly had no bearing on the procedure used or the substance of decision reached”) (internal quotation marks and citations omitted); Gifford Pinchot Task Force v. U.S. Fish and Wildlife Service, 378 F.3d 1059 , 1071 (9th Cir.2004) (noting that in the context of agency review, the role of harmless error is constrained and may be employed only “when a mistake of the administrative body is one that clearly had no bearing on the procedure used or the substance of decision reached”); and Buschmann v. Schweiker, 676 F.2d 352, 358 (9th Cir.1982) (holding that an agency can rely on harmless error only when a mistake of the administrative body is one that clearly had no bearing on the procedure used or the substance of decision reached) (internal quotations marks and citations omitted). 14

The dissent, however, posits that the Supreme Court’s recent decision in Shinseki v. Sanders, — U.S. -, 129 S.Ct. 1696 , 173 L.Ed.2d 532 (2009), requires a different definition of “harmless error.” We do not agree. In Sanders , the Supreme Court clarified that in agency cases, as in appellate review of civil cases, “the burden of showing an error is harmful normally falls upon the party attacking the agency’s determination.” Id. at 1706 . The Supreme Court struck down the Federal Circuit’s “harmless error” framework, in part because it “require[d] the YA, not the claimant to explain why the error [was] harmless.” Id. at 1705 .

The Court, however, did not redefine “harmless error,” but rather embraced a commonsense approach to the concept.

129 S.Ct. at 1706 .

We do not think that this approach to harmless error allows us to depart from *1092 our consistent case law holding that “harmless error” requires a determination that the error “had no bearing on the procedure used or the substance of[the] decision reached.” Paulsen, 413 F.3d at 1006 (quoting Sagebrush, 790 F.2d at 764-65 ). In Miller v. Gammie, 335 F.3d 889 (9th Cir.2003) (en banc), we held that “where the reasoning or theory of our prior circuit authority is clearly irreconcilable with the reasoning or theory of intervening higher authority, a three-judge panel should consider itself bound by the later and controlling authority, and should reject the prior circuit opinion as having been effectively overruled.” Id. at 893 . Certainly, Sanders clarifies that the burden of showing an agency’s deviation from the APA was not harmless rests with the petitioner, but we see nothing in Sanders that is “clearly irreconcilable with the reasoning or theory” underlying our definition of harmless error. 15

Moreover, here our concern with process is very different from the issue in Sanders . There, the Supreme Court’s exclusive focus on the harmlessness of the error in affecting the final outcome — the determination of disability — was a logical application of the statutory and regulatory requirements of notice by the agency to the veteran as to how to pursue his claim. The notice requirement was not an end in itself; it was a means to permit the veteran to develop his claim for disability. If a failure of notice had no effect on the determination of the disability claim, there could be no harm; the failure of notice by itself was of no consequence.

In contrast, here, as in Riverbend Farms, the congressional notice requirement reflects the desirability of the interactive process itself. 16 958 F.2d 1479 . The consultative process dictated by Congress serves the purpose of permitting the States to participate in the formulation of federal policy in an area of major interest to the States. As in Riverbend Farms, Congress did not require the agency to accept the views of the States; its requirement was directed at process and not merely a final result (although early consultation often will lead to a better result). Id. at 1487. Under such a scheme, we are not free to depart from our consistent case law holding that a finding of “harmless error” requires a determination that the error “had no bearing on the procedure used or the substance of [the] decision reached.” Paulsen, 413 F.3d at 1006 (quoting Sagebrush, 790 F.2d at 764-65 ) (emphasis added).

The dissent, however, prefers the D.C. Circuit’s approach in Gerber v. Norton, 294 F.3d 173, 182 (D.C.Cir.2002), that to show “that error was prejudicial a plaintiff must indicate with reasonable specificity what portions of the documents it objects to and how it might have responded if given the *1093 opportunity.” (internal quotation marks and citations omitted). The dissent asserts that other circuits have adopted a similar approach for showing prejudice from procedural errors. 17

Although we are compelled to follow our prior case law, we would reach the same conclusion applying the standard proffered by the dissent. We find that the petitioners have demonstrated that they were excluded from the decisionmaking process, have indicated what evidence and information they would have provided if given the opportunity, and have shown how their interests were harmed by their exclusion.

Placing the burden of showing that DOE’s failure to consult was not harmless on the petitioners, we determine that the affected States have shown that DOE’s failure to comply with Congress’s mandate in § 216 was harmful. First, we note that although the nature of consultation makes it difficult to determine the precise consequences of its absence, the prejudice to the party excluded is obvious. Consultation requires an exchange of information and opinions before the agency makes a decision. This requirement is distinct from the opportunity to offer comments on the agency’s decision. The essential verity of this distinction is illustrated by posing the question: would any attorney forgo the opportunity to argue his client’s case before a judge renders a decision in favor of seeking reconsideration after the judge has made a decision? Of course not; such a decision might well amount to malpractice. Similarly, here, the opportunity to comment on DOE’s completed Congestion Study does not compensate for the lost opportunity of consulting with DOE in the formation of that study. 18

Another drawback of the dissent’s approach is its failure to appreciate that DOE’s failure to consult has a substantive, as well as procedural component. The exclusion of the affected States from the decisionmaking process not only limited the information available to DOE, it altered the way in which DOE made its discretionary decisions. In Kurzon v. United States Postal Service, 539 F.2d 788 (1st Cir.1976), the First Circuit indicated that where there was an alleged substantive error, it would remand “only if the court is in substantial doubt whether the administrative agency would have made the same ultimate finding with the erroneous finding removed from the picture.” Id. at 796 (internal quotation marks and citations omitted). Here, we are left with substantial doubt as to whether DOE would have made the same findings had it consulted with the affected States.

Second, the impact of the lack of consultation before a decision is made as contrasted to commenting after the agency has made a decision is particularly severe here because, as DOE admits, its decisions were for the most part discretionary. In its May 7, 2007, Notice, DOE stated that “there is no generally accepted under- *1094 standing of what constitutes ‘constraints or congestion that adversely affects consumers,’ ” 72 Fed.Reg. at 25,843, and noted that “[t]he statute provides little direction on how the Department should draw the boundaries of a National Corridor,” id. at 25,848. Thus, DOE admits that its determinations and conclusions in the Congestion Study were not decisions compelled by some mathematical formulae, but important discretionary decisions for which there was little guidance. 19 The value of consulting with an agency before it makes a decision is greatest when the agency is tasked with adopting a “novel approach” that will then affect all stakeholders. In such a situation, as here, a court can hardly conclude that the agency’s refusal to consult with the affected States had no bearing on the substance of the decision reached. See Paulsen, 413 F.3d at 1006 .

The dissent’s recitation of the States’ “opportunity to comment” reflects its failure to appreciate how consulting is different from commenting. Of course, the States were aware that Congress had direefed DOE to conduct a Congestion Study. They had to protect themselves, and perhaps try to reduce the potential harm from a lack of consultation, by responding to DOE’s request for comments. But, contrary to the dissent’s suggestion, there is no evidence that DOE ever consulted with any State. 20

Finally, a review of the objections filed by the affected States and others reveals that DOE’s failure to consult with the affected States in developing the Congestion Study was not harmless error and that consultation would probably have resulted in a different study. 21 Among the over 400 comments that DOE received when it published the Congestion Study were assertions that: (1) “the focus of the Congestion study is too narrow to accommodate State laws and policies on renewable portfolio standards”; (2) DOE had adopted too broad a definition of “adverse effects”; (3) designations should only be made for areas actually experiencing congestion adversely affecting consumers and not areas that *1095 may experience congestion in the future; and (4) DOE should clarify the criteria it would use in deciding whether to designate a National Corridor. 72 Fed.Reg. at 25,-842-43.

DOE does not claim that its decisions in the Congestion Study were compelled or that consultation could not have produced variations. Rather, DOE’s responses to the comments stress its “discretion,” implicitly recognizing that consultation might well have resulted in different decisions. For example, the first line of DOE’s response stresses that § 216 gives it “discretion.” Id. at 25,843. It admits that the term “constraints or congestion that adversely affects consumers ... is ambiguous and the statute attaches no modifiers to the term to specify the particular type of magnitude of adverse effect intended.” Id. DOE’s admission that a term is ambiguous and that the statute provides little guidance in its interpretation indicates that its decisions might well have been altered through consultation with the affected States.

The likelihood that consultation would produce different results may be seen, for example, in the petitioners’ objections to DOE’s designation of the entire Mid-Atlantic region as a NIETC Corridor and to their assertions that DOE failed to consider regional efforts to address energy constraints and congestions. This is not to suggest that DOE’s determinations were unreasonable. Rather, it appears that petitioners’ objections are not frivolous, may well have some merit, and thus, we cannot conclude that DOE, were it to exercise its discretion when informed by consultation with the affected States, would not modify its decisions.

The failure to consult was not some technical error, but resulted in a decision-making process that was contrary to that mandated by Congress and one that deprived DOE of timely substantive information. We conclude that DOE’s failure to consult with the affected States, as directed by Congress, was not harmless error.

IV. THE REMEDY

When a court determines that an agency’s action failed to follow Congress’s clear mandate the appropriate remedy is to vacate that action. See, e.g., Nw. Envtl. Advocates v. EPA, 537 F.3d 1006, 1026-27 (9th Cir.2008) (explaining that the district court’s decision to vacate the EPA’s action was the proper remedy when EPA acted outside of its authority and in defiance of Congress’s clear intent). Similarly, where a regulation is promulgated in violation of the APA and the violation is not harmless, the remedy is to invalidate the regulation. See Paulsen, 413 F.3d at 1008 . Accordingly, as we have determined that § 216 required more than the notice- and-comment procedure adopted by DOE, and that DOE’s failure to consult with the affected States was not harmless error, precedent and reason require that we vacate the Congestion Study and remand for the DOE to prepare a Congestion Study “in consultation with the affected States.”

DOE’s suggestion that we might vacate only those portions of the Congestion Study for which the States have shown prejudice, misconstrues the nature of the right to consultation as well as the deference we owe to DOE’s decisions. As noted, it is almost impossible to determine the precise impact of a decisionmaker’s failure to consult prior to making a discretionary decision. Because the Congestion Study invokes DOE’s sound discretion for which there are few, if any, objective criteria, we simply cannot know what DOE would have decided had it considered the affected State’s perspectives before it completed the study, or foretell what it will decide after consulting with the affected States. *1096 Accordingly, DOE must prepare a Congestion Study in consultation with the affected States which thereafter may be judicially reviewed. We express no opinion as to the form or results of the collaboration. Indeed, presumably DOE could, in the exercise of its sound discretion, come to the same or similar conclusions that it did in the initial study. Of course, it might reach very different conclusions. What is critical is that it follow the statute’s mandate and consult with affected States, particularly as § 216 requires DOE to prepare a congestion study every three years.

V. THE FAILURE TO UNDERTAKE AN ENVIRONMENTAL STUDY

We next address DOE’s failure to prepare an Environmental Impact Statement (“EIS”) or an Environmental Assessment (“EA”) for either of the NIETCs. We do so for two reasons. First, even if we had not determined that the Congestion Study must be vacated, we would nonetheless hold that the NIETCs must be vacated because DOE violated the law in failing to consider the environmental consequences of the NIETCs. Second, because DOE will now have to prepare new NIETCs based on a new Congestion Study, our guidance on this issue should be useful for all concerned.

A. The Applicable Law

All parties agree that pursuant to the National Environmental Policy Act, 42 U.S.C. § 4882 (2)(C), DOE, like any other federal agency, must include “in every recommendation or report on proposals for legislation and other major Federal actions significantly affecting the quality of the human environment, a detailed statement by the responsible official on” the potential environmental consequences of the action. 22 Id.

In Marsh v. Oregon Natural Resources Council, 490 U.S. 360 , 109 S.Ct. 1851 , 104 L.Ed.2d 377 (1989), the Supreme Court noted that “NEPA promotes its sweeping commitment to prevent or eliminate damage to the environment and biosphere by focusing Government and public attention on the environmental effects of proposed agency action” so that the “agency will not act on incomplete information, only to regret its decision after it is too late to correct.” Id. at 371 , 109 S.Ct. 1851 (internal quotation marks and citations omitted). In Winter v. Natural Resources Defense Council, Inc., 555 U.S. 7 , 129 S.Ct. 365 , 172 L.Ed.2d 249 (2008), the Supreme Court reiterated that “[p]art of the harm NEPA attempts to prevent in requiring an EIS is that, without one, there may be little if any information about prospective environmental harms and potential miti *1097 gating measures.” 129 S.Ct. at 376; see also Monsanto v. Geertson Seed Farms, - U.S. -, 130 S.Ct. 2743, 2768 , 177 L.Ed.2d 461 (2010) (Stevens, J., dissenting) (noting that an EIS is especially important where the environmental threat is novel). Ultimately, our role “is to insure that the agency has taken a ‘hard look’ at environmental consequences....” Kleppe v. Sierra Club, 427 U.S. 390 , 410 n. 21, 96 S.Ct. 2718 , 49 L.Ed.2d 576 (1976).

In accord with this approach, we have reiterated that the “agency bears the primary responsibility to ensure that it complies with NEPA.” ‘Ilio‘ulaokalani Coal. v. Rumsfeld, 464 F.3d 1083 , 1092 (9th Cir.2006) (quoting Dep’t of Transp. v. Pub. Citizen, 541 U.S. 752, 765 , 124 S.Ct. 2204 , 159 L.Ed.2d 60 (2004)). We reiterated in Alaska Ctr. for Environment v. U.S. Forest Service, 189 F.3d 851, 859 (9th Cir.1999), that “[w]hen an agency decides to proceed with an action in the absence of an EA or EIS, the agency must adequately explain its decision.” Id. (internal citation omitted). We commented that “[a]n agency cannot avoid its statutory responsibilities under NEPA merely by asserting that an activity it wishes to pursue will have an insignificant effect on the environment.” Id. (quoting The Steamboaters v. FERC, 759 F.2d 1382, 1393 (9th Cir.1985)).

In Klamath Siskiyou Wildlands Center v. Boody, 468 F.3d 549 (9th Cir.2006), we noted that an EIS “must be prepared if substantial questions are raised as to wdiether a project may cause significant degradation of some human environmental factor.” Id. at 562 (internal citation omitted). We explained that “[t]he plaintiff need not show that significant effects will in fact occur, but if the plaintiff raises substantial questions whether a project may have a significant effect, an EIS must be prepared,” and noted that “[t]his is a low standard.” Id. (internal citation omitted). In addition, we stated:

Klamath, 468 F.3d at 562 .

Thus, our precedents hold that an agency cannot merely assert that its decision will have an insignificant effect on the environment, but “must adequately explain its decision.” Alaska Ctr., 189 F.3d at 859 . In The Steamboaters , we reversed and vacated the agency’s order because the agency failed to prepare even an EA and did “not discuss the evidence presented by the various agencies or how the particular conditions placed on the project would prevent environmental damage.” 759 F.2d at 1393 . We explained that the “agency must supply a convincing statement of reasons why potential effects are insignificant.” Id. (internal citation omitted). “The appellate court must be able to determine whether the agency took a ‘hard look’ at the potential environment impacts of the project” *1098 and “[t]he statement of reasons is crucial to such a determination.” Id.

B. Evaluating DOE’s assertion of no environmental impact

We apply these standards to DOE’s assertion that, although NEPA applies, it was not required to undertake any review of potential environmental consequences because the NIETCs do not have any environmental effects. 23 We are compelled to reject DOE’s assertion because (1) its conclusory statement does not allow us to determine whether DOE took a “hard look” at the potential environmental consequences; and (2) although the effects of the NIETCs may be uncertain and difficult to quantify, the potential consequences of such effects are significant enough to undermine DOE’s conclusory determination that no EA need be prepared. 24

1. The NIETCs do not determine the siting of any particular facility

DOE argues that the NIETCs do not have any environmental effect because they do not approve of the siting of any transmission facility, and furthermore, any particular siting will be subject to NEPA review. Our precedent, however, provides that agency action may constitute a “major Federal action” even though the program does not direct any immediate groundbreaking activity.

In Forelaws on Board v. Johnson, 743 F.2d 677 (9th Cir.1984), the petitioners challenged the Bonneville Power Administration’s (“BPA”) offers of long-term contracts for power. Id. at 679 . Although BPA argued that its actions “that merely allocate federal power to different customers do not significantly affect the environment,” we held that the contracts raised “considerations of far greater historic and regional import and significantly affect the environment.” 25 Id. at 682 . Accordingly, we concluded that BPA’s action was not sufficient and required the preparation of an EIS. Id. at 686 .

DOE recognizes the relevancy of Fore-laws, but seeks to distinguish the case on the ground that DOE has no authority to site electric transmission facilities. This distinction is not persuasive because the NIETCs, in essence, influence the areas in which electric transmission facilities will be located, even though they do not determine the precise locations of the facilities. As in Forelaws, the locations of those areas could have great historic and regional consequences that significantly affect the environment. Thus, the fact that the NIETCs do not approve the actual sitings of specific transmission facilities does not excuse DOE from considering the NIETCs’ environmental impacts.

Furthermore, Forelaws does not stand alone in holding that broad agency programs may constitute “major Federal actions,” even though the programs do not direct any immediate ground-disturbing activity. See Oregon Natural Desert Ass’n v. BLM, 531 F.3d 1114, 1116 (9th Cir.2008) *1099 (finding EIS inadequate for land use plan covering a large portion of Oregon); N. Alaska Envtl. Ctr v. Kempthorne, 457 F.3d 969, 973 (9th Cir.2006) (noting that an EIS was prepared for agency action making entire Northwest Petroleum Reserve available for oil and gas leasing despite the lack of “site specific analysis for particular locations where drilling might occur”); Friends of Yosemite Valley v. Norton, 348 F.3d 789, 800-01 (9th Cir.2003) (evaluating programmatic EIS for land use plan for national park); 26 Blue Mountains Biodiversity Project v. Blackwood, 161 F.3d 1208, 1213 (9th Cir.1998) (finding EA inadequate and requiring EIS for log-salvaging plan for national forest). 27

2. The NIETCs are major federal actions

a. Northcoast Environmental Center v. Glickman

DOE asserts that the NIETCs are not major federal actions because it would be pure speculation to predict their environmental impacts. Citing Northcoast Environmental Center v. Glickman, 136 F.3d 660 (9th Cir.1998), DOE asserts that an agency action that has only speculative environmental impacts is not a major federal action.

Our holding in Northcoast is more nuanced than suggested by DOE. At issue in that case was a proposal by the Forest Service (“FS”) to establish guidelines for research, management strategies, and information sharing concerning a root rot fungus on federal land in Oregon. Id. at 670 . The district court found that the programs did not constitute final agency action subject to judicial review and that even if they did, they “were not major federal actions significantly affecting the environment.” Id. at 668 .

On appeal, we first noted that where review is sought under the general review provision of the APA, the agency’s decision must be a final agency action and the plaintiffs “must establish they have suffered a legal wrong, or will be adversely affected or aggrieved within the meaning of the relevant statute.” Id. We proceeded to comment that the agency action “must (1) be federal, (2) ‘major’, and (3) have a significant environmental impact.” 28 Id. The opinion focused on the third requirement. We concluded that the *1100 district court “properly recognized” that none of the activities had an actual or immediately threatened effect on the environment and “correctly decided” that the FS “reasonably found” that its actions did not “significantly affect the quality of human environment.” 29 Id. at 669-70 . Although sympathetic to plaintiffs’ concern that agencies should conduct a full NEPA analysis when management plans are implemented or proposed, we concluded that the current forest management programs did not “call for specific enough action to trigger NEPA’s procedural requirements,” and noted that plaintiffs could “challenge the sufficiency of an agency EIS when discrete agency action is called for.” Id. at 670 .

Northcoast offers several points of guidance. First, in determining whether the program had a significant environmental impact, we implicitly held that the program was a final agency action subject to review under the APA, even though we recognized that plaintiffs could “challenge the sufficiency of an agency EIS when discrete agency action is called for.” 30 Id. Second, we determined that the program was, at least potentially, a “major” Federal action. It is not clear whether the requirement that agency action be “major” was considered separately from the requirement that the action have “significant environmental impact,” or whether the latter was treated as an element of the former. In any event, the opinion certainly implies that if the program did have a significant environmental impact, it would have been a “major” federal action.

Third, although Northcoast states that an agency need not prepare an environmental study when its action does not have a significant environmental impact, it also holds that the record must be sufficient to allow the court to determine that the agency’s conclusion was reasonable. See 136 F.3d at 670 .

Here, the NIETCs are undoubtedly final agency actions. The NIETCs conclude DOE’s responsibilities under § 216. They establish the boundaries for two national electric transmission- corridors. Once the NIETCs become final, any question as to the actual siting of a facility within the corridors will be addressed to FERC. See 73 Fed.Reg. at 12,969 (“DOE agrees that the effect of a National Corridor is to delineate geographic areas within which, under certain circumstances, FERC may ultimately authorize the construction or modification of electric transmission facilities.”).

*1101 Both the intent and impact of the NIETCs support the conclusion that they constitute major Federal action. They create “National Interest” corridors to address national concerns. The NIETCs cover over a 100 million acres in ten States. Moreover, they create new federal rights, including the power of eminent domain, that are intended to, and do, curtail rights traditionally held by the states and local governments. See 16 U.S.C. § 824p(b), (e). In sum, we hold that the NIETCs are final agency actions that constitute major Federal actions.

b. The NIETCs raise significant environmental impacts

The remaining question is whether the NIETCs could have significant environmental impacts or, more accurately, whether DOE has created a record sufficient to allow us to evaluate whether its “no effects” determination is reasonable.

DOE proffers four arguments against being required to undertake an environmental study. First, DOE contends that no potential project-specific impacts are reasonably foreseeable or caused by the NIETCs. DOE contends that the NIETCs are not decisions to add transmission capacity to solve the problems of congestion or to site transmission facilities along preselected routes. DOE claims that these decisions remain to be made by multiple independent actors, and given “the vast range of options available ... it would be pure speculation to predict environmental impacts or assign them (as a matter of causation) to the Designation Order.” DOE further asserts that under § 216 its limited task was to “determine the conditional availability of a federal forum for siting transmission projects,” and it would have been premature for DOE “to evaluate the potential environmental impacts of new transmission facilities when deciding merely whether a federal forum should be made available.” DOE also contends that even if the NIETCs were certain to result in specific projects being submitted to FERC, DOE was not required “to prejudge the potential impacts of those projects” because a project-specific NEPA review is required before a permit issues.

Second, DOE claims that the NIETCs have no foreseeable programmatic effects. DOE admits that in some instances NEPA may require review of “programmatic” decisions that prescribe future actions, even though project-specific NEPA review will occur before a particular project is undertaken. 31 Nonetheless, DOE maintains that the NIETCs are not programmatic deci *1102 sions with reasonably-foreseeable future effects because each NIETC “is not a plan to guide land management or energy policy decisions,” but “merely makes available a federal procedural remedy (i.e., a forum for the consideration of interstate transmission lines), in the event that FERC finds relevant State forums to be inadequate per the standards set by Congress.” DOE maintains that the “addition of a backstop federal forum” does not mean that States and FERC will approve a greater number of projects and it “does not favor transmission solutions over non-transmission alternatives ... nor particular generation sources over others.” DOE recognizes that it is tasked with choosing the geographic boundaries of the National-Interest Corridors, but asserts that petitioners have failed to show that these boundaries circumscribe relevant alternatives as they place no limits on State siting authorities.

Third, DOE denies that the NIETCs could have any impacts on sensitive areas such as critical habitat for endangered species, scenic rivers, wilderness areas, and historic sites. DOE points out that an EIS must be prepared whenever substantial questions are raised about whether a specific project may have a significant effect. DOE further asserts that petitioners have the burden of showing that the potential impacts to sensitive resources are a “reasonably foreseeable” result of the designations. DOE maintains that “[t]he very breadth of these designations belies any suggestion that impacts can be meaningfully evaluated at the designation stage, even if it is assumed that the designation will prompt additional transmission projects.” DOE disagrees with petitioners’ claim that the inclusion of land within a corridor will discourage conservation, opining that a NIETC “might as readily spur the expansion of parks and conservation easements within the Corridors, as interested parties seek to protect sensitive resources.” 32 DOE further argues that claims of potential habitat fragmentation within a corridor cannot be meaningfully reviewed because of the many variables and wide range of alternatives. DOE claims that any suggestion that environmentally sensitive areas might be excluded from the corridors “confuses DOE’s threshold task (designating areas with congestion problems) with the States’ and FERC’s subsequent task (evaluating proposed solutions).”

Fourth, DOE argues that the NIETCs do not diminish any legal protections because “Congress provided that nothing in § 216 alters federal environmental laws, including laws requiring special authorization for use of federal lands or federal permits for impacting air and water resources.” The NIETCs do not allow power companies “to run away” from state and federal environmental and land use laws because they, in themselves, have “no preemptive effect, and FERC’s authority to preempt State law under § 216(b) is project-specific and limited to circumstances enumerated by Congress.” According to DOE, there are no “foreseeable adverse effects from the mere threat of federal intervention” because “potential acceleration of State proceedings does not dictate the outcome of those proceedings” and because DOE does not have “any discretion to alter the statutory time frames,” which might preclude meaningful “review of their potential effects.”

There may be merit to some of DOE’s arguments in terms of limiting the scope of *1103 an EIS or in explaining why an EA and not an EIS should be prepared, but they fail both as a matter of law and fact to justify DOE’s failure to undertake any study of the potential environmental impacts.

DOE’s primary argument appears to be that because the NIETCs do not approve any specific sites, they have no meaningful environmental impact. This perspective fails to appreciate that a decision to encourage, through a number of incentives, the siting of transmission facilities in one municipality rather than another has effects in both municipalities in terms of the values of land and proposed and potential uses of land. The effects may be difficult to measure and may be determined ultimately to be too imprecise to influence the Designation, but this is precisely the type of determination that only can be intelligently made after the preparation of at least an EA.

Recognition of these consequences flowing from the NIETCs defeats most of DOE’s reasons for not preparing an EA or EIS. Without such a study, it is impossible to fairly determine whether project-specific impacts are reasonably foreseeable, whether there are “programmatic effects,” 33 and whether the Designation has any impact on sensitive areas. Furthermore, the NIETCs do diminish legal protections at least as to whether any particular geographic area should be included in a corridor. The particular siting of a transmission facility may be challenged before a State or FERC, but a challenge to a specific site cannot challenge the inclusion of the area involved in the NIETCs by DOE. Thus, the alleged impact of the NIETCs’ inclusion of particular areas as within the corridors, and the exclusion of other areas, are subject to review for environmental impacts at this time or not at all.

c. The relevance of the environmental study for the West-unde Corridors

Any remaining doubt as to whether it is possible to consider the environmental impacts of the NIETCs dissipates in light of DOE’s preparation of a Programmatic Environmental Impact Statement (“PEIS”) for its designation of the West-wide Corridors for federal lands in eleven western states. See U.S. Department of Energy et al., Programmatic Environmental Impact Statement, Designation of Energy Corridors on Federal Land in the 11 Western States (DOE/EIS-0386), 2007 (hereinafter “PEIS”). A separate and distinct provision in the EPAct, § 368 of the Energy Policy Act of 2005 ( Pub.L. No. 109-58, § 368 , 119 Stat. 727 , codified at 42 U.S.C. § 15926 ), directs federal land-management agencies to identify rights-of-way across lands they administer to serve as energy corridors. DOE points out that the statute provides that any “corridor designated under this section shall, at a minimum, specify the centerline, width, and compatible uses of the corridor.” 42 U.S.C. § 15926 (e). Together with the Department of Interior, DOE prepared the required PEIS. See 42 U.S.C. § 15926 (a)(2) (providing for the preparation of “any environmental reviews that may be required *1104 to complete the designation of such corridors”).

The federal agencies issued the PEIS in October 2007. Two aspects of the PEIS are of particular relevance to this case. First, in response to the question “why conduct an environmental review under NEPA and prepare a programmatic analysis,” the PEIS’s executive summary states:

PEIS, supra, Executive Summary, at ES.8 (footnote omitted).

Second, after identifying an “ ‘unrestricted’ conceptual West-wide network of energy transport paths,” the executive summary explained:

PEIS, supra, at ES.12.2.1.

We recognize that the PEIS and the West-wide Designation were undertaken pursuant to a separate and distinct provision of the EPAct. Nonetheless, the creation of the PEIS and its impact on the resulting corridor designation is strong evidence both that it is possible to determine the environmental impacts of a proposed energy corridor and that the study of such environmental impacts may result in modifications of a corridor’s boundaries. The West-wide Corridors Designation, like the NIETC Designation, did not approve any specific sites, but designated specific areas for sites. Nonetheless, the lead agencies, including DOE, reshaped the corridors in *1105 response to the PEIS to exclude certain sensitive lands. Certainly § 15926 contains a more specific requirement for a study of environmental impacts than § 216, but DOE’s ability to undertake a PEIS for West-wide Corridors, and to modify the boundaries based on the PEIS, undermines its assertion that it is not possible to evaluate the environmental impacts of a NIETC.

3. DOE has not adequately documented its decision

DOE also asserts that it has adequately documented its decision not to undertake any review under NEPA. It argues that similar to the situation in Northcoast, neither NIETC is a “specific proposal with environmental consequences that can be meaningfully evaluated at this time.” 136 F.3d at 663 (internal quotations omitted). We doubt that a NIETC is similar to the management guidelines at issue in North-coast, but even if we were to engage in this fiction, this case does not contain the critical factual element present in Northcoast: a record that supports the reasonableness of the agency’s decision not to prepare an EIS or EA. We cannot accept DOE’s unsupported conclusion that its final agency action that covers ten States and over a 100 million acres does not, as a matter of law, have some environmental impact. See Alaska Ctr., 189 F.3d at 859 ; The Steamboaters, 759 F.2d at 1393 . If the smaller West-wide Corridors are worthy of a PEIS, as detailed in the statement’s executive summary, then a much larger NIETC is also presumptively worthy of an EA or EIS. In any event, DOE has failed to present the documentation necessary to allow us to determine that there are no environmental impacts or that DOE' took a “hard look” at the environmental impacts.

4. The failure to undertake an environmental study is not harmless error

Finally, DOE suggests, citing 40 C.F.R. § 1500.3 , that even if we were to determine that a formal EA was required to document DOE’s “no effects” determination, DOE’s failure to do so was, at most, harmless error. As noted in Section III C, supra, following the issuance of the Supreme Court’s opinion in Sanders, 129 S.Ct. 1696 , we place the burden on petitioners to show that the failure to undertake an environmental study is not harmless error. Here, even a cursory review of petitioners’ contentions raises “substantial questions ... as to whether [the NIETCs] may cause significant degradation of some human environmental factor.” Klamath Siskiyou Wildlands, 468 F.3d at 562 (internal citation omitted). 35

For example, petitioners note that the Southwest Corridor includes the Joshua Tree National Park and the Sonoran Desert National Monument. It includes more than three million acres of national wildlife refuge as well as national parks and 57 state beaches, reserves and recreational areas. The Mid-Atlantic Corridor encompasses four national forests, over a million acres of national reserves, historic properties, and environmentally sensitive lands. In light of the agencies’ sensitivity to envi *1106 ronmental impacts in their creation of the West-wide Corridors, we cannot conclude that the DOE’s failure to undertake a study of the NIETCs’ environmental impacts constitutes harmless error.

In sum, NEPA requires that for all “major Federal actions significantly affecting the quality of the human environment” an agency must prepare a detailed statement on the environmental impact of the action and any adverse environmental effects. 42 U.S.C. § 4332 (c). DOE did not prepare an EIS or even an EA for its NIETC Designation. Its proffered reasons for not doing so — the NIETC Designation is not a major Federal action, NEPA review will take place in subsequent requests for specific sitings, and there are no significant impacts from the Designation — are not persuasive as a matter of law and are not supported by the record. Accordingly, because DOE has not shown that it has taken the requisite “hard look” at the environmental consequences of the NIETCs, we vacate the NIETC Designation and remand the matter to DOE to prepare at least an EA to determine whether there are any environmental impacts that significantly affect the quality of human environment, and whether, if so, the impacts warrant adjustments. 36

VI. THE ENDANGERED SPECIES ACT AND THE NATIONAL HISTORIC PRESERVATION ACT

Petitioners also argue that DOE violated the Endangered Species Act (“ESA”), 16 U.S.C. § 1531 , by failing to consult with the Secretary of Interior pursuant to 16 U.S.C. § 1536 (a)(2). We have recently stated that “[t]he threshold for triggering the Endangered Species Act is relatively low; consultation is required whenever a federal action ‘may affect listed species or critical habitat.’ ” Cal. ex rel. Lockyer v. USDA, 575 F.3d 999, 1018 (2009) (quoting 50 C.F.R. § 402.14 (a)) (emphasis added). DOE responds that petitioners’ arguments concerning the ESA “echo their NEPA arguments.”

Petitioners also contend that DOE violated the National Historical Preservation Act (“NHPA”) by failing to comply with 16 U.S.C. § 470f, which requires that it accept comments from certain entities, including the Advisory Council on Historic Preservation (“ACHP”), prior to the approval of the expenditure of any Federal funds on an undertaking that has the potential to adversely affect historic properties. See 36 C.F.R. §§ 800.1 (c); 800.3(a); 800.16(y). DOE argues that it reasonably declined to initiate historic-preservation review under NHPA, and further claims that it adequately responded to the two letters it received from the ACHP.

As we hold that the Congestion Study and the NIETCs Designation must be vacated and the matter remanded to the DOE, we need not consider petitioners’ claims under the ESA and NHPA. Should DOE on remand designate NIETCs in a manner that petitioners believe violates either the ESA or NHPA, they can then seek judicial review of those decisions.

VII. CHALLENGES TO SPECIFIC ASPECTS OF THE NIETCs

Petitioners have raised numerous challenges to particular aspects of the Mid-Atlantic Corridor and the Southwest Corridor. However, as the Designation of these corridors is vacated, these challenges are moot and we need not address them. Petitioners will have the opportunity to *1107 present their concerns to DOE in the proceedings on remand. We are confident that whatever actions DOE takes, the subsequent challenges to those actions (if any) will turn, at least in part, on facts and arguments that are not before us now.

CONCLUSION

In the Energy Policy Act of 2005, Congress specifically directed DOE to undertake a Congestion Study “in consultation with affected States.” 16 U.S.C. § 824p(a)(l). It further directed that in undertaking this study and in designating any national interest electric transmission corridors, DOE was to comply with NEPA. We determine that DOE failed to consult with the affected States prior to issuing its Congestion Study and that this failure was prejudicial to the States. Accordingly, the Congestion Study is vacated. We further find that DOE’s statement that its designation of NIETCs “does not significantly affect the quality of the human environment” is not supported by sufficient evidence to show that DOE has taken the requisite “hard look” at the environmental consequences. See Kleppe v. Sierra Club, 427 U.S. at 410 n. 21, 96 S.Ct. 2718 ; California ex rel. Lockyer, 575 F.3d at 1012 . We further find that the record does not allow us to conclude that the failure to take a hard look at the environmental consequences was harmless. Accordingly, the Designation of the NIETCs is vacated. In light of our vacation of the Congestion Study and the NIETCs Designation, we decline to consider the petitioners’ challenges (1) under the Endangered Species Act, (2) under the National Historic Preservation Act, and (3) to specific aspects of the Mid-Atlantic Corridor and the Southwest Corridor.

The petitions for review are GRANTED, the Congestion Study and Designation of NIETCs are VACATED, and the matter is REMANDED for further proceedings consistent with this opinion.

. Section 824p(b) reads in part:

. The February 2 Notice set forth eight draft criteria for identifying NIET Corridors: (1) "action is needed to maintain high reliability’'; (2) "action is needed to achieve economic benefits for consumers”; (3) "actions are needed to ease electricity supply limitations in end markets served by a corridor, and diversify sources”; (4) "targeted actions in the area would enhance the energy independence of the United States"; (5) "targeted actions in the area would further national energy policy”; (6) "targeted actions in the area are needed to enhance the reliability of electricity supplies to critical loads and facilities and reduce vulnerability of such critical loads or the electricity infrastructure to natural disasters or malicious acts”; (7) "the area’s projected need (or needs) is not unduly contingent on uncertainties associated with analytic assumptions, e.g. assumptions about future prices for generation fuels, demand growth in load centers, the location of new generation facilities, or the cost of new generational technologies”; and (8) "the alternative means of mitigating the need in question have been addressed sufficiently.” 71 Fed.Reg. at 5662.

. According to DOE, the invitation-only meeting had three purposes:

. DOE reasoned as follows:

72 Fed.Reg. 25,851.

. The parties note a single change in coverage; Clark County, Nevada, was removed from the Southwest Corridor. See 72 Fed. Reg. at 25,923; 72 Fed.Reg. 57,017-18.

. Section 706 reads:

In making the foregoing determinations, the court shall review the whole record or those parts of it cited by a party, and due account shall be taken of the rule of prejudicial error.

. After it had completed the Congestion Study, DOE did invite the governors to consult. 72 Fed.Reg. at 56,996 n. 18. (“The Department sent a letter to the Governor of each of the States within the draft National Corridors and the Mayor of the District of Columbia on April 26, 2007, requesting an opportunity to consult with them on the draft designations.”).

. NARUC wrote:

. The e-mail cited by DOE in support of meetings with New York Public Service Commission and the Florida Public Service Commission simply notes that meetings took place. There is no identification of which individuals attended the meetings or the subject of the meetings. At the cited meetings with the Connecticut Department of Public Utility Control, the New Jersey Board of Public Utilities, and the Pennsylvania Public Utility Commission, DOE was represented by Ellen Lutz. In the e-mail, she explains that she is DOE's "Representative to the Eastern Interconnect,” that this “is a new position that I began in late January,” and that she is "in the process of coming up to speed on the issues.” Since at the time of this e-mail, March 8, 2006, the Congestion Study was nearly complete, it seems unlikely that the meetings provided any real opportunity for consultation. Also, the document cited by DOE to support its contention that it met with the CPUC indicates only that CPUC submitted further comments, and does not state that any meeting was held.

. The Congestion Study’s Executive Summary states that it is DOE’s "first congestion study in response” to the EPAct, and that it is based "on examination of historical studies of transmission conditions, existing studies of transmission expansion needs, and unprecedented region-wide modeling of both the Eastern and Western Interconnections.” (Emphasis added).

. DOE's intimation that it disclosed "the technical studies and data on which it relied — in the February 2006 Notice of Inquiry” is misleading. A review of the Notice of Inquiry discloses that it lists, by name or title only, over 50 documents. 71 Fed.Reg. at 5663-64. It is not possible to glean the substance, let alone the particulars, of the modeling studies from the listed titles.

. The court explained:

524 F.3d at 236 (citations omitted). The court further noted that:

*1090 Id. at 236-37 (citation omitted). The court further observed that it was "a fairly obvious proposition that studies upon which an agency relies in promulgating a rule must be made available during the rulemaking in order to afford interested persons meaningful notice and an opportunity for comment.” Id. at 237 .

. DOE states that it "evaluated electric transmission congestion by: (1) reviewing historical data related to congestion from dozens of sources, (2) developing projections of future congestion, and (3) comparing the two.” Moreover, DOE consistently reiterates that impacts from persistent congestion "are not readily subject to empirical measurement” and that the "task of drawing boundaries around 'geographic areas’ experiencing constraints or congestion is not one that lends itself to technical precision.”

. Furthermore, we are not alone in using this standard. It was set forth by the D.C. Circuit in Braniff Airways, Inc. v. C.A.B., 379 F.2d 453, 462 (D.C.Cir.1967), when it stated that an error "does not mechanically compel reversal 'when a mistake of the administrative body is one that clearly had no bearing on the procedure used or the substance of decision reached.” Massachusetts Trustees of Eastern Gas and Fuel Associates v. United States, 377 U.S. 235, 248 , 84 S.Ct. 1236 , 12 L.Ed.2d 268 (1964).’’ See also Silverton Snowmobile Club v. U.S. Forest Service, 433 F.3d 772 (10th Cir.2006) (noting that harmless error may be employed only when a mistake of the administrative body is one that clearly had no bearing on the procedure used or the substance of the decision reached) (internal quotation marks and citations omitted); and Conservation Law Foundation v. Evans, 360 F.3d 21, 29 (1st Cir.2004) (noting that the omission of a formal public comment before it "clearly had no bearing on the procedure used or the substance of [the] decision reached”).

. We note that the law review article cited by the dissent, Craig Smith, Taking "Due Account” of the APA’s Prejudicial-Error Rule, 96 Va. L.Rev. 1727 (2010), is in accord. It opines that Sanders "declared something already widely understood: the burden of demonstrating harm is borne by the parties challenging agencies' decisions,” but "left unexplored the interesting and important question of how parties can persuade a court that an error was prejudicial.” Id. at 1728 .

. We noted in Riverbend Farms:

Riverbend Farms, 958 F.2d at 1482 .

. The dissent cites Conservation Law Found, v. Evans, 360 F.3d 21, 29-30 (1st Cir.2004); Texas v. Lyng, 868 F.2d 795, 800 (5th Cir.1989); Miami-Dade Cnty. v. EPA, 529 F.3d 1049, 1061 (11th Cir.2008); and Owner-Operator Indep. Drivers Ass'n, Inc. v. Fed. Motor Carrier Safety Admin., 494 F.3d 188, 202-03 (D.C.Cir.2007).

. In Sanders , the Court noted that it had “previously made clear that courts may sometimes make empirically based generalizations about what kinds of error are likely, as a factual matter, to prove harmful.’’ 129 S.Ct. at 1707. A reasonable argument can be made that a failure to consult prior to making a discretionary decision, when such consultation is mandated by law, is the type of error that is likely to prove harmful. However, in light of the affected States’ showing of prejudice we need not consider such an argument.

. The May 7, 2007 Notice commented that DOE "recognizes that FPA section 216(a) adopted a novel approach to addressing the need for new transmission infrastructure, an approach that poses challenges to all stakeholders as we collectively work to address this problem." 72 Fed.Reg. at 25,845.

. The dissent claims that DOE participated in conference calls from state entities and "met and corresponded with” commissions from five states. It bases this assertion on DOE's lists of its contacts with states. However, there is no indication that these provided any real opportunity for consultation. Less than half of the contacts listed at 72 Fed.Reg. 25,850 n. 35 were actually with state officials rather than various organizations that may have included state officials. More telling, a review of the list compiled by DOE of "outreach meetings held regarding the Congestion Study,” reveals that only two of the six1y-two meetings were with officials of a state.

. The dissent’s failure to appreciate the practical impact of DOE’s failure to consult with the affected States renders the Congressional directive to consult with the affected States unenforceable. Under the dissent’s approach, an agency could refuse to consult in its decisionmaking process, but no State could show "harm” because the agency would argue that it allowed for comments after it rendered its decision and it declined to adjust its decision on the basis of those comments. Because no State could show that the agency would have made a different decision if it engaged in consultation, no State could shoulder the burden of showing prejudice. Indeed, this is precisely the argument that the dissent makes in rejecting the Eastern States claim that DOE had "incorrect and flawed documentation.” The dissent accepts as dis-positive DOE’s post-Congestion Study explanation that the discrepancies and reporting errors "did not affect the analysis and findings of the Congestion Study.” Dissent at 1114.

. Section 4332(2)(C) provides that:

. See 72 Fed.Reg. 57,022; 73 Fed.Reg. 12,-968.

. For these reasons, as elaborated in the following sections, although recognizing that this is a close case, we disagree with the dissent's perspective that the Designation Order constitutes an adequate EA (dissent at 1117).

. Among the other potential effects warranting review under the NEPA were the fact that "by defining the federal base system and ‘new large single loads’ the contracts help determine the magnitude of BPA power obligations in the future and thus will have an impact upon long-range regional energy plans” and that "the contracts significantly affect the environment because they involve important policy choices affecting energy conservation.” 743 F.2d at 682 .

. In Friends of Yosemite, we recognized that: [A]n agency’s planning and management decisions may occur at two distinct administrative levels: (1) the "programmatic level” at which the [agency] develops alternative management scenarios responsive to public concerns, analyzes the costs, benefits and consequences of each alternative in an [EIS], and adopts an amendable [management] plan to guide management of multiple use resources; and (2) the implementation stage during which individual site specific projects, consistent with the [management] plan, are proposed and assessed.

348 F.3d at 800 . We also recognized that an agency must prepare an EIS at each level. "An EIS for a programmatic plan (such as the CMP) must provide sufficient detail to foster informed decision-making, but site-specific impacts need not be fully evaluated until a critical decision has been made to act on site development.” Id. (internal quotation marks and citations omitted).

. We noted that a “project may have significant environmental impacts where its effects are highly uncertain or involve unique or unknown risks” and “warned that general statements about possible’ effects and 'some risk’ do not constitute a ‘hard look’ absent a justification regarding why more definitive information could not be provided.” Blue Mountain, 161 F.3d at 1213 (quotation marks and citations omitted).

. We further noted that an "EIS is not necessary where a proposed federal action would not change the status quo.” 136 F.3d at 668 . Here there is no question that the NIETCs change the status quo.

. We noted:

136 F.3d at 670 .

. The district court had held that the program "was not a final agency action and, thus, not subject to review.” 136 F.3d at 668 . However, in determining whether the program had a significant environmental impact, we implicitly rejected this determination. If the program had not been final, then it would not have been subject to judicial review under the APA and plaintiffs' complaint would have been dismissed.

. DOE cites 40 C.F.R. § 1508.18 (b), which reads:

Id. DOE appears to recognize that NIETCs may be federal actions under subsection (b)(3).

. This argument exposes a weakness in DOE’s position. In essence, DOE here reasons that because people see the Designation as threatening conservation efforts, they will redouble their efforts to “protect sensitive resources.”

. DOE's argument that the Designation "is not a 'programmatic' decision” because it "is not a plan to guide land management or energy policy decisions” and "merely makes available a federal procedural remedy” does not withstand scrutiny. This argument assumes that making available a federal "procedural” remedy is not a plan to guide land management or energy policy. Here, the converse is true. The federal remedy was created as part of an energy plan. Moreover, the Designation is more than a "procedural” remedy because it also creates authority for federal action (by FERC) where no such authority previously existed, and arms that authority with the power of eminent domain.

. The executive summary goes on to state:

. We note that DOE’s reference to a part of the regulation is no substitute for its obligation to supply "a convincing statement of reasons why potential effects are insignificant.” The Steamboaters, 759 F.2d at 1393 . The regulation does state, as DOE notes, that "any trivial violation of [CEQ] regulations [does] not give rise to any independent cause of action.” However it also provides that the "provisions of the Act and of these regulations must be read together as a whole in order to comply with the spirit and letter of the law.” 40 C.F.R. § 1500.3 . Here, the failure to undertake any environmental review unsupported by a record that factually supports such a decision is not a "trivial” violation.

. In Northwest Environmental Advocates v. EPA, we noted that when the EPA acted outside its authority and failed to follow Congress’s clear mandate, the appropriate remedy was to vacate that action. 537 F.3d 1006, 1026-27 (9th Cir.2008).