United States Court of Appeals For the First Circuit

No. 24-1138 NATIONAL TRUST FOR HISTORIC PRESERVATION; FRIENDS OF THE FRANK J. WOOD BRIDGE, Plaintiffs, Appellants, HISTORIC BRIDGE FOUNDATION; WATERFRONT MAINE, BRUNSWICK, LLC, Plaintiffs, v. PETER PAUL MONTGOMERY BUTTIGIEG, in his official capacity as Secretary of the United States Department of Transportation; SHAILEN BHATT, in his official capacity as Administrator, Federal Highway Administration; TODD JORGENSEN, in his official capacity as Administrator, Federal Highway Administration Maine Division; BRUCE VAN NOTE, in his official capacity as Commissioner, Maine Department of Transportation, Defendants, Appellees.

APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF MAINE [Hon. Lance E. Walker, U.S. District Judge]

Before Gelpí, Kayatta, and Aframe, Circuit Judges.

Christopher M. Cody, with whom Elizabeth S. Merritt and Andrea C. Ferster were on brief, for appellants.

John Emad Arbab, Attorney, Environment and Natural Resources Division, U.S. Department of Justice, with whom Todd Kim, Assistant Attorney General, Environment and Natural Resources Division, U.S. Department of Justice, Joan M. Pepin, Attorney, Environment and Natural Resources Division, U.S. Department of Justice, and Silvio J. Morales, Attorney, U.S. Department of Transportation, were on brief, for appellees Peter Paul Montgomery Buttigieg, Shailen Bhatt, and Todd Jorgensen.

Thomas A. Knowlton, Deputy Attorney General of Maine, with whom Aaron M. Frey, Attorney General of Maine, and Anne M. Pare, Chief Counsel, Maine Department of Transportation, were on brief, for appellee Bruce Van Note.

January 10, 2025 KAYATTA, Circuit Judge. We consider for a fourth time the fate of the Frank J. Wood Bridge ("the Bridge"), a key piece of Maine's infrastructure connecting the municipalities of Topsham and Brunswick. Built in 1932, the Bridge is no longer capable of safely performing its needed role, and traffic restrictions and eventual closure will likely result absent substantial rehabilitation or replacement.

Maine's Department of Transportation (MDOT) concluded that replacement makes more sense than rehabilitation. Because federal funds will be used to replace the Bridge, MDOT was therefore required to obtain the approval of the Federal Highway Administration (FHWA) and to comply with several federal statutes, including -- because the Bridge is part of a historic district and is itself historic -- section 4(f) of the Department of Transportation (DOT) Act. See DOT Act, Pub. L. No. 89-670, § 4(f), 80 Stat. 931, 932 (1966), codified as amended at 49 U.S.C. § 303(c) (limiting approvals of projects which "use" a protected property to situations where there is "no prudent and feasible alternative"); 23 C.F.R. § 774.17 (2024) (providing that an alternative is not "prudent" under section 4(f) if it "results in additional construction, maintenance, or operational costs of an extraordinary magnitude"); see also 54 U.S.C. § 306108 (requiring federal agencies to "take into account the effect of [any] undertaking on any historic property"). Over the objections of

- 3 - pro-preservation groups -- including the plaintiffs prosecuting this appeal -- the FHWA granted MDOT approval under section 4(f) to replace the Bridge.

A petition for review to the district court followed, Hist. Bridge Found. v. Chao, 517 F. Supp. 3d 9 (D. Me. 2021), in which plaintiffs and others presented "a slew of line-item challenges to various [agency] calculations," Hist. Bridge Found. v. Buttigieg, 22 F.4th 275, 281 (1st Cir. 2022). The district court rejected all of plaintiffs' challenges. Hist. Bridge Found. v. Chao, 517 F. Supp. 3d at 32.

Plaintiffs' first appeal to this court followed. We rejected all of plaintiffs' challenges, save one. Hist. Bridge Found. v. Buttigieg, 22 F.4th at 282–86. The surviving challenge pertained to the manner in which MDOT compared the costs of replacement and rehabilitation in concluding that rehabilitation would represent a cost of "extraordinary magnitude" under 23 C.F.R. § 774.17 (2024), so as to render rehabilitation not "prudent" under section 4(f). Id. at 284. As explained in more detail in our opinion on that prior appeal, MDOT concluded that the rehabilitation alternative would cost over double the cost of replacement, for a shorter service life. Id. at 279. In so concluding, MDOT did not rely on the differential that would result from discounting future costs to present value -- although it calculated that differential, roughly 53%, in a calculation known

- 4 - as life-cycle cost analysis. Id. at 284–86. Nor did the agency explain why it did not rely on discounted costs, apart from "rel[ying] on non-discounted future costs as the better basis upon which to compare the alternatives," id. at 279, despite evidence that the DOT and the FHWA's own guidelines identify discounting as "the standard and preferred way to compare future costs," id. at 284–85 (citing, among others, Off. of Asset Mgmt., FHWA, Pub.

No. FHWA IF-02-047, Life-Cycle Cost Analysis Primer 9 (2002)). We therefore vacated the FHWA decision in part and remanded the case "for the strictly limited purpose of allowing the agency to further justify use of the service-life analysis and/or to decide whether a 53% price differential represents a cost of an extraordinary magnitude under 23 C.F.R. § 774.17." Id. at 286.

The district court then did as we ordered, remanding the case to the FHWA with identical instructions. On remand, the FHWA also did as we ordered, training its attention on deciding whether a 53% price differential between new construction and rehabilitation represented a cost of an extraordinary magnitude.

The FHWA concluded that the 53% differential did constitute a cost of an extraordinary magnitude.

Rather than contesting that finding, plaintiffs sought to reframe the inquiry by pointing to the fact that MDOT had recently increased its own estimate of the cost of building a new bridge from $13 million to $42 million. That increase in

- 5 - construction costs, plaintiffs contended, eliminated -- indeed reversed -- the 53% differential. And to parry any assumption that the cost of rehabilitating the bridge had similarly escalated, plaintiffs submitted calculations of engineer Robert Shulock, who "updated" MDOT's 2016 cost estimate for rehabilitation in light of inflation. According to Shulock and plaintiffs, the updated cost of rehabilitation using life-cycle cost analysis would be approximately $25 million, a roughly $18-million savings from MDOT's updated estimate for reconstruction.

In its final Section 4(f) Limited Scope Re-Evaluation ("Re-Evaluation"), the FHWA did not accept plaintiffs' invitation to conduct its own calculation of the updated costs of each alternative; nor did it recalculate the differential between rehabilitation and reconstruction. It did "acknowledge[]" that its own cost estimates for replacing the Bridge had increased from $13 million in 2019 to $42 million in 2022. But it pointed to our mandate that remand was to be "strictly limited to either justifying its prior use of non-discounted costs, and/or deciding whether a 53% price differential represented a cost of extraordinary magnitude." Moreover, relying on an internal evaluation of Shulock's calculations, the agency stated that it found Shulock's "updated" rehabilitation costs to be "flaw[ed]." "Among other things," the agency opined that:

- 6 - • Shulock's inflation rate of 23% was too low, as compared to steel component rates of 66–90% revealed by producer price indices and a general increase in construction costs of approximately 40%, all of which indicated that "the ratio of increased costs reflected in the recent work plan estimates for the replacement alternative would apply at least equally to [the] rehabilitation [a]lternative." • Shulock's estimate failed to include major items that were included in the replacement estimate, such as "local amenities, utility work, and full right-of-way costs." • Shulock underestimated the costs of "painting, site access/trestle, removal of existing slab, and construction of the temporary bridge" that would be required to rehabilitate the Bridge. • Shulock overlooked the difference in "long-term maintenance costs [which is] what truly separate[s] the two alternatives." • Shulock relied on a bridge rehabilitation project in New Hampshire as evidence of costs but failed to account for the fact that its square footage was

- 7 - one-third that of the Bridge and its scope was not as extensive.

The agency concluded that it was "not persuaded that the effects of inflation and other cost considerations presented in public comments materially change the conclusions reached in its draft [section] 4(f) Re-Evaluation." It did explain that "[i]n addition to dollar values, there is added risk and uncertainty to keeping a fracture critical bridge in service beyond its intended service life, even if it is extensively rehabilitated."1 This additional risk is not "fully capture[d]" by cost estimates because "there is no consistent way to continuously monitor for defects." And "[i]f a fracture critical component were to fail, it could cause a bridge to collapse." Regular inspections would be necessary to reduce this risk, and those would be time-intensive and highly disruptive.

I.

In February 2023, plaintiffs again sued in United States district court, alleging that it was arbitrary and capricious under the Administrative Procedure Act (APA) for the FHWA not to have considered the updated cost information in MDOT's own

- 8 - possession -- and that if it had, it would have found that rehabilitation is a prudent and feasible alternative.2 After fielding plaintiffs' varied efforts at enjoining construction on the project,3 the district court granted summary judgment for the

After the district court denied their motion, plaintiffs appealed to this court and sought an emergency motion for an injunction pending an appeal. We denied plaintiffs' emergency motion for an injunction pending appeal without prejudice for non-compliance with Federal Rule of Appellate Procedure 8(a), instructing that plaintiffs should first seek an injunction pending appeal from the district court. In response to this court's order, plaintiffs filed an emergency motion for an injunction pending appeal in the district court. They also requested a temporary restraining order to halt work on bridge reconstruction, and in the case of a denial, a stay of construction activities pending appeal. The district court denied both requests; plaintiffs then renewed their emergency motion for an injunction pending appeal with us. We denied this last request by stating, "[o]ur strictly limited remand did not provide the occasion for a new foray by the plaintiffs and, in any event, the new foray now attempted does not look especially promising, even were its pursuit otherwise appropriate notwithstanding our strictly limited remand."

- 9 - agencies. Nat'l Tr. for Hist. Pres. v. Buttigieg, No. 2:23-cv- 00080, 2024 WL 69855, at *1–2 (D. Me. Jan. 5, 2024). In relevant part, the district court found that the FHWA "was required to use [MDOT's] 2016 cost estimates because those estimates were before the agency 'at the time of the agency action.'" Id. at *9 (quoting Dep't of Homeland Sec. v. Regents of the U. of Cal., 591 U.S. 1, 20–21 (2020)). The district court also reasoned that the agency "complied with [this court's] remand instructions," which specifically instructed the FHWA to either justify its use of non-discounted figures or consider whether a 53% price differential was an extraordinary cost. Id. Because the district court found that the agencies properly used the 2016 cost estimates, it found it "unnecessary to consider [p]laintiffs' arguments concerning whether the rehabilitation alternatives would be a prudent and feasible alternative if 2022 cost estimates were used." Id. at *9 n.7. Plaintiffs appealed.

II.

"Agency determinations under . . . section 4(f) are reviewed under the [APA] and accordingly shall not be overturned unless arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." Hist. Bridge Found. v. Buttigieg, 22 F.4th at 282 (cleaned up); see 5 U.S.C. § 706(2)(A). Under this standard, a reviewing court should "determine whether the agency has examined the pertinent evidence [and] considered the relevant

- 10 - factors." Airport Impact Relief, Inc. v. Wykle, 192 F.3d 197, 202 (1st Cir. 1999) (quoting Penobscot Air Servs., Ltd. v. FAA, 164 F.3d 713, 719 (1st Cir. 1999)). "The agency must explain the evidence which is available, and must offer a 'rational connection between the facts found and the choice made.'" Motor Vehicle Mfrs.

Ass'n of U.S. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 52 (1983) (quoting Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168 (1962)). We review a district court's decision under the APA de novo. Hist. Bridge Found. v. Buttigieg, 22 F.4th at 282.

III.

On appeal, plaintiffs do not contest that a 53% differential is a cost of extraordinary magnitude. Instead, as they argued before the agency and in the district court, they contend that the FHWA should not have so limited its inquiry on remand. Rather, say plaintiffs, the FHWA should have reassessed the relative costs of the alternatives in view of a significant increase in construction costs during the six years that had passed since the record before the FHWA was first compiled. And they contend that the FHWA's decision to continue to assume that rehabilitation would cost roughly 53% more than new construction was arbitrary and capricious.

As the district court concluded, our mandate in the prior appeal provides the simple answer to this argument. There, the

- 11 - issue before us reflected the agency's 2019 comparison of 2016 cost estimates of the principal rehabilitation and new construction alternatives. Hist. Bridge Found. v. Buttigieg, 22 F.4th at 284. Without discounting to present value, those costs generated a roughly 103% price differential, id. at 279, 281, and it was that differential that the agency found to be a cost of extraordinary magnitude. But discounting costs to present value generated only a roughly 53% differential. Thus, we issued a clear mandate: The remand proceeding was to be "strictly limited" to either justifying the decision not to use discounted costs or to opining on whether that differential was a cost of extraordinary magnitude. Id. at 286. Plaintiffs now resist this straightforward reading by arguing that our mandate allowed the agency to decide much more -- i.e., to "deal with the problem afresh." See SEC v. Chenery Corp. (Chenery II), 332 U.S. 194, 201 (1947). But this reading of our mandate makes no sense. An order that a remand proceeding be "strictly limited" to deciding an explicitly identified and narrow issue does not ordinarily grant license to consider other issues "afresh." Nevertheless, say plaintiffs, the agency inherently retained the ability to reopen the proceeding to reconsider its prior conclusions based on new evidence. As plaintiffs point out, the Supreme Court has noted the possibility that, upon remand for further justification of its decision, an "agency can 'deal with

- 12 - the problem afresh' by taking new agency action." Regents of the U. of Cal., 591 U.S. at 20–21 (2020) (citing Chenery II, 332 U.S. at 201). Whether it can do so when the remand mandate is limited as it is here, we need not decide. Rather, we will assume without deciding that the agency did have the discretion to reopen the proceeding.

Primed with that assumption, plaintiffs' argument proceeds in two alternative forms. First, and principally, they argue that the agency did in fact reopen the record to consider new cost information by responding to Shulock's estimates, and therefore was required to recalculate the differential between reconstruction and rehabilitation. Second, and much less clearly, plaintiffs seem to suggest that, if the agency did not reopen the proceeding to consider updated cost information, it acted arbitrarily and capriciously. Because we reject both arguments, we need not reach plaintiffs' contention that, if the agency had recalculated the differential, it would have necessarily found that rehabilitation was a prudent and feasible alternative.4

- 13 - A.

Plaintiffs argue that, by considering and rebutting Shulock's cost estimates, the FHWA reopened the record -- in other words, it took into account new cost information that was not in the record at the time of our remand. See Pub. Emps. for Env't Resp. v. Hopper, 827 F.3d 1077, 1089–90 (D.C. Cir. 2016) (finding that the Fish and Wildlife Service reopened the record by choosing to rely on new, post-remand analysis proffered by its in-house economist). And, plaintiffs say, because the agency considered new costs, it was required to take the next step and recalculate the differential. See id. at 1090 (holding that it was arbitrary and capricious for the Fish and Wildlife Service to "disregard" new, post-remand data submitted by plaintiffs where it had otherwise reopened the record).

EPA, 540 F.2d 1023, 1034 (10th Cir. 1976) (upholding an EPA regulation based on the record before the EPA at the time it took action and finding that "data obtained after promulgation" strengthened that conclusion by showing "the validity of the EPA actions"); Amoco Oil Co. v. EPA, 501 F.2d 722, 729 n.10 (D.C. Cir. 1974) (while reviewing an EPA regulation, supplementing the record with post-promulgation testimony before a congressional committee because the testimony "b[ore] directly upon the plausibility of certain predictions made by the [EPA] Administrator in promulgating the Regulations"). Here, MDOT awarded a construction contract for $50 million after the Re-Evaluation. See Nat'l Tr., 2024 WL 69855, at *4. But at the time of the Re-Evaluation, MDOT had already increased its estimate for the project to upwards of $48 million. Thus, supplementing the record with the price of the construction contract would not significantly alter the analysis, much less show that the agency's section 4(f) determination was "wrong."

- 14 - We disagree. As plaintiffs point out on reply, the agency did engage an in-house expert to review Shulock's estimates.

But we view that engagement with Shulock's estimates as a belt-and-suspenders explanation for why the agency was not reopening the proceeding to take into account new costs.

Plaintiffs' entire rationale for reopening was predicated on a contention that the rehabilitation costs had not risen commensurately with new construction costs. That contention, in turn, rested on Shulock's submission. So by explaining why it found that submission to be seriously flawed, the FHWA effectively explained why plaintiffs had not shown cause to reopen the proceeding to consider new costs even were it an option to do so. Hence, the premise of plaintiffs' principal argument -- that the agency reopened the record to consider new cost information -- is simply not correct.

B.

Largely latent in plaintiffs' briefs is an alternative argument that the agency acted arbitrarily and capriciously by failing to reopen the record to consider updated cost information.

We disagree. The fact that, in general, the costs of construction rose between 2016 and 2022 is hardly surprising. So MDOT's new 2022 construction estimate provided no cause for questioning its own comparison between construction and rehabilitation alternatives unless one thought that the costs of rehabilitating

- 15 - the bridge did not also rise commensurately. Presumably for this reason, plaintiffs proffered Shulock's cost estimate to show that there was cause to expect significantly different increases in the relative costs. The FHWA, however, was unimpressed with that submission. And it explained why. Not only did Shulock omit certain construction, right-of-way, and maintenance costs from his rehabilitation estimate, but he also used an apparently very low inflation rate and relied on lower out-of-state bid prices in calculating his estimate. These perceived flaws, among others, are like the thirteenth chime of the clock -- they not only seemed wrong, but they also cast doubt on all of Shulock's reasoning that preceded them.

Plaintiffs' counsel seeks to parry this line of reasoning by attempting to quantify the FHWA's stated criticisms of Shulock's analysis. But much of this attempted quantification was not in the record at the time of the Re-Evaluation, and thus is not appropriate for our review. See Chenery II, 332 U.S. at 196 ("[A] reviewing court . . . must judge the propriety of [agency] action solely by the grounds invoked by the agency.").

Moreover, counsel's attempts do not address all of the agency's concerns -- notably, counsel's attempts do not seem to use an inflation rate that moots the agency's breakdown of greater inflation rates for specific components, nor do they meet the thrust of the agency's observation that a key difference between

- 16 - rehabilitation and reconstruction was the cost and challenge of maintenance. As a result, we find that the record supports the agency in its rejection of Shulock's estimates as compelling reasons to expand the scope of the remand proceeding.

In sum, the FHWA was acting in response to a "strictly limited" instruction from this court. Hist. Bridge Found. v. Buttigieg, 22 F.4th at 286. By deciding whether a 53% price differential was a cost of extraordinary magnitude, the agency did precisely what we told it to do -- plaintiffs hardly argue otherwise. And on this record we cannot say that its failure to recalculate the differential between rehabilitation and reconstruction using updated cost information was arbitrary and capricious.

Our disposition of this appeal finds support in contemplating the ramifications of a contrary holding. Many months and years can pass between the end of an agency proceeding and the end of judicial challenges to such a proceeding. Furthermore, multi-issue appeals raising a slew of challenges are not unusual.

Were we to adopt plaintiffs' view that agencies on remand can be compelled, on a record like this one, to reopen their proceedings and "update" their assessments, finality and repose would diminish as proceedings might never end.

IV.

For the foregoing reasons, we affirm.

- 17 -