OPINION
This case arises from a longstanding dispute between the National Highway Traffic Safety Administration (NHTSA) and the National Truck Equipment Association (NTEA). NHTSA is a federal agency within the Department of Transportation that writes and enforces safety standards for motor vehicles. NTEA is a trade organization representing more than one thousand manufacturers who customize bodies for special-purpose commercial vehicles. In 2005, NHTSA initiated a rulemaking proceeding at Congress’s behest to upgrade the safety standard establishing strength requirements for passenger compartment roofs in certain vehicles. NHTSA proposed, among other things, extending the scope of the safety standard to include a previously unregulated class of vehicles, many of which are produced by NTEA’s members. NTEA resisted this proposal throughout the process, but to no avail. In 2009, NHTSA promulgated Federal Motor Vehicle Safety Standard (FMVSS) No. 216a, which NTEA now challenges on several grounds. For the reasons that follow, we deny the petition for review.
I.
A.
Congress passed the National Traffic and Motor Vehicle Safety Act of 1966 (Safety Act) with the express purpose of “reducing] traffic accidents and deaths and injuries to persons resulting from traffic accidents.” Pub. L. No. 89-563, 80 Stat. 718 (codified at 49 U.S.C. §§ 30101-30183 (2010)). Toward this end, Congress directed the Secretary of Transportation to prescribe safety standards for motor vehicles, see 49 U.S.C. § 30111(a), a statutory duty the Secretary has delegated to NHTSA since 1980, see 49 C.F.R. § 501.2 (2010). NHTSA’s standards serve as mandatory performance benchmarks for motor vehicles and their constituent parts. See 49 U.S.C. § 30112(a)(1). As such, manufacturers must certify that the vehicles *664they produce comply with each standard that applies. See id. § 30115.
While most of NHTSA’s standards are aimed at single-stage manufacturers of non-commercial vehicles — the kind most of us drive everyday' — -this group is not the only one subject to regulation. The Safety Act also applies to final-stage manufacturers — who complete vehicles assembled in two or more stages — and alterers — who make modifications to already-completed vehicles prior to sale. See id. § 30115(b). NTEA challenges FMVSS No. 216a only as it relates to the latter two groups.
A little background is in order here. Final-stage manufacturers are for the most part small companies that supply a market for custom-made work trucks driven by end-users with specialized needs, like contractors and utility companies. See generally Nat’l Truck Equip. Ass’n v. Nat’l Highway Traffic Safety Admin., 919 F.2d 1148, 1150-51 (6th Cir.1990) (describing the market for multi-stage vehicles in considerable detail). A final-stage manufacturer filling an order usually begins with a truck chassis produced by a major manufacturer such as Ford or General Motors (GM). Oftentimes the chassis will consist, in its entirety, of an engine, transmission, axles, wheels, and a completed passenger compartment with bare frame rails in the rear (where the bed would be on a non-commercial truck). This form of chassis is known as a “chassis-cab,” see 49 C.F.R. § 567.3, and it is functionally a work-in-progress. A final-stage manufacturer then attaches a service body or other specialized equipment to complete the vehicle and meet the end-user’s specific needs. One familiar example might be adding a flat bed and a winch to complete a tow truck.
Like other manufacturers, final-stage manufacturers are required to certify that the vehicles they complete are in compliance with all applicable safety standards. 49 U.S.C. § 30115(a)-(b). The Safety Act provides two avenues to do this. The most straightforward avenue is to independently certify compliance. See id. § 30115(b)(2). Taking this route generally entails conducting whatever test is specified in the relevant standard — e.g., crashing a vehicle into a wall. Because the cost of conducting such tests is almost always insurmountable for small companies, this is not a popular option among final-stage manufacturers. Thankfully Congress anticipated this problem and included a second avenue for final-stage manufacturers to certify compliance that amounts to piggybacking. See id. § 30115(b)(1). Final-stage manufacturers may rely on an initial manufacturer’s certification statement that the incomplete vehicle it delivered conforms to applicable standards — in other words, they are allowed to “pass through” the initial certification rather than certifying independently. Id. Both the initial manufacturer and the final-stage manufacturer of a vehicle have some responsibility in this scenario.
Pass-through certification is at the heart of the dispute between NTEA and NHTSA. This avenue is made possible because initial manufacturers of chassis-cabs are required to deliver their products with an incomplete vehicle document (IVD) that lists the applicable safety standards “in effect at the time of manufacture.” 49 C.F.R. § 568.4(a)(7). For each standard, initial manufacturers must supply one of three kinds of statements regarding an incomplete vehicle’s compliance. Id. A “Type 1” statement certifies “that the vehicle when completed will conform to the standard if no alterations are made in identified components of the incomplete vehicle.” Id. § 568.4(a)(7)(i). A “Type 2” statement provides “specific conditions of final manufacture under which the manufacturer specifies that the completed vehicle will conform to the standard.” Id. *665§ 568.4(a)(7)(ii). Finally, a “Type 3” statement simply notes that “conformity with the [applicable] standard cannot be determined based upon the components supplied on the incomplete vehicle, and that the incomplete vehicle manufacturer makes no representation as to conformity with the standard.” Id. § 568.4(a)(7)(iii). These statements specify the restrictions and conditions that an initial manufacturer places on the certification of a completed vehicle built on its chassis. A final-stage manufacturer need only certify “that it has complied with the specifications set forth” in an IVD to satisfy its duty under the Safety Act. 49 U.S.C. § 30115(b)(1); see also 49 C.F.R. § 567.5(d)(1) (noting that final-stage manufacturers are in the clear if they “complete the vehicle in accordance with the prior manufacturers’ IVD”).
The Safety Act — and the attendant duty to certify — also applies to alterers. They are generally small companies that make aftermarket modifications to single-stage vehicles before those vehicles are first sold to end-users. See 49 C.F.R. § 567.3. Alterers are distinct from final-stage manufacturers insofar as vehicles arrive to alterers fully built; that is, the vehicles arrive with load-bearing and other specialized equipment already attached rather than as an incomplete chassis-cab. For this reason, alterers face a different certification regime under the Safety Act, one in which pass-through certification is irrelevant. Single-stage manufacturers are required to certify that their completed vehicles are compliant with all applicable safety standards. See 49 U.S.C. § 30115(a). Thus, there is no need to issue an IVD when delivering such vehicles to alterers or anyone else. Alterers do not run afoul of the Safety Act so long as they avoid making changes that take the completed vehicle out of conformity with any particular standard. See 49 C.F.R. § 567.7(a)(1).
B.
The one and only safety standard relevant to this petition is FMVSS No. 216a, which establishes strength requirements— known as “roof crush resistance” — for passenger compartment roofs in certain vehicles. See id. § 571.216a. The standard regulates roof pillars and roof materials to improve occupant safety in the event of a rollover. NHTSA promulgated the first iteration of this standard in 1971 as FMVSS No. 216, see Motor Vehicle Safety Standards, 36 Fed. Reg. 23,299 (Dec. 8, 1971), and amended it in 1991, see Roof Crush Resistance, 56 Fed. Reg. 15,510 (Apr. 17,1991).
In 2005, Congress directed NHTSA to revisit the roof crush resistance standard once again. See 49 U.S.C. § 30128(a), (d); see also Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users, Pub. L. No. 109-59, Title X, § 10301(a), (d), 119 Stat. 1144, 1939 (Aug. 10, 2005). The statute mandated new rule-making proceedings, including one to “establish performance criteria to upgrade Federal Motor Vehicle Safety Standard No. 216 relating to roof strength for driver and passenger sides.” 49 U.S.C. § 30128(d). This specific mandate came as part of a broader effort to promulgate new “rules or standards that will reduce vehicle rollover crashes and mitigate deaths and injuries associated with such crashes for motor vehicles with a gross vehicle weight rating of not more than 10,000 pounds.” Id. § 30128(a).
Less than a month later, NHTSA heeded Congress’s call and issued a Notice of Proposed Rulemaking (NPRM) regarding the roof crush resistance standard. 70 Fed. Reg. 49,223 (Aug. 23, 2005). The NPRM set forth three significant changes intended to upgrade existing FMVSS No. 216, one of which raised the weight limit for vehicles subject to the standard from a *666Gross Vehicle Weight Rating (GVWR) of 6,000 pounds to 10,000 pounds. Id. at 49,223-24. This particular change promised to have a large impact on many multistage and altered vehicles not subject to the existing standard. In an effort to mitigate the burden of demonstrating compliance with the new standard, NHTSA stipulated that certain multi-stage vehicles could certify to an easier standard (FMVSS No. 220) but excluded those built on chassis-cabs because their manufacturers could take advantage of pass-through certification. Id. at 49,235.
In 2009, after considering public comments from numerous groups, NHTSA modified its proposals and promulgated the final rule — designated as FMVSS No. 216a — which is the subject of this petition. See Roof Crush Resistance, Phase-In Reporting Requirements, 74 Fed. Reg. 22,348 (May 12, 2009); see also Final Rule, Correcting Amendment, 75 Fed. Reg. 17,604 (Apr. 7, 2010). NTEA draws our attention to just one aspect of the final rule: NHTSA’s decision to extend the scope of the roof crush resistance standard to include vehicles with a GVWR up to 10,000 pounds. See 74 Fed. Reg. at 22,348-49; see also 49 C.F.R. § 571.216a. At the same time that NHTSA decided to extend the scope of the standard, it made two important concessions regarding heavier vehicles vis-a-vis its position in the NPRM. First, the final rule only requires that roofs on heavier vehicles be able to withstand an amount of force equal to 1.5 times the vehicle’s weight — that is, half the strength-to-weight ratio imposed on vehicles with a GVWR of 6,000 pounds or less. See 74 Fed. Reg. at 22,348-49; see also 49 C.F.R. § 571.216a, S5.2. Second, the final rule only applies to multi-stage vehicles “built on either a chassis cab or an incomplete vehicle with a full exterior van body.” See 74 Fed. Reg. at 22,349; see also 49 C.F.R. § 571.216a, S3.1 (a)(4). NHTSA also provided plenty of lead time in the final rule. It decided to phase-in the upgraded roof crush resistance standard over a lengthy period, giving manufacturers until September 2016 to begin certifying already-regulated vehicles and an extra year for newly regulated vehicles. See 74 Fed. Reg. at 22,348; see also 49 C.F.R. § 571.216a, S8.8, S9(b).
C.
These concessions notwithstanding, one group that remains concerned about the impact of FMVSS No. 216a is NTEA, an association counting among its members many final-stage manufacturers and alter-ers who work on vehicles with GVWRs between 6,001 pounds and 10,000 pounds. These members face the daunting task of demonstrating compliance with the roof crush resistance standard for the first time. NTEA voiced their concerns throughout the rulemaking process in response to NHTSA’s many requests for comments. See, e.g., Further Response to Comments, 75 Fed. Reg. 17,590 (Apr. 7, 2010). When those concerns were not addressed to their satisfaction in the final rule, NTEA immediately filed this petition for review. We granted a motion to hold the litigation in abeyance while NHTSA responded to further NTEA comments regarding compliance with FMVSS No. 216a.
In 2010, NHTSA published its response to NTEA’s core concern: that pass-through certification did not provide a realistic avenue for final-stage manufacturers using chassis-cabs to comply with the upgraded roof crush resistance standard. See id. NTEA had requested that NHTSA completely exclude its members from FMVSS No. 216a rather than leave them subject to its requirements with no way to certify compliance. See id. at 17,-594. But NHTSA rejected both the premise and the conclusion. NHTSA noted that final-stage manufacturers had been *667using pass-through certification to successfully comply with FMVSS No. 216 and similar standards for decades. Id. at 17,-599. NHTSA thus determined that final-stage manufacturers could use this method to demonstrate compliance with the upgraded standard in FMVSS No. 216a without incurring significant new costs. Id. at 17,604. NHTSA also determined that excluding all multi-stage vehicles from FMVSS No. 216a would be inconsistent with Congress’s cardinal desire to improve occupant safety during rollovers. See 49 U.S.C. § 30128(a).
Immediately following NHTSA’s response, NTEA filed a petition for reconsideration, which NHTSA predictably denied. See Response to Petition for Reconsideration, 76 Fed. Reg. 15,903 (Mar. 22, 2011). The parties continued to trade proposals regarding amendments to FMVSS No. 216a. But a mutually agreeable solution proved elusive. When the settlement process finally broke down, this litigation recommenced.
NTEA now challenges NHTSA’s adoption of FMVSS No. 216a insofar as it extends the scope of the roof crush resistance standard to include vehicles with a GVWR up to 10,000 pounds. NTEA asserts that NHTSA’s decision is arbitrary and capricious and that the final rule fails to satisfy the minimum substantive criteria prescribed in the Safety Act. In addition, NTEA asserts that NHTSA improperly delegated its statutory authority. By contrast, NHTSA maintains that it conducted an appropriately thorough rulemaking, involving input from many groups and numerous modifications to its proposals, and ultimately promulgated a final rule that conforms to the language of the Safety Act.
II.
Before we reach the merits, we begin by identifying the appropriate standards of review. This is not always an easy task, especially in the administrative law context, but it is a critical one. Here, our review of the final rule in question is to proceed along two tracks: the first can be described as generally procedural, the other substantive.
The first track is defined by the traditional standards set forth in the Administrative Procedure Act (APA). See Chrysler Corp. v. Dep’t of Transp., 472 F.2d 659, 670 (6th Cir.1972). Section 706 requires us to uphold agency action unless it is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A). This inquiry is principally concerned with the agency decisionmaking process. It is a familiar refrain that the process is considered arbitrary and capricious if
the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.
Motor Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983). The arbitrary-and-capricious standard is a narrow one inasmuch as we are “not to substitute [our] judgment for that of the agency.” Id. Instead, our role is limited to reviewing the administrative record “to determine whether there exists a ‘rational connection between the facts found and the choice made.’ ” Alliance for Cmty. Media v. F.C.C., 529 F.3d 763, 786 (6th Cir.2008) (quoting State Farm, 463 U.S. at 43, 103 S.Ct. 2856). The facts found, of course, must be supported by substantial evidence in the record as a whole. Chrysler, 472 F.2d at 668; see also *668State Farm, 463 U.S. at 44, 103 S.Ct. 2856. Put another way, our role under the arbitrary-and-capricious standard is to ensure that NHTSA “complied with all applicable procedural requirements” in making its decisions. Chrysler, 472 F.2d at 671 (emphasis added).
The second track is defined by the statutory limitations set forth in the Safety Act itself, which is the source of NHTSA’s grant of authority to promulgate standards for newly manufactured vehicles. Importantly, the grant is conditioned upon standards that “shall be practicable, meet the need for motor vehicle safety, and be stated in objective terms.” 49 U.S.C. § 30111(a). We have characterized these factors as the “minimum substantive criteria against which each ... standard must be tested.” Chrysler, 472 F.2d at 668. Whether or not a standard passes the statutory test is an inquiry distinguishable from traditional arbitrary-and-capricious review. Id. at 671. Our focus here is principally substantive rather than procedural. Still, under the APA, a final rule that conflicts with the language of the statute relied upon by the agency is considered “not in accordance with the law.” City of Cleveland v. Ohio, 508 F.3d 827, 838 (6th Cir.2007); see 5 U.S.C. § 706(2)(A).
While both of these standards of review are premised on deference, it is important to be clear about the particular brand of deference. To that end, this case is not one in which the Chevron analytical framework controls — notwithstanding NHTSA’s assertion to the contrary. See Chevron U.S.A Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984). The Chevron framework is generally appropriate for use when an agency’s authority to act under a statute is questioned. See id. at 842-45, 104 S.Ct. 2778; see also Arent v. Shalala, 70 F.3d 610, 615 (D.C.Cir.1995). Such a dispute often manifests as competing interpretations of the statutory text demarcating the bounds of Congress’s delegation to the agency. Arent, 70 F.3d at 615. But it is hardly controversial that the Safety Act authorizes NHTSA to promulgate a safety standard regulating roof crush resistance in certain vehicles. See 49 U.S.C. § 30111. Nor do the parties dispute the proper meaning of the text of the Safety Act. What is ultimately in doubt is whether NHTSA discharged its delegated authority in a justifiable manner. And we review that question under the standards set forth in the APA and State Farm. See Arent, 70 F.3d at 615-16.
In the instant petition, we are thus presented with three broad questions for decision — some easier to answer than others. (1) Was the process by which NHTSA promulgated FMVSS No. 216a arbitrary and capricious, or did NHTSA conduct its rulemaking proceedings in a sufficiently thorough manner? (2) Setting aside the decisionmaking process, does FMVSS No. 216a fall short of the minimum substantive criteria prescribed in the Safety Act, or does it satisfy all three? (3) Finally, did NHTSA improperly delegate its statutory authority as part of FMVSS No. 216a? We answer each question in turn.
III.
The first question is essentially a procedural one. We must ask whether NHTSA conducted the rulemaking proceeding that resulted in the adoption of FMV SS No. 216a in accordance with the APA’s procedural requirements. Our review is limited by the narrow scope of the petition, concerning only the decision to regulate final-stage manufacturers and al-terers. After canvassing the record, we are satisfied that the agency decisionmak-ing process was adequate.
*669Under the APA, we have the power to set aside agency action if it is “arbitrary and capricious.” 5 U.S.C. § 706(2)(A). Here, our review turns in part on whether NHTSA amassed sufficient record evidence to support its conclusions. See State Farm, 463 U.S. at 43-44, 103 S.Ct. 2856. Our review turns in other part on whether NHTSA properly relied on the factors Congress intended it to consider when promulgating a new safety standard. See id. at 43, 103 S.Ct. 2856. These factors are clearly articulated in the Safety Act. Congress directed NHTSA to consider the “relevant available motor vehicle safety information”; “whether [the] proposed standard is reasonable, practicable, and appropriate for the particular type of motor vehicle”; and “the extent to which the standard will carry out” the purposes of the Safety Act. 49 U.S.C. § 30111(b). These factors underscore the notion that “[t]he Safety Act’s mandate is not ... categorical,” Ctr. for Auto Safety v. Peck, 751 F.2d 1336, 1343 (D.C.Cir.1985), as the Act does not contemplate the elimination of all accidents or injuries, only those that are “unreasonable.” See 49 U.S.C. § 30102(a)(8); S.Rep. No. 89-1301, at 2714 (1966), 1966 U.S.C.C.A.N. 2709, 2714 (noting that the Secretary “will necessarily consider reasonableness of cost, feasibility and adequate lead time”). Nonetheless, Congress has maintained that “safety shall be the overriding consideration” in the regulatory calculus. S.Rep. No. 89-1301, at 2714, 1966 U.S.C.C.A.N. 2709 at 2714.
Turning now to the record, it seems obvious to us that NHTSA promulgated the final rule at issue only after engaging in an exhaustive and well-considered deci-sionmaking process. We begin with the evidence. NHTSA compiled multiple studies and reports “indicating] that a significant number of serious and fatal injuries occur during rollovers of light trucks” — a class of vehicles including mul-ti-stage and altered trucks with a GVWR up to 10,000 pounds — and that many of those injuries could be attributed to roof crush. See 70 Fed. Reg. at 49,232-34 (finding that “412 belted, not fully ejected occupants are killed or seriously injured every year in light trucks” with a GYWR between 6,001 pounds and 10,000 pounds “involved in rollover crashes resulting in roof intrusion”); see also id. at 49,228 (noting that heavier vehicles experience higher-than-usual roof crush). NHTSA also connected the rollover problem in heavier vehicles with occupant safety. Specifically, NHTSA found that extending the scope of the roof crush resistance standard to include heavier vehicles would help “prevent 135 fatalities and 1,065 nonfatal injuries annually.” 74 Fed. Reg. at 22,349. All of which goes to show that NHTSA amassed sufficient evidence of a problem and moved to enact a reasonable solution.
In addition, we have little doubt that NHTSA weighed the proper factors in crafting this solution — one part of which was to extend the reach of the roof crush resistance standard to heavier vehicles. The record attests to NHTSA’s sensitivity to manufacturers’ concerns about demonstrating compliance with the upgraded standard. See, e.g., 70 Fed. Reg. at 49,234. The record also attests to NHTSA’s desire to seek a workable balance between such concerns and Congress’s “overriding” interest in safety. Among other things, NHTSA exempted multi-stage vehicles “where the final-stage manufacturer would need to complete the roof structure,” see 76 Fed. Reg. at 15,910; allowed all multistage vehicles not built on chassis-cabs to certify to an easier standard, see id. at 15,908; and modified the test in FMVSS No. 216a to avoid any compression of the frame rails, see id. at 15,918-19. Moreover, NHTSA made these concessions after engaging in an extensive dialogue with groups of all stripes, including NTEA. We do not think the APA or State Farm com*670pel anything more from the decisionmak-ing process.
NTEA disagrees and raises two objections that we view as inherently procedural. First, NTEA charges that NHTSA “did not even bother to test multi-stage or altered vehicles” at any point in the process. But NHTSA did perform a roof strength test on a chassis-cab prior to adopting FMVSS No. 216a. See 76 Fed. Reg. at 15,918 n.77; 75 Fed. Reg. at 17,602 & n.53 (noting that NHTSA’s test of a single-stage pickup truck would have sufficed because an attached bed does not affect roof strength). This objection misses the mark in any event. NTEA fails to identify statutory authority for the so-called testing requirement and similarly fails to explain adequately how testing might undermine the overwhelming record evidence supporting NHTSA’s decision to regulate heavier vehicles.
Second, NTEA charges that NHTSA did not sufficiently consider whether the supposedly unique attributes of multi-stage and altered vehicles justified exemption from the roof crush resistance standard. This objection is based on NHTSA’s Final Regulatory Impact Analysis (FRIA) — conducted in accordance with Executive Order 12,866 — in which the agency found that injury-causing rollovers appear to affect multistage and altered vehicles at relatively lower rates and speculated that factors like frequent use at job sites and operation in low-speed environments might explain why. But NTEA’s focus on the FRIA is misplaced. Executive Order 12,866 does not create judicially enforceable rights, nor does it provide a basis for rejecting final agency action. See § 10, 3 C.F.R. 638, 649 (1993). Regardless, NHTSA did consider whether to apply the upgraded standard to all vehicles within the newly regulated weight class or whether to exempt multistage and altered work trucks. The agency found that the latter class of vehicles was not categorically immune from rollovers and roof crush — reasonably refusing to make the same inferential leap as did NTEA based on the FRIA. See 74 Fed. Reg. at 22,373.
So, to recap, we conclude that NHTSA conducted a more-than-adequate investigation and reached a reasonable decision based on the results of that investigation. There is ample evidence in the record showing a connection between the problem of injuries resulting from roof crush during rollovers of heavier vehicles, including multistage and altered vehicles, and the solution put forth in FMVSS No. 216a. In the end, we must defer to the agency when it has engaged in such an involved process. Congress entrusted NHTSA with an important proactive regulatory mission, and it is not our role in the institutional scheme of things to make that mission unnecessarily difficult to accomplish. Indeed, the agency’s job is to promulgate standards and ours is simply to ensure that it does so responsibly. To ask for more process in a situation like this would render NHTSA’s standard-setting mission a practical impossibility. Because that is not what Congress had in mind when it passed the Safety Act, we decline to do so here.
IV.
Our review does not end with the determination that NHTSA followed the APA’s procedural requirements. This case presents a second question for decision, which is essentially a substantive one. We must ask whether FMVSS No. 216a complies with the “statutory limitations” of the Safety Act. Chrysler, 472 F.2d at 669. There are three relevant limitations — or “minimum substantive criteria.” Congress specified that any new safety standard “shall be practicable, meet the need for motor vehicle safety, and be stated in objective terms.” 49 U.S.C. § 30111(a). *671NTEA argues that FMVSS No. 216a falls short on all three counts. We disagree.
A.
NTEA’s principal argument is that FMVSS No. 216a is not “practicable” within the meaning of the Safety Act. Our guidance on this issue comes from a prior case involving these very same parties, in which we held that a safety standard is practicable only if it “offer[s] the regulated party a chance to demonstrate compliance.” Nat’l Truck, 919 F.2d at 1153. NTEA says that FMVSS No. 216a does not. It argues that performing the prescribed test is too costly for its members and that pass-through certification is inherently unworkable for multi-stage vehicles as a general matter. This line of argument is hardly novel: NTEA has made it in one form or another throughout the rulemaking proceeding, as well as in other rulemaking proceedings going back decades. We must decide whether the argument has any merit here — in other words, whether FMVSS No. 216a offers final-stage manufacturers and alterers a fair shot to demonstrate compliance via pass-through certification or other means. We think that it does.
We begin with some background. Most manufacturers can demonstrate compliance with FMVSS No. 216a by conducting the dynamic test prescribed in the standard itself. See 49 C.F.R. § 571.216a, S5-7. We do not doubt that this test is stated in objective terms and can be repeated. But NTEA and NHTSA agree that conducting the test is not a realistic option for final-stage manufacturers and alterers for other reasons. As small businesses with a fraction of the financial and engineering resources available to major manufacturers like Ford and GM, the cost of independently certifying is prohibitive. See Nat’l Truck, 919 F.2d at 1154-55. Final-stage manufacturers and alterers cannot afford to produce vehicles for the sole purpose of crushing roofs, something the test requires. See 49 C.F.R. § 571.216a, S7. Their compliance difficulties are only exacerbated by the fact of limited production runs. In extreme cases, a final-stage manufacturer or alterer might have to produce two vehicles just to sell one. (A major manufacturer, on the other hand, can spread the cost of testing across thousands of vehicles in a particular model line.) The upshot is that final-stage manufacturers and alterers need an alternative avenue for demonstrating compliance.
For final-stage manufacturers, NHTSA says that pass-through certification is the solution to their compliance problem — and we agree. It is clear from the record that final-stage manufacturers certifying to the upgraded roof crush resistance standard may rely on documentation provided by initial manufacturers, see 76 Fed. Reg. at 15,904-06, 15,913-16, just as Congress envisioned, see 49 U.S.C. § 30115. Significantly, NHTSA noted that most chassis-cabs would arrive with Type 1 or Type 2(IVD) statements regarding FMWSS No. 216a, and final-stage manufacturers would be able to conform to the listed conditions because the attachment of cargo-carrying or load-bearing equipment to bare frame rails has no obvious effect on roof strength in the cab. See 76 Fed. Reg. at 15,912-13. We know this to be true from years of practice and the accumulation of agency expertise in administering the existing roof crush resistance standard. See id. at 15,-911; 75 Fed. Reg. at 17,593, 17,602-03. Indeed, manufacturers of multi-stage vehicles with a GVWR of 6,000 pounds or less have been taking advantage of pass-through certification to successfully comply with FMVSS No. 216 since 1991. See 75 Fed. Reg. at 17,599. And manufacturers who truly cannot take advantage of pass-through certification are exempted in the final rule. See 74 Fed. Reg. at 22,349. Thus, we see no compelling reason why the *672upgraded roof crush resistance standard would render pass-through certification inherently unworkable.
An additional virtue of pass-through certification is that it does not come at a prohibitive cost, quite unlike testing. We owe considerable deference to NHTSA’s well-informed predictive judgment that the availability of pass-through certification means final-stage manufacturers will not face substantially higher compliance costs as a consequence of the upgraded standard. See Cellnet Commc’ns, Inc. v. F.C.C., 149 F.3d 429, 441 (6th Cir.1998); see also Public Citizen, Inc. v. Nat’l Highway Traffic Safety Admin., 374 F.3d 1251, 1260-61 (D.C.Cir.2004) (deferring to a NHTSA prediction on the basis that it was a policy judgment best left to the agency). Because this judgment “is within the agency’s field of discretion and expertise,” we are required to treat it favorably, even though a predictive judgment, by its nature, “cannot necessarily be proved by the record.” Cellnet, 149 F.3d at 441 (“This deference has been explained as deriving from the fact that such judgments tend to be infused with policy considerations that are not appropriate for close judicial scrutiny.”). It goes some distance toward convincing us that pass-through certification is a practicable solution to the compliance problem NTEA complains of here.
NTEA pushes back against this conclusion with two arguments — neither of which sways us. First, NTEA contends that pass-through certification is no solution at all for final-stage manufacturers because it is “not viable” as a general matter. Its allegedly fatal flaw is that self-interested initial manufacturers dictate the terms of compliance; under the Safety Act’s certification scheme, initial manufacturers have every incentive to draw up the most restrictive IVDs possible, thereby pushing much of the burden of certification (and risk of liability for failing to comply with applicable standards) onto final-stage manufacturers. See Nat’l Truck, 919 F.2d at 1155. Final-stage manufacturers complain that they exceed the bounds of these IVDs the second they go to work on a chassis-cab and cannot take advantage of pass-through certification as a result. This problem might warrant additional discussion if it found any support in the record. But whatever questions regarding pass-through certification may exist in theory have been answered by NHTSA’s experience in practice. See 75 Fed. Reg. at 17, 599 (noting that final-stage manufacturers have been using the pass-through method to certify compliance with various safety standards for decades). NTEA’s fears regarding too-restrictive IVDs appear to us unfounded. In any event, final-stage manufacturers remain free to tell dealers and customers that they will only work on chassis-cabs from initial manufacturers who provide accommodating IVDs.
Second, NTEA cites two federal appellate court decisions setting aside safety standards for failing to satisfy the practicability criterion and argues that their logic compels a similar result here. The first is Paccar, Inc. v. National Highway Traffic Safety Administration, 573 F.2d 632 (9th Cir.1978), in which an association of final-stage manufacturers successfully challenged a NHTSA standard regulating brake systems. Id. at 634-35. The court held that manufacturers were “entitled to testing criteria that they can rely upon with certainty” in demonstrating compliance with the standard, meaning “formal and reasonably specific ... criteria.” Id. at 644-45. The second decision is one of our own, in which NTEA sought review of a NHTSA standard regulating steering column displacement, based on virtually identical practicability concerns. See Nat’l Truck, 919 F.2d at 1149-50. We held that a valid safety standard must provide a “realistic way for the final-stage manufac*673turers to comply with it and thereby meet their duty to certify the safety of the vehicle.” Id. at 1155. Because the displacement standard did not provide a non-testing certification option for all final-stage manufacturers, we set aside the standard as impracticable as applied to those “final-stage manufacturers that cannot pass through the certification of the initial manufacturer.” Id. at 1158. NTEA claims that these decisions taken together require NHTSA to provide some alternative method for final-stage manufacturers to demonstrate compliance with FMVSS No. 216a.
But this claim does not stand up to scrutiny. NTEA fails to note that Paccar contains nary a word about pass-through certification, a fact that renders it largely inapposite. For our purposes, Paccar simply establishes that NHTSA must provide a reasonably definite method for manufacturers to certify compliance. See 573 F.2d at 644-45. And pass-through certification fits the bill. While our decision in the prior case between these parties is arguably on-point, it is not the silver bullet NTEA makes it out to be. There, we actually upheld the standard at issue as it applied to manufacturers able to pass through the certification of the initial manufacturer, a group that included final-stage manufacturers of vehicles built on chassis-cabs. See Nat’l Truck, 919 F.2d at 1158. Having accepted the basic premise that pass-through certification is workable, see id., it is significant that the method has been made available to all final-stage manufacturers subject to the upgraded roof crush resistance standard. In the end, these two decisions are little more than reminders that NHTSA must carefully consider the unique problems facing final-stage manufacturers and must ultimately provide them with a formal and definite method to demonstrate compliance with any new safety standard. We are satisfied that NHTSA has done both here.
For alterers, it is a different story entirely. Alterers cannot even nominally avail themselves of pass-through certification because they modify already-completed single-stage vehicles that do not arrive from the manufacturer with an IVD attached. However, for the very same reason that pass-through certification is not an option, alterers do not need any non-testing option at all. Alterers do their work on vehicles already certified under the Safety Act. See 49 C.F.R. § 567.3. Single-stage manufacturers are required to take the greater step of certifying compliance with applicable standards, see 49 U.S.C. § 30115(a), which is precisely why they do not take the lesser step of issuing an IVD. Alterers need not certify independently so long as they do not make the kind of changes that affect roof strength, see 76 Fed. Reg. at 15,905 & n. 11, and most alterers work on the rear of a vehicle, away from the cab. Even those that attach aftermarket parts to the cab rarely do so in a manner that compromises the strength of the roof pillars or roof materials. As with final-stage manufacturers and pass-through certification, the current regime reflects a status quo in which alter-ers have been successfully complying with existing FMVSS No. 216 for years. See id. at 15,905-06. Thus, we have no reason to doubt that the upgraded standard offers alterers a similarly fair shot at demonstrating compliance.
In sum, we conclude that FMVSS No. 216a is “practicable” within the meaning of the Safety Act because it provides final-stage manufacturers and alterers with reasonable means of demonstrating compliance. To conclude otherwise would disregard Congress’s instruction to put a thumb on the scale for safety in considering the substantive limitations of the Act. See Public Citizen, Inc. v. Mineta, 340 F.3d 39, 58 (2d Cir.2003). After all, Congress intended for manufacturers to adjust to the regu*674latory demands of the industry rather than the other way around. Cf. Chrysler, 472 F.2d at 671 (describing the Safety Act as technology-forcing legislation). Where there are seemingly definite and formal means for manufacturers to demonstrate compliance, as here, we ought not “second-guess the ‘practicability’ determination of an agency charged with the responsibility of studying and developing safety standards through an involved fact-finding process.” Nat’l Truck, 919 F.2d at 1159 (Guy, J., dissenting).
B.
That brings us to the two remaining substantive criteria. The Safety Act specifies that all safety standards must “meet the need for motor vehicle safety” and must be stated in “objective terms.” 49 U.S.C. § 30111(a). NTEA says that FMVSS No. 216a comes up short in both respects, at least as applied to final-stage manufacturers and alterers. In addressing NTEA’s claims, our discussion focuses on the record and thus tends to overlap with our procedural review at the margins.
NTEA’s first contention — that “NHTSA did not establish a safety need” for extending the scope of the roof crush resistance standard to include vehicles with a GVWR up to 10,000 pounds — is contradicted by the record. As we discussed earlier, NHTSA identified a general “rollover problem” in which roof crush plays a substantial role. See 70 Fed. Reg. at 49,226-29 (noting that roof crush during rollovers contributes to more than one thousand serious and fatal injuries every year). In addition, NHTSA found that heavier trucks not subject to existing FMVSS No. 216, multi-stage and altered vehicles among them, were a conspicuous part of the problem. Id. at 49,283. Finally, NHTSA produced evidence showing that its proposed changes would “prevent 135 fatalities and 1,065 nonfatal injuries annually.” 74 Fed. Reg. at 22,349; see also 70 Fed. Reg. at 49,233 (estimating that applying the standard to vehicles with a GVWR between 6,001 pounds and 10,000 pounds would prevent at least 129 of those fatalities and serious injuries). This evidence strikes us as sufficient alone to prove that FMVSS No. 216a meets a safety “need.”
We also think it worth noting that NHTSA promulgated FMVSS No. 216a pursuant to a specific congressional mandate. See 49 U.S.C. § 30128(a); see also id. § 30128(d) (directing NHTSA to initiate a rulemaking proceeding to “upgrade [FMVSS] No. 216 relating to roof strength for driver and passenger sides”). Nothing suggests that the mandate was conditioned upon a finding that heavier vehicles experience rollovers at a greater rate than vehicles subject to the existing standard. As Congress made clear, the occupants of heavier vehicles, including multi-stage and altered vehicles, are equally entitled to the safety benefits secured by an upgraded standard. See 76 Fed. Reg. at 15,919. Thus, the “need” requirement in the Safety Act cannot be said to prohibit NHTSA from acting as it did in this instance.
Likewise, NTEA’s second contention— that NHTSA failed to provide an “objective method” by which final-stage manufacturers and alterers can demonstrate compliance with the upgraded standard — is unavailing. There is no question that the dynamic test prescribed in FMVSS No. 216a is set forth in sufficiently objective terms. See 49 C.F.R. § 571.216a, S5.1. And this test forms the basis for certifying that any particular vehicle complies with the upgraded standard. Pass-through certification is only available to a final-stage manufacturer working on a multi-stage vehicle because the initial manufacturer properly tested the chassis-cab prior to delivering it. We thus conclude that FMVSS No. 216a is “objective” within the meaning of the Safety Act.
*675V.
The final question for review differs in nature from the prior two. We must decide whether pass-through certification, as incorporated in FMVSS No. 216a, constitutes an improper delegation of NHTSA’s statutory duties. Specifically, NTEA argues that initial manufacturers are exercising agency powers when they write IVDs without oversight because the terms of these IVDs dictate whether final-stage manufacturers can meet their duty to certify compliance with applicable safety standards. We are not convinced.
Under U.S. Telecom Association v. F.C.C., 359 F.3d 554 (D.C.Cir.2004), agency “subdelegations [of statutory powers] to outside parties are assumed to be improper absent an affirmative showing of congressional authorization.” Id. at 565. From this premise, NTEA concludes that NHTSA has used pass-through certification as a way to avoid promulgating a safety standard that is practicable and objective. But NTEA misunderstands the improper delegation doctrine on several levels. Most fundamentally, NTEA fails to specify any particular power, conferred on NHTSA by Congress, that the agency has turned around and actively delegated to initial manufacturers. NHTSA quite clearly promulgated FMVSS No. 216a. In addition, unregulated IVDs do not imper-missibly reallocate the statutory division of responsibility for certifying compliance. Congress in fact explicitly endorsed the pass-through certification regime in 2000. See 49 U.S.C. § 30115(b)(1). Allowing initial manufacturers to write their own IVDs is not the same thing as allowing them to pass off their risk of liability onto helpless final-stage manufacturers. Thus, we conclude that NHTSA has not improperly delegated its statutory duties.
VI.
For the foregoing reasons, we reject NTEA’s challenges to FMVSS No. 216a and deny the petition for review.