Three commercial truck drivers and the Owner-Operator Independent Drivers Association (OOIDA) have petitioned for review of a final rule issued by the Federal Motor Carrier Safety Administration (FMCSA or Agency) about the use of electronic monitoring devices in commercial trucks. Electronic On-Board Recorders for Hours-of-Service Compliance, 75 Fed. Reg. 17,208 (Apr. 5, 2010). Though the briefing raises a litany of issues that would make for a difficult and exhaustive Administrative Law final exam, in the end we find that we can dispose of the petition on a narrow basis. We conclude that the rule cannot stand because the Agency failed to consider an issue that it was statutorily required to address. Specifically, the Agency said nothing about the requirement that any regulation about the use of monitoring devices in commercial vehicles must “ensure that the devices are not used to harass vehicle operators.” 49 U.S.C. § 31187(a). We therefore grant the petition and vacate the rule.
I
Federal regulators have long limited the number of hours during which commercial truck drivers may operate their vehicles in a given day and over the course of a week. Between 1940 and 2003, the permissible “hours of service” (HOS) went largely unchanged for most drivers. The basic idea has remained constant: to protect driver health and to ensure highway safety by reducing driver fatigue and thus fatigue-related accidents. To keep track of a trucker’s time on the road and, to the extent possible, his time spent sleeping, the regulations require a driver to document four statuses: (1) driving; (2) on duty, not driving (e.g., sitting at a loading dock or filling up the gas tank); (3) in the sleeper-berth (a small compartment in the cab of the truck with a bed); and (4) off duty. 49 C.F.R. § 395.8(b) (2010). The regulations set out daily limits for time spent either driving or otherwise on duty, and they establish a daily minimum for consecutive hours off duty. That minimum is subject to an exception for time spent in the sleeper-berth. This allows drivers to split their off-duty hours into two parts if they rest in the truck. The regulations also cap the total hours a driver may spend on duty in a given week (which can be measured in either seven- or eight-day units depending on the carrier).
Traditionally, drivers have recorded their hours in paper logbooks (referred to a driver’s “record of duty status,” id. at § 395.8) to demonstrate compliance with the HOS regulations. Individual drivers must keep copies of their records-of-duty status for seven days. They then submit the records to their motor carrier, which must retain them for six months. Id. at § 395.8(k). As one might imagine, this paper-based system is not free from problems of manipulation and falsification, and those problems have long been a subject of concern.
In 2003, the FMCSA issued a final rule that substantially, and controversially, changed the HOS numbers and how they would be measured. Hours of Service of Drivers; Driver Rest and Sleep for Safe Operations, 68 Fed. Reg. 22,456 (Apr. 28, 2003). The 2003 rule increased the daily driving limit, reduced the daily on-duty limit, increased the daily off-duty requirement, retained the sleeper-berth exception, and created a 34-hour restart rule as a new exception to the weekly on-duty cap. See id. at 22,457, 501-02. Following a petition for review, however, the D.C. Circuit held that the 2003 revised rule was arbitrary and capricious because the “agency failed to consider the impact of the rules on the health of drivers, a factor the agency must consider under its organic *583statute.” Public Citizen v. FMCSA, 374 F.3d 1209, 1216 (D.C.Cir.2004).
Although the court’s decision ordinarily would have required vacatur of the rule, Congress overrode that consequence in legislation that granted the Agency temporary relief and kept the 2003 rulemaking in effect until the earlier of either a new rule from the Agency or one year. Surface Transportation Extension Act of 2004, Part V, Pub. L. 108-310, § 7(f), 118 Stat. 1144, 1154. The Agency went back to the drawing board, but it ultimately made only one small change to the rule. In 2005 it issued a final rule with the revised HOS regulations. Hours of Service of Drivers, 70 Fed. Reg. 49978 (Aug. 25, 2005). Again, the new HOS rule did not survive judicial review in one piece. This time, the Agency erred because (1) it failed to allow meaningful comment on the driver-fatigue model it used to justify increasing the daily driving limit and in creating the 34-hour restart provision, and (2) it failed adequately to explain its reasons for adopting this model, which figured heavily in the Agency’s cost-benefit analysis. These two flaws, the D.C. Circuit concluded, were serious enough to require more changes to the rule. OOIDA v. FMCSA 494 F.3d 188, 199-206 (D.C.Cir.2007).
Nestled within the Agency’s larger consideration- of the HOS rules is the- more narrow, but still controversial, regulatory issue before us. In the notice of proposed rulemaking for the 2003 rule, the Agency considered requiring truckers to use electronic on-board records (EOBRs) instead of logbooks for documenting their records of duty status. The Agency defines an EOBR as “an electronic device that is capable of recording a driver’s hours of service and duty status accurately and automatically.” 49 C.F.R. § 395.2 (2011). An EOBR must be “integrally synchronized” with a truck’s engine, id; this allows the device to be linked simultaneously with both the engine and the driver’s telephone so that contemporaneous updates can be sent either through cellular technology or via satellite to a remote server. To meet the Agency’s performance requirements, the amount of data an adequate EOBR must be capable of recording is extensive: the truck’s registration number, the date and time, the location of the truck, the distance traveled, the hours in each duty status for a 24-hour period, the motor carrier’s name and Department of Transportation number, the weekly basis used by the motor carrier (either seven or eight days) to calculate cumulative driving time, and even the document numbers or name of the shipper and goods being shipped. Id; 49 C.F.R. § 395.16 (2010). At a less technical level, an EOBR is essentially a device implanted into a truck that records significant amounts of data about the truck’s location, how it is being used, how it has been used over time, and that uses satellite technology to allow nearly instant electronic transmission of this data to the trucker’s employer (that is, the motor carrier).
During the 2003 rulemaking procedure, the Agency determined that “falsification of logbooks ... [is] widespread.” Public Citizen, 374 F.3d at 1214 (citing 65 Fed. Reg. 25,540, 25,558 (2000)). Ultimately, however, it decided not to require EOBRs as part of its comprehensive overhaul of the HOS rules, see 68 Fed. Reg. at 22,488-89. It gave several reasons for that decision. The most significant was that it believed that, as of then, it could not adequately estimate the costs or benefits of an EOBR, in part because the market was small (which made cost estimates difficult), and in part because it had not tested available devices or those in current use (which made benefits, like increased compliance, difficult to estimate). Id In addition, the Agency wanted more time to address the concerns that had been expressed about *584secondary uses of data and about the effects of EOBRs on privacy. Id. at 22,489. Even though it concluded that it could not justify a general EOBR requirement in 2003, the Agency promised “to continue research on EOBRs and other technologies.” Id. at 22,488. Public Citizen criticized the Agency’s decision not to adopt an EOBR requirement, to the extent that the reason was a lack of information about the potential costs and benefits of an EOBR mandate. 374 F.3d at 1220-22. The court could not “fathom” why the Agency had not “even taken the seemingly obvious step of testing existing EOBRs on the road, or why the agency ha[d] not attempted to estimate their benefits on imperfect empirical assumptions.” Id. at 1222. The court characterized the absence of information about how much EOBRs could increase compliance as “willful,” given the lack of testing. Id. This was especially glaring since the Agency had allowed voluntary EOBR use for over 15 years at the time. Id. After Public Citizen, in the rule-making conducted under the temporary stay from Congress, the Agency did not address EOBRs.
In 2004, the Agency made good on its promise to investigate EOBRs further when it issued an optional advanced notice of proposed rulemaking indicating that it was still considering an EOBR mandate. Electronic On-Board Recorders for Hours-of-Serviee Compliance, 69 Fed. Reg. 53386 (Sept. 1, 2004). That advance notice led to a formal notice of proposed rulemaking in 2007, which considered three regulatory issues: (1) new performance standards for EOBR technology; (2) the use of EOBRs to “remediate regulatory noncompliance”; and (3) incentives to promote voluntary use of EOBR technology. Electronic On-Board Recorders for Hours-of-Service Compliance, 72 Fed. Reg. 2340, 2343 (Jan. 18, 2007). The noncompliance measure was the most significant of the three and is the focus of our review. The Agency proposed requiring EOBR use for carriers found to have an HOS violation of greater than 10 percent for any two compliance reviews in a two-year period. The Agency calls this the “2 x 10 remedial directive.” As the Agency described this option in its final rule-making, the 2x10 remedial directive would apply to “a relatively small population of companies and drivers with a recurrent HOS compliance problem.” 75 Fed. Reg. at 17,211. In the section of the notice of proposed rulemaking that lists statutes authorizing the FMCSA to promulgate regulations regarding monitoring devices, the Agency acknowledged that Congress contemplated rules mandating electronic monitoring devices in the Truck and Bus Safety and Regulatory Reform Act of 1988. 72 Fed. Reg. at 2341.. This statute, the Agency realized, also “requires the Agency to ensure that any such device is not used to ‘harass vehicle operators.’ ” Id. (quoting 49 U.S.C. § 31137(a)).
After receiving and considering a significant number of comments to the proposed rule, the Agency promulgated the 2010 final rule now before us. In so doing, the Agency produced two reports related to its reasons for not requiring EOBRs in 2003 — a Regulatory Impact Analysis that weighed potential costs and benefits of requiring EOBRs, and a Privacy Impact Assessment. In the end, it decided on a rule under which motor carriers “that have demonstrated serious noncompliance with the HOS rules will be subject to mandatory installation of EOBRs.” 75 Fed. Reg. at 17,208. The 2010 rule abandoned the proposed 2x10 directive in favor of a stricter “1 x 10 remedial directive.” Under the 1 x 10 directive, the Agency will issue a remedial order — requiring installation of EOBRs for all the trucks in a motor carrier’s fleet — to any carrier found to have a greater than 10 percent rate of noncompliance with HOS rules in any single “compli*585anee review.” Id. A remedial directive would make the EOBRs mandatory for a period of two years; after that, a carrier, if it desired, could in theory return to using the logbooks. Id. at 17,211. The Agency adopted the stricter rule because 1x10 carriers have a 40 percent higher crash rate than the general motor carrier population, id., and because it agreed with the numerous commenters that the “proposed 2x10 trigger would not mandate EOBR use by enough carriers, given the total population.” Id. at 17,215.
The 2010 rule was entered April 5, 2010, and this petition followed the day before the rule became effective, June 4, 2010. Though final, the rule set an HOS compliance date — with the specter of a remedial directive following a 1 x 10 violation — two years down the road — June 4, 2012.
II
Before reaching the merits, we must consider several key preliminary points: jurisdiction, standing, and ripeness. The Agency challenges the latter two, and we have an independent obligation to assure ourselves that jurisdiction is secure. Because the FMCSA is a part of the Department of Transportation, we start with the Hobbs Act, which supplies our jurisdiction and guides our consideration of standing: “Any party aggrieved by the final order may, within 60 days after its entry, file a petition to review the order in the court of appeals wherein venue lies.” 28 U.S.C. § 2344. No one contests that the petition was timely, that the 2010 rule was a “final order,” or that venue is proper (the individual petitioners live in this jurisdiction).
Our attention, therefore, is on the question whether the petitioners qualify as “parties aggrieved” by the rule. This is an inquiry that incorporates elements of both prudential and constitutional standing. Brotherhood of Locomotive Engr’s v. United States, 101 F.3d 718, 723 (D.C.Cir.1996). We have no doubt, and the Agency does not challenge, that the petitioners have satisfied the purely prudential concerns— that a challenger be a “party” to the agency proceedings, id., and “arguably within the zone of interests” regulated by the agency. Association of Data Processing Serv. Orgs., Inc. v. Camp, 397 U.S. 150, 153, 90 S.Ct. 827, 25 L.Ed.2d 184 (1970). Petitioners’ participation in the notice-and-comment process below establishes their party status, while the fact that the organization is composed of and represents the truckers being regulated (and the individual petitioners are themselves subject to the new rule) easily brings them within the “zone of interests” of the 2010 rule.
The “aggrieved” requirement of the Hobbs Act implicates constitutional standing, which, through Article III, requires a petitioner to demonstrate a concrete and particularized injury, which has been caused by the agency, and which a court can redress. Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-61, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992). Only injury is at issue here, and we find this requirement easily met. As Lujan explained, when the complainant is “an object of the action (or forgone action) at issue ... there is ordinarily little question that the action or inaction has caused him injury.” Id. at 561-62, 112 S.Ct. 2130. The three truck drivers are the “objects of the action” here. It is truck drivers who will be required to install EOBRs should they work for a carrier subject to a remedial directive, and the central purpose of the rule is to increase their compliance with HOS regulations. Although compliance is measured at the carrier level, it is the individual truck drivers whose sleep is really at issue under the HOS rules and whose status is being tracked on a day-today basis. As a trade association that includes truckers and represents their in*586terests, OOIDA meets the requirements of associational standing; it has Article III standing because the individual truckers do. Hunt v. Washington State Apple Adver. Comm’n, 432 U.S. 333, 343, 97 S.Ct. 2434, 53 L.Ed.2d 383 (1977); Sierra Club v. EPA, 292 F.3d 895, 898 (D.C.Cir.2002).
The Agency’s argument that the petitioners’ injury is pure “speculation” is premised on the theory that because the petitioners are not currently under a remedial directive they cannot demonstrate an injury. Extending this premise, the Agency argues that only drivers or motor carriers “currently subject” to a remedial directive have standing to challenge the Agency’s rule. The Agency’s standing argument, however, ignores the very idea that it advances to justify adopting the EOBR rule in the first place: a punitive stick (it says) is necessary to increase compliance with HOS regulations. The 2010 rule aims to alter truck drivers’ behavior now to avoid a remedial directive in the future. Cf. 75 Fed. Reg. at 17,211 (noting the rulemaking “provides immediate safety benefits to society” (emphasis added)). Indeed, the very reason the Agency chose the 1x10 directive rather than the 2 x 10 option was to raise the stakes in an effort to achieve more compliance by 2012. In the end, it strikes us as odd that the Agency is arguing that it must have a strict rule now to get truck drivers to be more compliant with HOS rules, but at the same time it is asserting that these rules are not meant to change anyone’s immediate behavior enough to confer standing to challenge that regulation. Cf. Stilwell v. Office of Thrift Supervision, 569 F.3d 514, 518 (D.C.Cir.2009) (finding it “more than a little ironic” that an agency “would suggest Petitioners lack standing and then, later in the same brief,” label the petitioner a “prime example” of the “very problem the [r]ule was intended to address” (internal quotation marks omitted)).
As for ripeness, the Agency relies on the same theory as it did for standing: any challenge to the rule is unripe because the petitioners are not currently under a remedial directive. This argument ignores the well-established rule of Abbott Laboratories, which permits preenforcement challenges of final agency rules so long as the claim is fit for judicial decision and delay will cause some hardship to the parties. Abbott Labs. v. Gardner, 387 U.S. 136, 149, 87 S.Ct. 1507, 18 L.Ed.2d 681 (1967). In the decades since Abbott Laboratories, pre-enforcement review of final rules has become the norm. Where, as here, a petition involves purely legal claims in the context of a facial challenge to a final rule, a petition is “presumptively reviewable.” Sabre, Inc. v. Dep’t of Transp., 429 F.3d 1113, 1119 (D.C.Cir.2005) (internal quotation marks and citations omitted). In light of this presumption, a petitioner need not demonstrate individual hardship to show ripeness unless the agency identifies “institutional interests favoring the postponement of review.” Id. All agree that this case is presumptively fit for review (the issues are purely legal), and the Agency has not identified any institutional interest that would require a more exacting inquiry into hardship and thus would counsel for delay. Further, as Abbott Laboratories itself demonstrated, hardship need not take the form of an actual enforcement action; the threat of enforcement is sufficient because the law is in force the moment it becomes effective and a person made to live in the shadow of a law that she believes to be invalid should not be compelled to wait and see if a remedial action is coming. 387 U.S. at 150-54, 87 S.Ct. 1507; see also Metropolitan Milwaukee Ass’n of Commerce v. Milwaukee County, 325 F.3d 879, 882-83 (7th Cir.2003). This rationale is particularly powerful here because a compliant truck driver who works for a carrier *587that happens to have 11 out of 100 drivers out of compliance within one year will be forced to install an EOBR even if she never violated the rules. And, as we explained in our discussion of standing, the 1x10 mandate is meant to induce the trucking industry to “change their behavior or risk costly sanctions.” Clean Air Implementation Project v. EPA, 150 F.3d 1200, 1205 (D.C.Cir.1998). That is enough detriment to make a petition ripe.
Before leaving the topic of ripeness, we must make one last point. The Agency’s theory that no claim is ripe until a punitive sanction has been imposed not only ignores Abbott Laboratories, but also is flawed because it conflicts with our jurisdictional statute. The Hobbs Act provides that petitions for review must be filed within 60 days following the entry of a final rule. 28 U.S.C. § 2344. Congress wrote this statute for the very purpose of ensuring ex ante review of an entire rule (as opposed to a more narrow, ex post or as-applied challenge), just as it has done for final rules in a host of other regulatory contexts. See, e.g., Clean Air Implementation Project, 150 F.3d at 1204. If the Agency’s theory were correct, any final rule could be insulated from a pre-enforcement challenge by the simple expedient of setting an effective date 61 or more days after the rule was entered; ripeness would always stand as a bar to a petition. Such an interpretation of ripeness would directly oppose Congress’s determination that pre-enforcement review should be the norm for agency rules that fall under statutes like the Hobbs Act. Accordingly, we reject this argument.
Ill
Turning to the merits, the petitioners raise three reasons for vacating the 2010 final rule. First, they argue that the regulation is arbitrary and capricious because it does not “ensure that the devices are not used to harass vehicle operators,” as required by 49 U.S.C. .§ 31137(a); see 5 U.S.C. § 706(2)(A). Second, they argue that the Agency’s cost-benefit analysis is arbitrary and capricious because it fails to demonstrate the benefits of requiring EOBRs. Finally, the petitioners argue that mandating EOBRs violates the Fourth Amendment. We need address only the first issue.
Our starting point is the Supreme Court’s State Farm decision, which explained that normally, “an agency rule would be arbitrary and capricious if the agency has ... entirely failed to consider an important aspect of the problem” before it. Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983). This was the principle at issue in the earlier Public Citizen decision about these very rules. In Public Citizen, the D.C. Circuit applied the well-established rule that when an agency fails to consider a factor mandated by its organic statute, this omission is alone “sufficient to establish an arbitrary-and-capricious decision requiring vacatur of the rule.” 374 F.3d at 1216; see also Texas Oil & Gas Ass’n v. EPA, 161 F.3d 923, 934 (5th Cir.1998). When Congress requires an agency to address something before issuing a regulation, that factor is by definition an “important aspect of the problem” under State Farm. See Public Citizen, 374 F.3d at 1216. This is in contrast to factors an agency “may” consider, which might not always be an “important aspect” of a problem. Adherence to this rule is essential because it goes directly to the scope of the authority delegated to an agency by Congress; when an agency ignores a mandatory factor it defies a “statutory limitation on [its] authority.” United Mine Workers v. Dole, 870 F.2d 662, 673 (D.C.Cir.1989). Such an act is necessarily arbitrary and capricious.
*588Congress foresaw that monitoring devices on trucks might be used to enforce HOS rules, and that these devices could potentially be used to harass drivers. This led to the following provision:
Use of monitoring devices. — If the Secretary of Transportation prescribes a regulation about the use of monitoring devices on commercial motor vehicles to increase compliance by operators of the vehicles with hours of service regulations of the Secretary, the regulation shall ensure that the devices are not used to harass vehicle operators. However, the devices may be used to monitor productivity of the operators.
49 U.S.C. § 31137(a). There is no question that section 31137(a) is mandatory. The Agency concedes that it would be arbitrary and capricious not to consider this factor or to fail to explain its conclusion about the risk of harassment.
The Agency has taken on the difficult task of arguing that it adequately and expressly considered whether the EOBRs mandated by its rale would harass drivers. But its first argument can be set aside immediately. The FMCSA suggests that a single conclusory sentence in the final rulemaking to the effect that the Agency “has taken the[ ] statutory requirement!] into account throughout the final rule” is enough by itself to satisfy section 31137(a). It is not. Unfortunately, however, this sentence does represent the entirety of the Agency’s direct consideration of harassment, and even it is a bit elliptical. The word “harass” appears only once in the entire rulemaking, in the explanatory “legal basis for the rulemaking” section; otherwise it is not mentioned. This explanation is insufficient. The Agency must articulate a reason for its action that demonstrates a “rational connection between the facts found and the choice made.” State Farm, 463 U.S. at 43, 103 S.Ct. 2856 (internal quotation marks and citations omitted). Its explanation may not be superficial or perfunctory. Instead, the Agency must “cogently explain why it has exercised its discretion in a given manner,” id. at 48, 103 S.Ct. 2856, such that we can be sure that the decision was the product of “reasoned decisionmaking,” id. at 52, 103 S.Ct. 2856. When this standard has not been met, it is necessary to vacate the agency’s action. See, e.g., Advocates for Highway & Auto Safety v. FMCSA 429 F.3d 1136 (D.C.Cir.2005); OOIDA v. FMCSA, 494 F.3d at 206.
To provide an adequate explanation under section 31137(a), the Agency should have revealed how it drew the line between legitimate measures designed to assure productivity and forbidden measures that harass. These terms are undefined in the statute and thus require some amplification. At argument, counsel for the Agency told us that the devices would be used only for monitoring productivity, but we cannot accept this as a substitute for a proper explanation for at least two reasons. First, counsel’s post hoc rationalizations for the agency’s action are no substitute for the agency’s work. Citizens to Pres. Overton Park, Inc. v. Volpe, 401 U.S. 402, 419-20, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971); Smith v. Harvey, 458 F.3d 669, 674 (7th Cir.2006). Second, this particular statement actually conflicts with the Agency’s explicit representation that EOBRs would not be used to monitor productivity. 75 Fed. Reg. at 17,214 (“[T]he Agency is requiring EOBRs to monitor safety not workplace productivity.”).
In addition, an adequate explanation that addresses the distinction between productivity and harassment must also describe what precisely it is that will prevent harassment from occurring. The Agency needs to consider what types of harassment already exist, how frequently and to what extent harassment happens, and how *589an electronic device capable of contemporaneous transmission of information to a motor carrier will guard against (or fail to guard against) harassment. A study of these problems with EOBRs already in use, and a comparison with carriers that do not use these devices, might be one obvious way to measure any effect that requiring EOBRs might have on driver harassment.
The Agency’s back-up argument fares no better than its first one. For the first time in its consideration of EOBRs, the Agency’s brief before this court introduces the argument that its consideration of privacy and the Privacy Impact Assessment it produced also addresses the statutory factor of harassment. This argument is too little, too late. Petitioners point out that privacy and harassment are two different — even if related — concepts. As the D.C. Circuit held, the “relatedness of the concept discussed to the statutorily mandated factor that the agency does not discuss does not relieve the agency of the duty of compliance with the congressional instruction.” Public Citizen, 374 F.3d at 1217. No one would seriously take the position that “privacy” is a synonym for “harassment.” That being the case, section 31137(a) requires the Agency to consider the issue of harassment independently-
It is also significant that OOIDA raised the issue of harassment in its comments on the rale, related the concept of harassment to the sorts of pressure carriers exert over drivers, and gave examples illustrating the concerns of its members. See Comments of OOIDA, FMCSA-2004-18940-1094, at 30-31 (Apr. 18, 2007); Comments of OOI-DA, FMCSA-2004-18940-281, at 10-11 (Nov. 30, 2004). Other comments reported that drivers have been pressured by their motor carriers to perform at higher levels (and drive even when tired) as a result of the fact that an EOBR can send the carrier data in real time. Even if the 2010 rule does not require that level of reporting, the technology certainly allows it, and that is what motivated these comments. See, e.g., Comments of International Brotherhood of Teamsters, FMCSA-2004-189401105, at 5-6 (Apr. 18, 2007); Comments of Virginia L. Ganster, FMCSA-2004-189401093, at 3 (Apr. 18, 2007). The Agency’s failure to respond to this concern, which describes a form of harassment that the statute required it to address and that raises problems distinct from privacy, renders the rule arbitrary and capricious.
IV
We could say more about the other issues raised by the petitioners. For instance, some of the problems Public Citizen identified with the Agency’s cost-benefit analysis of EOBRs — like the failure to estimate the benefits of the EOBRs by looking at and testing the thousands currently in use — appear not to have been fully resolved. 374 F.3d at 1222. But this area of regulation is moving quickly. We note, for example, that the FMCSA has already proposed a rule requiring EOBRs for all motor carriers, that the technology and markets are rapidly changing, and that the Agency is apparently conducting new case studies on EOBR use. Rather than reach beyond what is strictly necessary here, prudence dictates that we leave for another day any questions that might arise in connection with whatever new rule the Agency decides to adopt.
For these reasons, we Vacate the Agency’s final rule and Remand for proceedings consistent with this opinion.