Thomas E. Perez v. Loren Cook Company

MELLOY, Circuit Judge,

with whom MURPHY, BYE, and KELLY, Circuit Judges, join, dissenting.

I. Introduction

The majority correctly notes that the Secretary’s deviation from a longstanding *944interpretation of a regulation can be a factor in assessing the reasonableness of a new interpretation. This factor appears to enjoy the support of two rationales. First, notice concerns arise if the regulated entity is unaware of the interpretation prior to taking action (notice concerns). And second, it is possible that through the pronouncement of a new interpretation in an administrative adjudication, the Secretary’s delegatees may be skirting normal agency process and advancing an interpretation that reflects an ad hoc rationalization for an action or more simply does not reflect the Secretary’s reasoned and considered judgment (agency-process concerns).

It does not follow, however, that we must address notice concerns through a wholesale rejection of the Secretary’s new interpretation. Rather, the Secretary’s interpretation may stand, and notice concerns may affect the penalty, if any, to be imposed. It also does not follow that a court should reject the new interpretation as failing to honor agency process when an administrative adjudication is a permissible vehicle for articulating an interpretation, the interpretation is consistent with the text of the regulation, and the Secretary’s past interpretation is not an expressly announced position, but rather a pattern of agency inaction.

The question at the heart of this appeal is whether our court should exercise restraint and defer to the evolving views of the Secretary in this matter. The majority appears to at least partially share the view that the Secretary possesses “expertise in workplace safety matters and ... need[s] flexibility in construing ambiguous regulations.” The majority, however, goes to great lengths in an attempt to make the Secretary’s clear and textually supported interpretation of the regulation appear unreasonable in order to justify a refusal to extend deference. In doing so, the majority relies heavily on materials outside the regulation — materials which may be evidence of inconsistent interpretations but which do not disprove the simple fact that the Secretary’s interpretation is a straightforward and clear reading of the regulation.

Ultimately, I disagree with the majority’s conclusion that its interpretation, rather than the Secretary’s interpretation, is more reasonable. I also believe that, because the Secretary’s interpretation enjoys substantial textual support, we should grant the agency the substantial deference it is owed in interpreting its own regulation. See Auer v. Robbins, 519 U.S. 452, 461, 117 S.Ct. 905, 137 L.Ed.2d 79 (1997) (holding that the Secretary’s interpretation of its “own regulations ... is ... controlling unless ‘plainly erroneous or inconsistent with the regulation’ ”) (quoting Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414, 65 S.Ct. 1215, 89 L.Ed. 1700 (1945)) (additional citation omitted). Any notice concerns should be dealt with in terms of the penalty, if any, to be imposed and not a wholesale rejection of the Secretary’s interpretation.

II. Background

I add upon the facts presented by the majority because I find additional facts to be material and because the ALJ failed to make certain findings that would have been material to the reasonableness of the Secretary’s interpretation. Loren Cook uses lathes of different sizes to form workpieces of different sizes. Large lathes employ barrier guards to protect workers from ejected objects. In the past, small lathes also employed such guards. By May 2009, however, Loren Cook had removed guards from the small lathes. At that time, the twelve-pound workpiece at issue broke loose, shot out, and struck a worker in the head, killing him. Although *945the parties dispute the frequency with which similar ejections of workpieces occurred in the past, it is undisputed that prior workpiece ejections had occurred. For example, approximately two weeks before the incident that killed the worker, a workpiece from a small lathe shot out and narrowly missed a worker twenty feet away.4

When Loren Cook sought review of the Secretary’s charge, the ALJ held a twenty-day hearing that resulted in an extensive record. The ALJ concluded that 29 C.F.R. § 1910.212(a)(1) did not apply in the context of the present case. According to the ALJ, the regulation only required guards on lathes to prevent debris or waste material from being ejected; it did not apply to guard against the ejection of the actual item being worked on, i.e., the ejection of the actual workpiece. As a result of this threshold determination, and notwithstanding the substantial record, the ALJ elected not to reach several other elements of the charge and defenses to the charge, stating, “it is not necessary to address several of the issues raised at the hearing, including the feasibility of abatement, fair notice, credibility of experts, willful classification, and collateral estop-pel.” (Emphasis added). Finally, the ALJ denied any pending motions not previously ruled on, presumably as moot, in light of the ALJ’s holding. The Commission declined further review, and the ALJ’s decision became a final order of the Commission. It is clear to me that, had the ALJ found the regulation to be applicable, the twenty-day hearing and the resulting record made the case fully ripe for the ALJ to rule on the issue of whether guards on small lathes would be feasible and whether fair notice concerns otherwise would have precluded, the imposition of a fine.. The ALJ simply chose not to address those issues.

Notwithstanding the ALJ’s failure to address the issue of feasibility, the facts strongly suggest the use of barrier guards on small lathes is feasible: they were used in the past; after the fatality at least one lathe operator unilaterally reinstalled such a guard; and such guards are still in use on the larger lathes. I view these facts as militating strongly against an assumption that use of the guards might be infeasible or ineffectual.5

III. Discussion

A. Martin

Normally, our review of a petition from a Commission order would be standard deferential review pursuant to the Administrative Procedures Act. See Omaha Paper Stock Co. v. Sec’y of Labor, 304 F.3d 779, 782 (8th Cir.2002) (“We will uphold the Commission’s legal conclusions unless they are ‘arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.’ ” (quoting 5 U.S.C. § 706(2)(A))). *946Here, however, the Secretary appeals as to a question of regulatory interpretation upon which the Secretary and the Commission have adopted competing positions. Assuming the competing interpretations are reasonable, we must defer to the Secretary. See Martin v. Occ. Safety & Health Review Comm’n, 499 U.S. 144, 152-53, 111 S.Ct. 1171, 113 L.Ed.2d 117 (1991) (resolving the relative authority of the Secretary and Commission); Solis v. Summit Contractors, Inc., 558 F.3d 815, 823-25 (8th Cir.2009) (applying Martin).

In Martin, the Court resolved a circuit split and held that “a reviewing court may not prefer the reasonable interpretations of the Commission to the reasonable interpretations of the Secretary.” 499 U.S. at 158, 111 S.Ct. 1171. In reaching this conclusion, the Court addressed Congressional intent in depth and examined the specific statutory division of adjudicatory and policymaking authority between the Commission and the Secretary. Id. at 151-54, 111 S.Ct. 1171. The Court emphasized that the Occupational Safety and Health Act (“OSHA”) did not create a typical unitary administrative, agency, but that the Commission and Secretary represented a separation of neutral, adjudicatory functions, on the one hand, from enforcement and policymaking functions, on the other. Id. The Court concluded unequivocally that deference in the interpretation of regulations was owed to the Secretary rather than the Commission, stating:

[T]he Commission is authorized to review the Secretary’s interpretations only for consistency with the regulatory language and for reasonableness. In addition, ... Congress expressly charged the Commission with making authoritative findings of fact and with applying the Secretary’s standards to those facts in making a decision. See 29 U.S.C. § 660(a) (Commission’s factual findings “shall be conclusive” so long as “supported by substantial evidence”). The Commission need be viewed as possessing no more power than this in order to perform its statutory role as “neutral arbiter.”

Id. at 154-55, 111 S.Ct. 1171.

Martin remains good law, although several courts have recognized the limited scope of Martin’s holding. For example, courts have refused to apply Martin in cases involving different agencies. See, e.g., Hinson v. Nat’l Transp. Safety Bd., 57 F.3d 1144, 1148 n. 2 (D.C.Cir.1995) (recognizing the narrow applicability of Martin and refusing to apply Martin in a case involving competing interpretations from the Federal Aviation Administration and the National Transportation Safety Board). And courts have determined that Martin was not controlling as to questions of statutory interpretation. See, e.g., Chao v. Occ. Safety & Health Review Comm’n, 540 F.3d 519, 525 (6th Cir.2008) (“Left undecided by Martin, however, is to whom doés a reviewing court defer when the Secretary and Commission offer conflicting interpretations of a provision of [OSHA].”). These refusals by other courts to expand Martin do not undercut Martin’s holding because the Supreme Court in Martin defined the issue narrowly and did not purport to issue a broad ruling that might apply in other contexts or to other agencies. Martin, 499 U.S. at 157, 111 S.Ct. 1171 (“We emphasize the narrowness of our holding. We deal in this case only with the division of powers between the Secretary and the Commission under the OSH Act.”). In fact, the nature of the issue raised in Martin was 'such that courts would not expect Martin to find application except in this very specific context: Martin rested on the careful division of authority Congress set out for the Secretary and the Commission, and that type of division of authority likely will vary from agency to agency and statute to statute.

*947The analysis in Martin itself also makes clear that the Secretary’s understanding of the effect of an interpretation may develop over time given the Secretary’s involvement with many more enforcement actions than the Commission.6 The Court identified this fact as one of the Secretary’s “structural advantages” over the Commission in the interpretation of regulations. Id. at 152, 111 S.Ct. 1171. Because the Court expressly anticipated that the Secretary may adjust its interpretation of a regulation over time, we should guard against championing the need for consistency at the expense of the Secretary’s flexibility. Like the Supreme Court, I believe a general review for reasonableness and for adherence to regulatory language is strong protection against surprising, biased, or abusive interpretations. See id. at 156, 111 S.Ct. 1171 (“Congress also intended to protect regulated parties from biased interpretations of the Secretary’s regulations. But this objective is achieved when the Commission, and ultimately the court of appeals, review the Secretary’s interpretation to assure that it is consistent with the regulatory language and is otherwise reasonable”).

Our review in this matter therefore requires that we address the Secretary’s interpretation of § 1910.212(a)(1) to determine whether it is a reasonable and textually supported interpretation that merits deference pursuant to Martin in the face of a competing and inconsistent interpretation by the Commission.

B. Interpretation of 29 C.F.R. § 1910.212(a)(1)

The regulation at issue provides:

Types of guarding. One or more methods of machine guarding shall be provided to protect the operator and other employees in the machine area from hazards such as those created by point of operation, ingoing nip points, rotating parts, flying chips and sparks. Examples of guarding methods are — barrier guards, two-hand tripping devices, electronic safety devices, etc.

29 C.F.R. § 1910.212(a)(1).

Like the ALJ, the majority concludes that “hazards such as those created by point of operation, ingoing nip points, rotating parts, flying chips and sparks” applied only to hazards in the form of ejected debris and not ejected workpieces. The majority also concludes the regulation applies only to machines in the normal course of operation, ejection of an actual workpiece could occur only in the event of a malfunction, and therefore, the regulation should not apply. The Secretary argues this same language contains no inherent limitation to protections only against ejected debris rather than workpieces and no inherent limitation to situations involving normal machine operation rather than machine malfunctions.

In reviewing the interpretation of a regulation, meaning should be accorded to all terms, and “[a]ny interpretation of [an OSHA regulation] generally should conform to the accepted rules of grammar.” Solis, 558 F.3d at 823-24. As an initial matter, the list “point of operation, ingoing nip points, rotating parts, flying chips and ■ sparks” is preceded by the phrase “hazards such as those created by.” Because the phrase preceding the list employs the *948term “created by,” the list necessarily refers to items or conditions that cause the hazards, rather than merely and narrowly setting forth an enumeration of actual hazards. When meaning is accorded to the phrase “created by,” it becomes apparent that the potential class of hazards covered by the regulation necessarily is larger than the enumerated causes. Each cause for a hazard could give rise to several different actual hazards. In conducting its analysis, the majority fails to fully account for this broad language. Supra at 940 (referring to the list as “five examples of hazards”).

Second, because the phrase preceding the list uses the term “such as,” the list is exemplary and' not exhaustive. Orion Fin. Corp. of S.D. v. Am. Foods Grp., Inc., 281 F.3d 733, 739 (8th Cir.2002) (“An objective reader would interpret the phrase ‘such as’ to mean ‘for example.’ ”); Donovan v. Anheuser-Busch, Inc., 666 F.2d 315, 327 (8th Cir.1981) (“The phrase ‘such as’ is not a phrase of strict limitation, but is a phrase of general similitude indicating that there are includable other matters of the same kind which are not specifically enumerated by the standard.”). This fact alone suggests it may be inappropriate to engage in a hypertechnical analysis as done by the majority when citing materials outside the regulation to limit the meanings of express terms and arrive at a narrow construction (and not merely to illustrate an inconsistent interpretation). See Donovan, 666 F.2d at 327 (concluding that use of the phrase “such as” required the court to interpret an OSHA standard as reaching beyond the enumerated items to cover other, similar items “of the kind specified”).

Turning to the exemplar list of items set forth as nonexclusive causes for hazards, the regulation itself defines “point of operation” as “the area on a machine where work is actually performed upon the material being processed.” 29 C.F.R. § 1910.212(a)(3)(i). The lathes at issue have several “rotating parts,” and the workpiece itself rotates in tandem with those parts when affixed to the machine— such is the essence of a lathe. The “point of operation” is the workpiece when rotating in tandem with the lathe. Accordingly, even without reaching the reference to flying chips and sparks, there exist two grammatically simple and clear routes that support finding the regulation applicable to the present case: the danger associated with a workpiece being ejected from the lathe is a “hazard[ ] such as [that] created by point of operation [or] rotating parts.”

Focusing on the language “rotating parts,” the majority concludes that, because a clear and undisputed danger of rotating parts is the possibility of a worker being crushed or pulled when a body part, hair, or clothing is caught on the rotating part, the term cannot also refer to danger in the form of ejected parts. To support its interpretation, the majority looks outside the regulation itself at OSHA’s online interpretative guidance. This source, however, does not purport to identify an exclusive type of danger associated with rotating parts. Rather, as quoted by the majority, this source simply stresses that even slowly rotating parts can be dangerous. The apparent need to stress this fact indicates that the same source implicitly recognizes — or assumes the clarity of — risk associated with rapidly rotating parts. In any event, the ejection of a workpiece is merely the consequence of inadequate connections and the centrifugal force associated with rapid rotation. Under the simplest and most straightforward interpretation of the regulation itself, and without resort to sources outside the regulation, the phrase “hazards ... created by ... rotating parts” can refer to items ejected from a rapidly rotating lathe. Nothing about the agency’s online guidance and a reference to additional risks from slowly rotating parts suggests *949exclusivity. And, the Secretary’s broad interpretation recognizes the regulation’s broadening language “such as” and “created by.”

Finally, use of the expansive language “such as” to indicate an exemplary rather than an exhaustive list comports with the undisputed purpose of the regulation: “to ‘assure so far as possible every working man and woman in the Nation safe and healthful working conditions.’ ” Donovan, 666 F.2d at 327 (quoting 29 U.S.C. § 651(b)); Arkansas-Best Freight Sys., Inc. v. Occ. Safety & Health Review Comm’n, 529 F.2d 649, 653 (8th Cir.1976) (“The legislative decision has been made to protect the health of employees even though increased production costs may result.”). The court in Donovan concluded that a “restrictive” interpretation of a term in an OSHA regulation would not be consistent with the broad and protective statutory purpose but that the regulatory interpretation “should extend to those [situations] which in the reasonable judgment of the Secretary need protection from injury by guardrails.” Donovan, 666 F.2d at 327. While this broad statement of purpose is by no means conclusive, the consistency between this broad purpose and the plain text as urged by the Secretary further demonstrates the reasonableness of the Secretary’s interpretation.

To reach the opposite conclusion, the majority appears to employ three tools. First, the majority identifies an unwritten distinction within the exemplar list of items that would separate the regulation’s list into distinct and mutually exclusive groups. Second, the majority identifies an unwritten distinction between protections required in common, everyday situations and protections required to guard against less common hazards. And third, the majority relies upon the Second Circuit’s opinion interpreting the regulation, Carlyle Compressor Company v. Occupational Safety & Health Review Commission, 683 F.2d 673 (2d Cir.1982).

The majority’s subdivision of the regulation’s list suffers from a lack of express textual support and fails to accord meaning to the regulation’s use of exemplar language. The majority concludes, “These five hazards create two distinct categories: sources or causes of the hazards (point of operation, ingoing nip points, and rotating parts) and by-products from routine operation of the machinery (flying chips and sparks).” Supra, at 940. It appears the majority concludes that the express listing of “flying chips and sparks” impliedly excludes from the list any other form of ejected material even if that other ejected material is a “hazard[ ] ... created by” the “point of operation, ingoing nip points, [or] rotating parts.”

This reading of the enumerated list as creating mutually exclusive categories is inconsistent with the balance of text that clearly indicates the list is exemplar and nonexclusive. This distinction is also curious, not only for its absence from the actual text, but in its creation of an inherently unworkable and apparently outcome-determinative dividing line acknowledged nowhere in the majority’s opinion. There can be no dispute that the regulation requires protection against hazards such as those created by “flying chips.” Certainly, then, a dividing line must exist somewhere between very small flying chips (triggering worker protection under the majority’s interpretation of the regulation) and larger pieces of flying debris (conclusively precluding application of the regulation). In either case, the debris is, in fact, a portion of the blank, i.e., the workpiece. I find no actual support for any method that might be used to identify this dividing line. I certainly find no support containing sufficient nuance to define the size limit at which a “chip” no longer invokes the need *950for a protective device under the regulation.

Nevertheless, the majority labels the Secretary’s straightforward interpretation as strained, unsupported by common sense, and hyperteehnical. All of these labels ring hollow given the majority’s unjustified and extra-textual division of the list into two discrete and mutually exclusive categories. In fact, I find just the opposite to be true; the majority’s interpretation which requires the invocation of outside materials and a discounting of the regulation’s broadening terms appears to be a hyperteehnical and unnatural reading that, departs from our general rules of construction.

C. Carlyle and the Normal Operation/Malfunction Distinction

Because the majority’s interpretation enjoys little support in the actual text of the regulation, the result today would appear to rest almost entirely on the history of agency acquiescence following the Carlyle opinion. This inertial support, however does not speak in any way to the question of whether the Secretary’s interpretation is textually sound. Moreover, even if it were appropriate to deny the Secretary flexibility and force adherence to past inaction, continued reliance on Carlyle is misplaced for two reasons. First, Carlyle is fundamentally flawed in that it rests upon an invented distinction between normal and abnormal operations and ignores the need to guard against harm arising from foreseeable and often-repeated machine malfunctions or human errors. Second, Carlyle simply does not apply to the present case because the record indicates fairly clearly that the ejection of workpieces was a common occurrence at Loren Cook that cannot truthfully be labeled “abnormal.”

In Carlyle, the Second Circuit addressed a situation involving a machine that held and rapidly rotated a shaft so that the shaft could be subjected to grinding. 683 F.2d at 674. There, the court interpreted the language of the regulation narrowly, found the. regulation inapplicable to a thrown workpiece, and recognized a distinction between “normal projectiles” and “abnormal projectiles.” Id. at 675 (“The ALJ apparently interpreted ‘flying chips’ to include shafts thrown by the machine .... [But h]ere, the standard is directed at the hazards attendant upon the wastage created by more normal projectiles such as flying chips and sparks, rather than abnormal projectiles such as flying workpieces.” (emphasis added)). The Court in Carlyle, nevertheless, found that the absence of a machine guarding device comprised a violation of a more general regulatory section regarding unsafe working conditions. See id. at 676 (“Despite our conclusion that the specific duty clause does not apply, we hold that the Commission properly determined that Carlyle violated the general duty clause, 29 U.S.C. § 654(a)(1) (1976), by not eliminating the danger of shafts expelled from the grinder.”).

In reaching a narrow and limited interpretation of the regulation, the Carlyle court rested its analysis upon an invented distinction between normal operations on the one hand (producing normal projectiles), and abnormal situations or machine malfunctions, on the other (producing abnormal projectiles). This distinction not only lacks support in the regulation’s text, it fails to comport with the common sense need for protection from machine malfunction, unsafe operating practices, operator error, or inadvertent mistake — often the primary sources of risk in the workplace. See Signode Corp., 4 BNA OSCH 1078, *2 (No. 3527, 1967) (“One purpose of the Act is to prevent accidents.... Although there *951is little chance of an injury if the machines are operated properly, the standard is plainly intended to eliminate danger from unsafe operating procedures, poor training, or employee inadvertence.” (internal citations omitted)); see also Donovan, 666 F.2d at 327-28 (addressing a regulation requiring railings on “platforms” and rejecting a limiting interpretation that would have excluded from coverage raised surfaces accessed by workers only on an intermittent rather than regular basis). If such a distinction were to be recognized consistently, it would beg the question of why access to confined spaces is carefully regulated or why workers at heights must wear harnesses. Walls collapse, workers fall, and workpieces fly from lathes. The purpose of OSHA is thwarted if protections extend only to workplace situations involving perfectly functioning machines and error-free workers.

Turning to the record, several employees testified as to the frequency with which the Loren Cook lathes ejected entire workpieces. While some dispute remains as to the exact percentage of workpieces ejected7, it is undisputed that the accident resulting in the fatality was not an isolated or unexpected event. As noted, guards are still employed on the large lathes, the small lathes for years employed such guards, and at least one worker re-installed a guard on his small lathe after the fatality (before a supervisor removed the. guard “for inspection” and before all such small-lathe guards disappeared). And, even Loren Cook’s work instruction manual stated, “The first rule of manual spinning lathe operations is CAUTION. This is one of the most difficult operations performed at Loren Cook Company. The very nature of manual spinning lathe equipment possesses the potential for accidents.” (emphasis in original).

The interpretation favored by the majority, then, enjoys historic inertia rooted in a flawed Second Circuit opinion easily distinguishable from the present case. It does not enjoy a claim to being the more reasonable interpretation of the regulation. Further, in this case, the majority employs rhetorical tools to “double down” on the normal operation/malfunction distinction, repeatedly characterizing ejection of the workpiece as a “catastrophic failure” in an effort to illustrate a difference between a rotating part slipping from the lathe and some other piece of material or chip or spark being ejected from the lathe.8

III. Conclusion

Because Martin makes clear the Secretary is specifically qualified and empowered to amend its own interpretations over time, I believe it is important to guard against an overeagerness to declare a new interpretation unreasonable. In Martin itself, the Supreme Court acknowledged *952that consistent application of an interpretation is “a factor bearing on the reasonableness of the Secretary’s position.” Martin, 499 U.S. at 157, 111 S.Ct. 1171. And, I do not deny that the Secretary’s failure to put forth the present interpretation earlier or in a different format creates notice concerns. Id. at 158, 111 S.Ct. 1171 (“[T]he Secretary’s interpretation is not undeserving of deference merely because the Secretary advances it for the first time in an administrative adjudication[, but] the decision to use a citation as the initial means for announcing a particular interpretation may bear on the adequacy of notice to regulated parties.).

The question that remains, however, is how we should address the notice concerns in this case. If the Secretary’s interpretation of the regulation were somehow extra-textual or strained, I likely would agree with the majority. As set forth at length above, however, the Secretary’s interpretation is not strained. The Secretary’s interpretation comports with the plain language of the statute, gives effect to the language “created by,” and interprets the phrase “such as” according to our normal construction of language setting forth exemplary lists. It does so without confusing the process of textual analysis with the process of examining a pattern of prior interpretations. Therefore, I would take the Supreme Court at its word and view consistency with past practice and interpretation as “a factor” rather than — as the majority advocates — the controlling factor or the only factor in assessing the reasonableness of an interpretation. In other words, the Secretary’s present advocacy of a different interpretation is not impermissible or per se unreasonable, although it may “bear on the adequacy of notice to regulated parties.” Id.

In conclusion, I find nothing about Carlyle or the Secretary’s past enforcement of the regulation sufficient to demonstrate that the Secretary’s current, plain-language interpretation is unreasonable. As such, I would defer to the Secretary rather than the Commission, allow the Secretary’s interpretation to stand, and address notice concerns on remand.

I would grant the petition for review, reverse the order of the Commission, and remand for further proceedings consistent with this opinion.

. After the May 2009 fatality, at least one lathe operator reattached a guard to, his small lathe. A Loren Cook supervisor • questioned the operator about the guard and later removed it. This guard, and other guards that previously had been used on small lathes, were purportedly removed for inspection. The guards, however, could not be located when demanded by the Secretary in this matter. The Secretary moved for sanctions alleging spoliation of evidence. The ALJ denied the motion, but stated he was "troubled by the disappearance of the guards.”

. The majority, supra at 941, states, "These guarding devices would do little to prevent the hazard for which the Secretaty cited Loren Cook: the high speed ejection of a workpiece nearly 3 feet in diameter and weighing 12 pounds.” This conclusion appears to ignore the fact that such guards on large lathes protect against the ejection of even larger workpieces. At a minimum, the majority’s commentary on feasibility addresses an issue not reached by the ALJ.

. The Court in Martin stated:

[B]y virtue of the Secretary’s statutory role as enforcer, the Secretary comes into contact with a much greater number of regulatory problems than does the Commission, which encounters only those regulatory episodes resulting in contested citations. Consequently, the Secretary is more likely to develop the expertise relevant to assessing the effect of a particular regulatory interpretation.

499 U.S. at 152-53, 111 S.Ct. 1171 (emphasis added) (internal citation omitted).

. Police reports of employee interviews after the fatality suggested as many as 20% of workpieces detached from lathes during operation. This number seems facially suspect given the low number of actual injuries and given that the percentage would appear to be an economically infeasible amount of waste. In any event, neither party suggests the accident in this case was an isolated event.

. The consequences of the ejection of the workpiece in this case were, no doubt, catastrophic. The majority's characterization of a workpiece ejection itself as catastrophic, however, suggests the ejection of a workpiece is somehow an isolated or highly unusual event that renders the lathe unusable. The facts as referenced above demonstrate otherwise and suggest appropriate labeling for the different types of ejected materials might be "commonly ejected materials” and "less commonly ejected materials.” In any event, the choice of the language "catastrophic” to distinguish the ejection of a workpiece from the ejection of tooling shavings disregards the record regarding the potential frequency of workpiece ejections.