The National Mining Association and the Alabama Coal Association ("NMA") dispute a finding of the Mining Safety and Health Administration ("MSHA") that allows testing the amount of coal dust in mines by using measurements taken over a single shift, rather than traditional multi-shift measurements. NMA challenges the new sampling method on substantive and procedural grounds. We vacate the finding.

One of the reasons Congress passed the Federal Coal Mine and Safety Act ("the Coal Act") in 1969 was to reduce the amount of coal dust inhaled by coal miners. The dust was known to cause Black Lung Disease. The Coal Act provided interim standards for the maximum amount of coal dust permitted in coal mines as well as guidance on how to measure the level of coal dust in a mine's atmosphere. The interim standards were effective until the Secretaries1 created improved health standards. Relevant provisions of the Coal Act were re-enacted in the Federal Mine Safety and Health Act of 1977 ("the Mine Act"). See 30 U.S.C. §§ 801-962 (1994).

This dispute revolves around several provisions of the Mine Act. Under 30 U.S.C. § 841(a) the Secretary has authority to supersede the "interim mandatory health and safety standards" of the Mine Act with "improved mandatory health and safety standards." But the Secretary must enact the new standards according to the provisions of § 811. See 30 U.S.C. § 811(a). Section 811(a)(6) is at the heart of the current controversy. It states the Secretary "shall set standards" that adequately assure, on the basis of the "best available evidence" that no miner will suffer "material impairment of health" under the new standard and that the Secretary shall also consider the "latest available scientific data in the field, the feasibility of the standards, and experience gained under this and other health and safety laws."

Other pertinent provisions of the Mine Act include § 842(b)(2) which requires that the "average concentration" of coal dust to which a miner is exposed during each shift not exceed 2.0 milligrams per cubic meter of air (2.0 mg/m3). Average concentration is defined as a concentration that

accurately represents the atmospheric conditions with regard to respirable dust to which each miner ... is exposed ... over a single shift only, unless [the Secretary] finds in accordance with ... Section 811 ... that such single shift measurement will not, after applying valid statistical techniques to such measurement, accurately represent such atmospheric conditions during such shift.

30 U.S.C. § 842(f).

In 1971, MSHA's predecessor, the Bureau of Mines, proposed a finding that single-shift sampling would not accurately represent the atmospheric conditions of a mine. See 36 Fed.Reg. 13286 (1971). The proposed finding was made final in 1972. See 37 Fed.Reg. 3833 (1972). MSHA now wishes to rescind the 1971/72 finding and to begin single-shift sampling.

In attempting to rescind the 1971/72 finding, MSHA published two notices in the Federal Register. The first, published in February 1994, stated MSHA's plan to rescind the 1971/72 finding and replace it with a single, full-shift measurement of the atmospheric conditions. See 59 Fed.Reg. 8357 (1994). The second, published simultaneously, stated that citations would be issued based on single-shift sampling. See 59 Fed.Reg. 8356 (1994).

Single-shift sampling--in part--grew out of MSHA's Spot Inspection Program ("SIP"), itself designed to defeat suspected tampering of dust samples by mine operators. See 63 Fed.Reg. 5664, 5667 (1998). After the SIP, MSHA concluded that multi-shift sampling was inaccurate because multi-shift sampling did not lead to citations in places where the SIP had shown miners to be overexposed. See id. at 5668. The Federal Mine Safety and Health Review Commission, however, vacated citations issued under the SIP because of MSHA's failure to comply with the rulemaking procedures in § 811. See Secretary of Labor v. Keystone Coal Mining Corp., 16 FMSHRC 6 (1994).

Another reason given by MSHA for rescinding the 1971/72 finding is the improvement in air sampling technology. See 63 Fed.Reg. 5664, 5666 (1998). Since 1971, significant improvements have been made to calibration procedures, weighing accuracy, and sampling pumps. See id.

The accuracy of single-shift sampling is hotly debated by the parties. NMA argues that single-shift sampling is so inaccurate that a large number of citations will be erroneously issued to coal mine operators. MSHA counters that single-shift measurements are more accurate because they tend to expose spatial or temporal peaks in dust levels that would, under a multi-shift measurement, be masked by some measurements below the 2.0 mg/m3 threshold when averaged with the peak values. See id. at 5689. MSHA supports this conclusion by pointing out that multi-shift measurements were always highest during the first measured shift: it was only after the first shift, says MSHA, that operators had time to affect dust production. See id. at 5668.

Because of this debate, the period for public comment was extended several months, and two public hearings were held about the notices. See, e.g., 61 Fed.Reg. 18158 (1996). As a result of the comments, MSHA defined "accurately represent[ ]" (as used in 30 U.S.C. § 842(f)), re-opened the comment period, and held a public hearing on the new definition. See 61 Fed.Reg. 10012, 10013 (1996). In February 1998, MSHA issued the subject of our review, the Joint Finding and Noncompliance Determination Notice ("the Joint Finding") which rescinded the 1971/72 finding. See 63 Fed.Reg. 5664 (1998).

NMA raises procedural objections under the Mine Act, the Administrative Procedure Act ("APA"), and the Regulatory Flexibility Act ("RFA"), and substantive objections to the Joint Finding. We will address only the procedural objections.

NMA says that the procedural requirements of the Mine Act, in 30 U.S.C. § 811, were not met by MSHA's Joint Finding. MSHA makes two arguments in response. First, the use of single-shift measurements is no mandatory health and safety standard and, therefore, does not need to comply with § 811. Second, if the Joint Finding is a mandatory health and safety standard, MSHA argues, the Joint Finding complied with the procedural requirements of § 811. In arguing that the Joint Notice complied with § 811, however, MSHA insists that portions of § 811 do not contain procedural requirements.

An agency's interpretation of its governing statute is often given significant deference. See Chevron, U.S.A., Inc. v. Natural Resources Defense Council, 467 U.S. 837, 842-43, 104 S. Ct. 2778, 81 L. Ed. 2d 694 (1984). But, when applying Chevron 's first step, we do not need to defer when the issue is a "pure question of statutory construction." See INS v. Cardoza-Fonseca, 480 U.S. 421, 446, 107 S. Ct. 1207, 94 L. Ed. 2d 434 (1987). Likewise, we need not defer to issues beyond the agency's expertise. See Morris v. CFTC, 980 F.2d 1289, 1293 (9th Cir.1992); see also Colorado Public Utils. Comm'n v. Harmon, 951 F.2d 1571, 1579 (10th Cir.1991) (not deferring on issue of preemption); Lynch v. Lyng, 872 F.2d 718, 724 (6th Cir.1989) (not deferring on issue of statute's effective date).

Because deciding if MSHA must address the requirements of § 811(a)(6) is a question of pure statutory construction, we need not defer to MSHA's interpretation. We conclude that MSHA's various interpretations of § 811(a)(6)--as we shall explain--are incorrect.

Use of single-shift measurements by MSHA is a health and safety standard. Mandatory health and safety standard is defined, in § 802(l ) as "the interim mandatory health or safety standards" between § 841 and § 846. Section 842(f) is the basis for single-shift sampling. Furthermore, § 841(a) refers to §§ 842-846 as "interim mandatory health standards." At a minimum, therefore, § 842(f) is an interim mandatory health standard. § 841(a) continues, however, to say that the interim mandatory health standards of §§ 842-846 are effective "until superseded in whole or in part by improved mandatory health standards." Single-shift sampling supersedes multi-shift sampling, which was based on § 842(f). Single-shift sampling, therefore, is an "improved mandatory health standard." See United Mine Workers v. Dole, 870 F.2d 662, 671 (D.C.Cir.1989) (the term "mandatory standard" includes standards adopted to replace an existing mandatory standard); id. at 672 (concluding § 811(a)(9) is a mandatory standard). According to § 841(a), any new standard must be "promulgated ... under the provisions of Section 811."2

The reasoning of the Federal Mine Safety and Health Review Commission ("the Commission") in Secretary of Labor v. Keystone Coal Mining Corp., 16 FMSHRC 6, 13 (1994) supports our conclusion that MSHA's new sampling method is a mandatory health standard. In Keystone, the Commission rejected MSHA's argument that single-shift measurements did not require following § 811 procedures. Section 842(f), said the Commission, explicitly requires MSHA to follow § 811 procedures if the Secretary decides not to use single-shift measurements. This intent--to use § 811 procedures if rejecting single-shift measurements--"bespeaks an equal intent that, once such a finding is made, it may be rescinded only" by following § 811 procedures. Id.

MSHA next argues that it did comply with § 811, but that MSHA must only comply with the procedure-setting portions of § 811. MSHA says § 811(a)(6) contains no procedure-setting provisions.3 We think MSHA's interpretation is incorrect.

The plain language of § 842(f) and 841(a) requires mandatory health or safety standards to be made "under" or "in accordance with" the "provisions of section 811." No restriction suggests that MSHA must comply only with the procedures in § 811. Where Congress sought to refer only to the procedural aspects of § 811, it did so clearly. See 29 U.S.C. § 811(b)(2) ("A temporary mandatory health or safety standard shall be effective until superseded by a mandatory standard promulgated in accordance with the procedures prescribed in [§ 811(a)(3) ].") (emphasis added).

Our conclusion using the statute's plain meaning is supported by three additional points. First, § 811(a)(6) says that MSHA shall consider the feasibility of the standards. The language is not discretionary. Second, MSHA, in more recent rulemakings, recognizes the requirement to address feasibility. See 63 Fed.Reg. 17492, 17558 (1998) (addressing feasibility of proposed rule on diesel engine exhaust in mines).4 Third, MSHA is reversing its prior policy on sampling. Proper procedures are particularly important where, as here, MSHA's predecessor studied and rejected single-shift sampling.

To use single-shift measurements, then, MSHA must follow all the provisions of § 811. We conclude MSHA has not done so.

§ 811 requires notice, the opportunity for public comment, public hearings if requested, and final publication in the Federal Register. There can be little doubt, as detailed in the facts above, that MSHA satisfied these requirements. But as we have explained, MSHA must also satisfy the requirements of § 811(a)(6). Therefore, MSHA must demonstrate that the new standard (a) adequately assures that no miner will suffer a material impairment of health, on the basis of the best available evidence; (b) uses the latest available scientific data in the field; (c) is feasible;5 and (d) is based on experience gained under the Mine Act and other health and safety laws. See 30 U.S.C. § 811(a)(6)(A).

After a review of the record, we conclude that the record contains no finding of economic feasibility. The absence of a showing of economic feasibility is not surprising because MSHA insisted, in the Joint Finding, that "there is no need to address feasibility." 63 Fed.Reg. 5664, 5669 (1998).6

We conclude, therefore, that MSHA failed to comply with § 811(a)(6) of the Mine Act. So we must vacate the Joint Finding.7

VACATED.