Collins v. National Transportation Safety Board

Notice: This opinion is subject to formal revision before publication in the Federal Reporter or U.S.App.D.C. Reports. Users are requested to notify the Clerk of any formal errors in order that corrections may be made before the bound volumes go to press. United States Court of Appeals FOR THE DISTRICT OF COLUMBIA CIRCUIT Argued October 10, 2003 Decided December 19, 2003 Nos. 02-1298 & 03–1001 THOMAS H. COLLINS, PETITIONER v. NATIONAL TRANSPORTATION SAFETY BOARD, RESPONDENT JOHN NITKIN, INTERVENOR On Petitions for Review of Orders of the United States Department of Transportation John S. Koppel, Attorney, U.S. Department of Justice, argued the cause for petitioner. With him on the briefs were Peter D. Keisler, Assistant Attorney General, Roscoe C. How- ard, Jr., U.S. Attorney, and Robert S. Greenspan, Attorney. Bills of costs must be filed within 14 days after entry of judgment. The court looks with disfavor upon motions to file bills of costs out of time. 2 Michael E. Robinson, Attorney, U.S. Department of Justice, entered an appearance. Andrew W. Anderson argued the cause and filed the brief for intervenor. Before: RANDOLPH, ROGERS, Circuit Judges, and WILLIAMS, Senior Circuit Judge. Opinion for the Court filed by Senior Circuit Judge WILLIAMS. WILLIAMS, Senior Circuit Judge: A Coast Guard adminis- trative law judge found that Captain John Nitkin committed misconduct by failing to sound the warning signal—five whis- tle blasts—required by treaty, specifically Rule 34(d) of the 1972 International Regulations for Preventing Collisions at Sea, 28 U.S.T. 3459 (‘‘COLREGS’’), when a collision threat- ens. The Commandant of the Coast Guard affirmed the ALJ’s decision, but the National Transportation Safety Board reversed, finding that the text of Rule 34(d) rendered the rule inapplicable to the circumstances of Nitkin’s accident. We reverse and remand. * * * On January 29, 1999 Captain Nitkin was serving as the pilot of the S/S Chelsea when it collided with the M/V Manzanillo in the Miami Harbor Channel. The two vessels had initially agreed to pass one another starboard-to- starboard, but as they approached each another, Captain Fernandez, the pilot of the Manzanillo, radioed Nitkin and announced his intention to attempt a port-to-port passing. Nitkin radioed Fernandez that a port-to-port passing was not possible and urged him to follow their original agreement by turning to port. But the Manzanillo in fact turned to star- board and, despite Nitkin’s efforts to maneuver to safety, the two ships collided about 21/2 minutes after the Manzanillo began its starboard turn. At the time of the accident Nitkin was operating under the authority of a Coast Guard pilot’s license. He was tried in a disciplinary proceeding before a Coast Guard ALJ for negli- gence and for violations of COLREGS Rule 14 (failure to turn 3 to starboard in a head-on meeting situation), Rule 8(e) (failure to reduce speed or reverse engines), and Rule 34(d) (failure to sound warning signal). The ALJ dismissed the negligence charge and the Rule 14 and 8(e) misconduct charges, but found that Nitkin committed misconduct by failing to comply with Rule 34(d). That rule provides: When vessels in sight of one another are approaching each other and from any cause either vessel fails to understand the intentions or actions of the other, or is in doubt whether sufficient action is being taken by the other to avoid collision, the vessel in doubt shall immedi- ately indicate such doubt by giving at least five short and rapid blasts on the whistle. COLREGS, 28 U.S.T. 3459, Part B, Section I, Rule 34(d). The ALJ suspended Nitkin’s license for five months, with four months remitted on probation. ALJ Final Order at 2. Nitkin appealed to the Commandant of the Coast Guard, raising four primary objections. Besides charging that the sanction was excessively harsh, Nitkin attacked the finding of a violation, arguing, first, that Rule 34(d) doesn’t apply when the danger of collision develops so late that the warning signal would be useless; second, that any duty he had to warn the Manzanillo of the collision risk was satisfied by his radio communications with Fernandez; and third, that his failure to sound the warning signal was excusable under COLREGS Rule 2(b)’s exception for ‘‘special circumstances TTT which may make a departure from these Rules necessary to avoid immediate danger.’’ According to Nitkin, special circum- stances existed in this case because the five-blast warning signal would have prevented communication with crew mem- bers at the Chelsea’s bow. The Commandant of the Coast Guard rejected all these claims and affirmed the ALJ’s decision. Appeal of Nitkin, 2001 WL 34080161. We need not address the Commandant’s reasoning, because although Nitkin raised essentially the same claims in his appeal to the NTSB under 49 U.S.C. § 1153, the Board reversed on quite different grounds and never reached the issues posed in Nitkin’s appeal to the Commandant. 4 Rather, the Board’s July 26, 2002 Opinion and Order, 2002 WL 1727347 (‘‘July 26 Order’’), rested on the following in- terrelated conclusions. First, the Board determined as a matter of law that Rule 34(d) is inapplicable in situations where a pilot is certain that the other vessel’s conduct makes collision inevitable. The Board explained that, because Rule 34(d)’s plain text specifies that the warning signal require- ment comes into play only when a pilot is ‘‘in doubt whether sufficient action [is] being taken by the [other vessel] to avoid collision’’ (emphasis added), the rule cannot apply where a pilot is certain that sufficient action is not being taken. July 26 Order, at 3–4 (quoting COLREGS Rule 34(d)). Second, the Board made a factual determination that Nit- kin was certain that the Manzanillo was not taking sufficient action to avoid a collision, rather than ‘‘in doubt’’ of the reverse. According to the Board, the ALJ had found that Nitkin ‘‘was not sure whether his vessel could avoid a collision after the unilateral decision of [Fernandez] to execute a port- side passing,’’ but the Board concluded that this finding was unsupported by the record. July 26 Order, at 2–3. The Coast Guard filed a notice of appeal (indeed, as we shall see, two) and we granted Nitkin’s motion to intervene. Before we reach the merits, we must consider whether our jurisdiction is undercut either by the possibility that the Coast Guard Commandant does not qualify as a ‘‘person’’ entitled to appeal under 49 U.S.C. § 1153(a), or by the Coast Guard’s prior filing of a request that the Board reconsider its July 26 Order. Jurisdiction. ‘‘Person.’’ Although the word ‘‘person’’ is usually pre- sumed not to include the sovereign, Vermont Agency of Natural Res. v. United States ex rel. Stevens, 529 U.S. 765, 780 (2000); United States v. Cooper Corp., 312 U.S. 600, 604 (1941), this presumption can be overcome by an affirmative showing of statutory intent to the contrary. See Vermont Agency of Natural Res., 529 U.S. at 781 (citing Int’l Primate Prot. League v. Administrators of Tulane Educ. Fund, 500 5 U.S. 72, 83 (1991)). Here 49 U.S.C. § 40102(a)(33) states that ‘‘ ‘person,’ in addition to its meaning under section 1 of title 1, includes a governmental authority,’’ and 49 U.S.C. § 1101 explicitly applies the definitions in § 40102(a) to the chapter containing 49 U.S.C. § 1153. Thus, the statute authorizes jurisdiction over an appeal by the Commandant, so long as he has a ‘‘substantial interest in the order,’’ 49 U.S.C. § 1153(a), as he clearly does. Timing. A petition for judicial review of a final NTSB order must be filed no later than 60 days after the order, see 49 U.S.C. § 1153(a), and a request for administrative recon- sideration must be filed within 30 days of the order, see Commandant v. Mintz, 4 N.T.S.B. 1976 (1984) (holding that the 30–day time limit established in the Rules of Practice for aviation proceedings, 49 C.F.R. § 821.50(b), should also apply to maritime proceedings). Here, the Coast Guard filed (1) an untimely request for administrative reconsideration on Au- gust 30, 2002 (35 days after the July 26 Order), which the NTSB dismissed as untimely on November 20, 2003, after the initial 60 days for seeking judicial review had run, (2) an initial petition for review on September 23, 2002 (59 days after the July 26 order), and (3) a second petition for review on January 3, 2003 (44 days after the Board’s dismissal of its request for NTSB reconsideration). Thus, the sequence of orders and petitions is as follows: 1 July 26, 2002: The Board issued its Order reversing the Coast Guard; 1 August 30, 2002: The Coast Guard requested administra- tive reconsideration by the Board; 1 September 23, 2002: The Coast Guard filed its first petition for judicial review; 1 November 20, 2002: The Board rejected the request for administrative reconsideration as untimely; 1 January 3, 2003: The Coast Guard filed its second petition for judicial review. Attacking our jurisdiction, Nitkin correctly notes that our cases treat a petition for review filed during the pendency of 6 a request for administrative reconsideration as ‘‘incurably premature,’’ see TeleSTAR, Inc. v. FCC, 888 F.2d 132, 133–34 (D.C. Cir. 1989), and in effect a nullity. As Nitkin acknowl- edges, the reconsideration request ordinarily tolls the running of the time limit for judicial review. Id. at 133. See also Outland v. Civil Aeronautics Bd., 284 F.2d 224, 227–28 (D.C. Cir. 1960); City of Pittsburgh v. Federal Power Comm., 237 F.2d 741, 749 (D.C. Cir. 1956). But, Nitkin argues, the Supreme Court said in Bowman v. Lopereno, 311 U.S. 262 (1940), that an untimely request for agency reconsideration could not extend the time for appeal (although it actually held that if the agency addressed the merits of such a request, its ultimate denial thereof would start a new period for seeking review), id. at 266. We can find jurisdiction without resolving the precise effect of the untimely request for NTSB reconsideration. If the request suspended the running of the time limit for appeal, then the Coast Guard’s second petition for review was timely; if it did not, then the initial petition was effective. Either way, we have jurisdiction. The only alternative—which Nitkin does not express in so many words, perhaps because of the embarrassment of doing so—is that the filing of an untimely request for administrative reconsideration renders the original agency order non-final for purposes of judicial review petitions filed before the agency dismisses the reconsideration request, but does not affect the finality of the original order for purposes of judicial review petitions filed after the dismissal of the reconsidera- tion request, so long as that dismissal is on procedural grounds. Thus, according to Nitkin’s theory, the July 26 order was non-final for the 82–day period between August 30 and November 20, 2002, but became final, retroactively, the moment the Board dismissed the untimely reconsideration request on November 20. This theory of finality makes little sense, and serves no purpose other than to create traps for the unwary. It is clearly not compelled by Bowman, as nothing in that case requires that the July 26 Order be regarded as non-final 7 during the 60 days after its issuance, or indeed at any other time. We find the Coast Guard appeal timely. Merits. Our first question is the appropriate scope of review. Three elements complicate the standard picture. First, the COLREGS are established by treaty rather than an ordinary statute. Second, their enforcement is divided vertically, in the sense that the Coast Guard acts as general regulator, as prosecutor, and as initial adjudicator of claims of violation, with the NTSB initially reviewing Coast Guard adjudications before they come before a court of appeals. Finally, the COLREGS’ enforcement is divided horizontally, as the Coast Guard shares responsibility for initial enforcement with the Navy and the states. We tackle these issues seriatim. That the COLREGS are established by treaty does not have a substantial impact on our analysis. We have held that treaty interpretation should be ‘‘guided by principles similar to those governing statutory interpretation,’’ Iceland S.S. Co., Ltd.-Eimskip v. Dept. of Army, 201 F.3d 451, 458 (D.C. Cir. 2000), and that courts should give great weight to the views of those executive agencies charged with enforcing particular treaties. See id. (citing Sumitomo Shoji America, Inc. v. Avagliano, 457 U.S. 176, 184–85 (1982)); Hill v. Norton, 275 F.3d 98, 104 (D.C. Cir. 2001) (applying Chevron framework to agency’s interpretation of treaty and implementing statute); cf. Collins v. Weinberger, 707 F.2d 1518, 1522 (D.C. Cir. 1983) (pre-Chevron case deferring to executive interpretation of treaty). The vertical split of enforcement clouds the matter, as here we have two conflicting interpretations of Rule 34(d) from two different executive agencies. According to the Board, ‘‘doubt’’ denotes uncertainty, and excludes certainty of the negative. In contrast, the Coast Guard proposed the follow- ing conclusion of law which was accepted by the ALJ: In regards [sic] to the failure of Captain Nitkin to sound the ‘‘doubt’’ signal as required by International Naviga- tion Rule 34(d), the Rule and the corresponding guidance is clear. LLANA and WISNESKEY [Christopher B. 8 Llana & George P. Wisneskey, Handbook of the Nautical Rules of the Road (1991)] comment: Paragraph (d) describes the ‘‘doubt’’ signal, also re- ferred to in the Inland Rules as the ‘‘danger’’ signal. The signal is five or more short and rapid blasts, which may be supplemented by a light signal. Give the signal as soon as you are in doubt about the action of another approaching vessel—when you don’t know what the other vessel is doing or when you think it is doing the wrong thing. ALJ decision at 35 (emphasis added). Thus, the Coast Guard interpreted ‘‘doubt’’ about the sufficiency of the other vessel’s actions to include not only cases where one is uncertain whether the other vessel’s actions are sufficient, but also cases where one is certain that those actions are not suffi- cient. The Coast Guard’s view, though probably the less frequent meaning, appears to be among the ordinary usages. Webster’s Third New International Dictionary (1981), for instance, includes among the meanings of doubt ‘‘an inclina- tion not to believe or accept,’’ which is at least akin to certainty of the negative. Assuming that some deference to the executive is proper in this case, the next question is which executive branch agen- cy—the Coast Guard or the NTSB—is entitled to it. Martin v. Occupational Safety & Health Rev. Comm’n, 499 U.S. 144 (1991), addressed a similar split-enforcement regime under the Occupational Safety and Health (‘‘OSH’’) Act and held that deference was owed to the interpretation of the primary executive branch enforcer (the Secretary of Labor) rather than to that of the independent review board (the Occupation- al Safety and Health Review Commission). The Court saw Congress as intending ‘‘to make a single administrative actor [the Secretary of Labor] ‘accountable for the overall imple- mentation’ of the Act’s policy objectives by combining legisla- tive and enforcement powers’’ in a single entity, id. at 156, a goal that would be marginally frustrated if the Commission could substitute its own (reasonable) interpretations for those of the Secretary, id. And the adjudicative administrative 9 reviewing agency (there, the Commission) could adequately accomplish the purpose of avoiding biased review of enforce- ment decisions if it merely exercised ‘‘the type of nonpolicy- making adjudicatory powers typically exercised by a court in the agency-review context,’’ reviewing ‘‘the Secretary’s inter- pretations only for consistency with the regulatory language and for reasonableness.’’ Id. at 154–55. Although the Martin Court was careful to limit its holding to the OSH Act and to ‘‘take no position on the division of enforcement and interpretive powers within other regulatory schemes that conform to the split-enforcement structure,’’ id. at 158, we find the logic of Martin applicable to this case. Nitkin, indeed, has offered no reason to suppose that Con- gress did not here intend the same essential trade-offs as the Court found present in the parallel OSHA scheme. The final deference complication arises from the horizontal division of authority between the Coast Guard (which en- forces the COLREGS against pilots operating U.S.-flagged vessels pursuant to their Coast Guard pilot’s licenses), the Navy (which enforces the COLREGS against Naval officers), and the various state maritime commissions (which have exclusive authority to enforce the COLREGS against pilots of foreign-flagged vessels pursuant to their state pilot’s licens- es). See Soriano v. United States, 494 F.2d 681, 684 (9th Cir. 1974). We must therefore decide what effect, if any, this shared enforcement authority has on the appropriate level of deference to Coast Guard interpretations of the COLREGS. Nitkin urges no deference to the Coast Guard, pointing to case law in this circuit holding that deference under Chevron, U.S.A., Inc. v. Natural Resources Defense Council, 467 U.S. 837 (1984), is inappropriate when multiple agencies are re- sponsible for implementing the same statute. But many of these cases involved generic statutes that apply to dozens of agencies, and for which no agency can claim any particular expertise. See Ass’n of Amer. Physicians & Surgeons v. Clinton, 997 F.2d 898 (D.C. Cir. 1993) (interpretation of Federal Advisory Committee Act (‘‘FACA’’)); Prof’l Reactor Operator Soc’y v. N.R.C., 939 F.2d 1047 (D.C. Cir. 1991) 10 (interpretation of Administrative Procedure Act (‘‘APA’’)); FLRA v. D.O.T., 884 F.2d 1446 (D.C. Cir. 1989) (interpreta- tion of Privacy Act); Reporters Comm. For Freedom of the Press v. D.O.J., 816 F.2d 730 (D.C. Cir. 1987), rev’d on other grounds, 489 U.S. 749 (1989) (interpretation of Freedom of Information Act (‘‘FOIA’’)). Although some of these cases state the no-deference principle rather broadly, they seem as a whole easily distinguishable from the case at hand. Where a statute is generic, two bases for the Chevron presumption of implied delegation are lacking: specialized agency exper- tise and the greater likelihood of achieving a unified view through the agency than through review in multiple courts. See Rehabilitation Ass’n of Virginia, Inc. v. Kozlowski, 42 F.3d 1444, 1471 (4th Cir. 1994); Allnet Communication Ser- vice, Inc. v. Nat’l Exch. Carrier Ass’n, Inc., 965 F.2d 1118, 1120 (D.C. Cir. 1992). More relevant to Nitkin’s claim are our cases dealing with enforcement of the Federal Deposit Insurance Act (‘‘FDIA’’), 12 U.S.C. § 1811. See Proffitt v. FDIC, 200 F.3d 855 (D.C. Cir. 2000); Rapaport v. D.O.T., Office of Thrift Supervision, 59 F.3d 212 (D.C. Cir. 1995); Wachtel v. Office of Thrift Supervision, 982 F.2d 581 (D.C. Cir. 1993). These cases establish that, because the FDIA is administered by four separate agencies (the Comptroller of the Currency, the Board of Governors of the Federal Reserve Board, the Fed- eral Deposit Insurance Corporation, and the Office of Thrift Supervision in the Treasury Department), the interpretation of any one of them is not entitled to Chevron deference. The FDIA scheme is somewhat analogous to COLREGS enforcement, in that each of the agencies responsible for FDIA enforcement has jurisdiction over a different set of regulated parties. See 12 U.S.C. §§ 1813(q)(1)-(4). But un- der the FDIA division of responsibility, as the statute itself notes, ‘‘more than one agency may be an appropriate Federal banking agency with respect to any given institution.’’ 12 U.S.C. § 1813(q). This is a potentially important distinction between the FDIA and the COLREGS. Since the respective jurisdictions of the Coast Guard, Navy, and state commissions 11 are exclusive as to specific sets of mariners, there appears no danger that any one regulated party will be faced with multiple and perhaps conflicting interpretations of the same requirement. We may then reasonably distinguish three types of shared- enforcement schemes. For generic statutes like the APA, FOIA, and FACA, the broadly sprawling applicability under- mines any basis for deference, and courts must therefore review interpretative questions de novo. For statutes like the FDIA, where the agencies have specialized enforcement responsibilities but their authority potentially overlaps—thus creating risks of inconsistency or uncertainty—de novo re- view may also be necessary. But for statutes where expert enforcement agencies have mutually exclusive authority over separate sets of regulated persons, the above concerns don’t work against application of Chevron deference. A facet of COLREGS enforcement does, however, suggest that interpretive uniformity across agencies may be important enough to undermine the case for Chevron deference to one agency’s interpretation. Even if every pilot knows what is expected of him, it is likely to be helpful—in some cases perhaps vital—that he also know the obligations of other pilots. Vessels whose masters behave in accord with well established common rules are more likely to communicate clearly with each other, and thus avoid accidents across the enforcement agencies’ jurisdictions, than if a variety of rules creates a Babel. When vessels under Navy, Coast Guard and state supervision converge, five whistle blasts should mean the same thing. Even if Chevron deference is not called for, we think there is at a minimum a case for deference under Skidmore v. Swift & Co., 323 U.S. 134 (1944), given the Coast Guard’s expertise both in maritime safety and in deciding the most efficient way to administer its licensing and discipline procedures. If the three enforcement agencies were found to have conflicting (though individually very reasonable) interpretations, the var- 12 ied positions’ ‘‘power to persuade’’ would sharply fall. See id. at 140. We need not try to assess the exact weight of Skidmore deference, which is obviously less than Chevron, see United States v. Mead Corp., 533 U.S. 218, 236–39 (2001), but more than acknowledgement that the agency’s position is more convincing than its adversaries’, as would be true any time it submitted the more convincing brief. Assuming some non- trivial boost to the Coast Guard, its interpretation of Rule 34(d) is sufficiently persuasive that the NTSB should have upheld its view that the Rule 34(d) obligation to sound the warning signal applies even when a pilot is certain that ‘‘sufficient action’’ is not ‘‘being taken by the other [vessel] to avoid collision.’’ Indeed, it seems to us that it is precisely in such cases that sounding the warning signal is most important, absent special factors such as Nitkin here invoked and the Board did not reach. Suppose that, in a collision between Vessel A and Vessel B, all agreed that Vessel A knew that Vessel B’s present course would cause a collision. It cannot be the case that Vessel A’s failure to sound the warning signal would nonetheless be excused because there was certainty, rather than ‘‘doubt,’’ about the insufficiency of Vessel B’s conduct. Yet the Board’s interpretation would allow precisely this result. We note that the Board subtly shifted the focus of the rule—from the risk that B’s conduct was insufficient, plainly suggesting the likely advisability of a warning whistle—to the actual likelihood of a collision, a related but separate point, and one perhaps captured in the three substantive defenses that the Board never addressed. Of course we take no position on those defenses, which the board must consider on remand. *** The Board erred in failing to defer to the Coast Guard’s reasonable application of Rule 34(d) to cases where mariners are certain that ‘‘sufficient action’’ is not ‘‘being taken by the other [vessel] to avoid collision.’’ We reverse and remand the 13 case so that the Board can consider Nitkin’s other factual and legal arguments. So ordered.