BOWNES, Senior Circuit Judge.
Plaintiffs-appellants, the Massachusetts State Automobile Dealers Association, Inc. and two trade groups of automobile manufacturers, appeal from an order denying their request for a preliminary injunction. Plaintiffs seek to stall the implementation of motor vehicle tailpipe emissions regulations adopted by defendant-appellee, the Commissioner of the Massachusetts Department of Environmental Protection (DEP).
See
Mass. Regs.Code tit. 310, §§ 7.40-7.60. Defendant-appellee the American Petroleum Institute, intervened in support of the regulations.
Prior to oral argument, plaintiffs moved to dismiss their appeal as to all issues but one: whether DEP’s 1995 model year requirements should be enjoined. DEP opposes the motion for partial dismissal and requests costs and attorney’s fees. We grant the motion for partial dismissal. We award DEP costs, but not attorney’s fees. With respect to the 1995 model year requirements, the order of the district court is affirmed.
i.
BACKGROUND
A.
Cars and the Clean Air Act
The exhaust from a gasoline-powered engine is a source of air pollution.
Motor Vehicle Mfrs. Ass’n v. New York Dep’t of Envtl. Conservation,
17 F.3d 521, 524 (2d Cir.1994) (hereinafter
MVMA).
Emissions from car tailpipes include hydrocarbons and nitrogen oxides (NOx), constituents of ground-level ozone, a major component of smog.
Id.
at 526.
The Clean Air Act is the federal legislation governing tailpipe emissions. The Act directs the United States Environmental Protection Agency (EPA) to establish national ambient air quality standards (NAAQS) for pollutants such as ground-level ozone. Under the Act, states are responsible for developing and enforcing a plan, subject to EPA approval, for attaining and maintaining the NAAQS by regulating sources of air pollution. 42 U.S.C. § 7410(a). States failing to meet the NAAQS risk sanctions, including the loss of federal highway funds.
Id.
§ 7509. EPA has designated the entire state of Massachusetts as a “serious” nonattainment area for the ozone NAAQS.
See
56 Fed.Reg. 56,694, 56,776 (Nov. 6, 1991).
Mobile sources of air pollution such as cars and trucks are subject to EPA regulation under §§ 202 and 207 of the Act, 42 U.S.C. §§ 7521, 7541. EPA emissions standards for hydrocarbons and nitrogen oxides apply to a given vehicle based on its weight, use classification, and model year.
See id.
§§ 7521, 7541;
MVMA,
17 F.3d at 525-26.
State regulation of motor vehicle emissions is generally preempted by the Clean Air Act, 42 U.S.C. § 7543(a), with one exception: California can enforce its own standards, subject to EPA approval by way of a waiver under § 209(b) of the Act,
id.
§ 7543(b) (the waiver requirement). Consequently, there can be only two types of cars “created” under emissions regulations in this country: “California” cars and “federal” (that is, EPA-regulated) cars.
See id.
§ 7507. Other states cannot take any action that would force manufacturers to create a “third vehicle.”
Id.
(the third vehicle requirement).
Section 177 of the Act allows other states to adopt- standards “identical” to California’s (the identically requirement), but only if there is a two-year time lapse between the time the standards are adopted and the first model year affected by those standards (the leadtime requirement).
Id.
Similarly, § 211 of the Act authorizes EPA to regulate motor fuels and preempts any unapproved state regulations, except for California, which may enact fuel standards without EPA approval.
Id.
§ 7545(c)(4)(B).
B.
DEP’s Adoption of California LEV Regulations
In September 1991, California enacted a novel set of vehicle emissions and clean fuels requirements called the “Low Emissions Ve-hieles/Clean Fuels” (LEV/CF) program. The LEV component of the program requires the creation of four categories of California cars to meet increasingly stringent emissions standards, to be phased in over time: Transitional Low-Emission Vehicles; Low-Emission Vehicles; Ultra-Law-Emission Vehicles; and Zero-Emission Vehicles, such as electric cars. California has also established annually descending “fleet average requirements,” based on sales targets for each category of vehicles. A fleet average requirement is a cap on the average emissions attributable to all classes of vehicles produced by a particular manufacturer in a given year (in other words, the manufacturer’s “fleet”). California’s requirements provide manufacturers with “flexibility to develop varying emissions within their entire fleet to meet [an] overall goal.”
MVMA,
17 F.3d
at 535. On January 7, 1993, EPA granted California a § 209(b) waiver for the program.
Meanwhile, on January 31, 1992, DEP adopted the LEV component of California’s standards, intending to apply the standards beginning with 1995 models. DEP regulations allow new California cars to be leased, bought, sold, and registered in Massachusetts, but ban the acquisition, sale, and registration of new federal cars in the state. DEP’s proposed regulations sent out for notice and comment contained fleet average requirements, but no such requirements appear in the final rule because DEP preferred to let the market determine the mix of new California cars in the state.
C.
Prior Proceedings
Plaintiffs filed an action in the District Court for the District of Massachusetts, arguing that DEP’s regulations are preempted by the Act because DEP allegedly failed to comply with § 177 of the Act, 42 U.S.C. § 7507. Plaintiffs moved for summary judgment and for a preliminary injunction, founding their motions on four claims: [1] the regulations are not “identical” to California’s, in that DEP did not adopt California’s clean fuels rules; [2] the regulations force manufacturers to create a “third vehicle” because of the higher sulfur content of gasoline in Massachusetts; [3] the regulations were adopted by DEP before EPA granted California a § 209(b) waiver; and [4] the two-year leadtime requirement precluded DEP from applying the regulations to any 1995 models because two automakers planned to begin producing 1995 cars before two years passed after the regulations were adopted.
With the parties’ consent, the court stayed the summary judgment proceedings and ruled first on the motion for a preliminary injunction. The court denied the motion without a hearing, ruling that while plaintiffs demonstrated a risk of irreparable injury-given the cost of vehicle emissions controls, the balance of equities and the risk of harm to the public interest did not clearly favor granting an injunction. The court also found that plaintiffs failed to demonstrate a likelihood of prevailing on the merits, which is the
“sine qua non”
of the preliminary injunction test.
Weaver v. Henderson,
984 F.2d 11, 12 & n. 3 (1st Cir.1993).
Three of the four Clean Air Act issues presented to the district court were later addressed by the Second Circuit in a case concerning a challenge to New York’s adoption of the LEV standards.
See MVMA,
17 F.3d at 521,
aff'g in part and rev’g in part Motor Vehicle Mfrs. Ass’n v. New York Dep’t of Envtl. Conservation,
831 F.Supp. 57 (N.D.N.Y.1993) (hereinafter
New York DEC).
The Second Circuit held in favor of the state on the identicality and waiver claims, but held in favor of the automakers on the leadtime claim.
Id.
at 532-35. The court did not consider the merits of the “third vehicle” claim because the district court found material facts at issue and set the claim down for trial.
Id.
at 530.
II.
PARTIAL DISMISSAL
Prior to oral argument, plaintiffs moved under Fed.R.App.P. 42(b) to dismiss their appeal as to the identicality, waiver, and third vehicle claims, thereby leaving the lead-time issue as the sole basis for interim relief. Plaintiffs’ action was prompted by the Second Circuit’s adverse ruling on the identicality and waiver claims, which came after plaintiffs’ opening brief was filed in this case. In addition, plaihtiffs maintain that the third vehicle claim requires testimony on the effects of sulfur on emissions systems, and that the evidence in the record is outdated and incomplete.
We have broad discretion to grant voluntary motions to dismiss. “An appeal may be dismissed on motion of the appellant upon such terms as may be ... fixed by the court.” Fed.R.App.P. 42(b);
see also
16 Charles A. Wright & Arthur R. Miller,
Federal Practice and Procedure
§ 3988, at 480 (1977). Such motions are generally granted, but may be denied in the interest of justice or fairness.
See HCA Health Servs. of Virginia v. Metropolitan Life Ins. Co.,
957 F.2d 120, 123 (4th Cir.1992);
United States v. Washington Dep’t of Fisheries,
573 F.2d 1117, 1118 (9th Cir.1978).
DEP contends that this case “presents one of the rare occasions where justice requires that a voluntary motion to dismiss be ... denied,” so that we might rule that the third vehicle claim fails as a matter of law. We are unpersuaded. None of the grounds that have compelled courts to deny voluntary motions to dismiss are present here.
See, e.g., Township of Benton v. County of Berrien,
570 F.2d 114, 118-19 (6th Cir.1978) (denying motion to dismiss filed by one of two appellants because dismissal “would be a meaningless gesture,” where both appellants pressed same arguments, and both would be affected by decision);
Blount v. State Bank & Trust Co.,
425 F.2d 266, 266 (4th Cir.1970) (denying appellant’s motion to dismiss, but granting appellee’s because appellant violated briefing schedule and caused appellee to file motion to dismiss);
Local 53, Int’l Ass’n of Heat and Frost Insulators v. Vogler,
407 F.2d 1047, 1055 (5th Cir.1969) (denying motion and affirming on the merits because motion to dismiss was based on unsound argument that appeal from injunction was moot since appellant was voluntarily refraining from enjoined conduct);
see also Washington Dep’t of Fisheries,
573 F.2d at 1118 (courts “might have grounds” for denying motion to dismiss if sought to evade appellate review and to frustrate court orders).
Furthermore, we note that granting the Rule 42(b) motion will not shelter the remaining claims from scrutiny. We will simply be accepting plaintiffs’ decision to let those claims be finally adjudicated before bringing them to this court.
Creaton v. Heckler,
781 F.2d 1430, 1431 (9th Cir.1986). The interests of fairness and judicial economy are well served by restricting our review to the leadtime issue, the sole claim both parties concede we must decide. Consequently, we grant the motion for partial dismissal and decline to reach the merits of the third vehicle claim.
III.
LEADTIME
We turn to whether the district court was correct in denying a preliminary injunction based on the leadtime claim. We will reverse only if the district court abused its discretion or made a manifest error of law.
Narragansett Indian Tribe v. Guilbert,
934 F.2d 4, 5 (1st Cir.1991).
At issue is the proper construction of the leadtime requirement. The statute at issue, § 177 of the Clean Air Act, 42 U.S.C. § 7507, empowers states to adopt and enforce California emissions standards for vehicles and motor vehicle engines “for any model year,” if the state adopts such standards “at least two years before commencement of such model year (as determined by regulations of the [EPA] Administrator).”
The parties agree that the model year designation of any particular vehicle depends on when that model or engine was produced. According to EPA, a model year is either the calendar year, or the manufacturer’s production period, lasting no longer than a day less than two years,
ie.,
from January 2 of the preceding year through December 31 of the calendar year for which the model year is named. 40 C.F.R. § 86.082-2 (“model year” means calendar year or “the manufacturer’s annual production period (as determined by the [EPA] Administrator)”); EPA, Office of Mobile Sources, Advisory Circular 6B (1987) (hereinafter Advisory Circular 6B) (defining annual production period).
The parties dispute whether or not the leadtime requirement applies on an industry-wide basis. According to plaintiffs, all 1995 models sold in Massachusetts must be federal cars because GM and Chrysler began producing 1995 models prior to January 31, 1994. In other words, the model year began less than two years after the LEV standards were adopted.
Basing its interpretation on Advisory Circular 6B, with support from EPA’s amicus brief, DEP demurs, maintaining that the leadtime requirement is satisfied as to any model in an “engine family” first produced after January 31,1994. The record indicates that an “engine family” is a classification used to group together vehicles that have the same emissions control design. DEP’s standards would apply only to models or engine families first produced after January 31, 1994. Plaintiffs characterize DEP’s interpretation as “splitting” the model year because the 1995 standards would apply to some, but not all 1995 cars.
The district court’s position approximated DEP’s (and EPA’s): “ ‘The failure to provide the statutory leadtime to a particular manufacturer for a particular model year does not
invalidate
the standards themselves. Instead, it merely renders them unenforceable as against those manufacturers which were not given the requisite two-years notice.’”
American Automobile Mfrs. Ass’n v. Greenbaum,
No. 93-10799-MA, slip op. at 23, 1993 WL 443946 (D.Mass. Oct. 27, 1993) (quoting
New York DEC,
831 F.Supp. at 64 (emphasis in original)). The court did not rule on whether each engine family has a different model year commencement date, but noted that Advisory Circular 6B “seems to support DEP’s understanding.”
Id.
at 23 n. 20.
Plaintiffs’ industry-wide date for the commencement of the model year prevailed in the Second Circuit. That court held that EPA’s position was not entitled to deference because it was “newly minted” for litigation and was not embodied in a regulation under § 177.
MVMA,
17 F.3d at 535. Moreover, the court found an industry-wide date to be consistent with Congressional intent, while EPA’s interpretation was unprecedented and “unreasonable” because it would be confusing to the industry and impractical to enforce.
Id.
at 535-36. Plaintiffs urge us to follow the Second Circuit. We decline to do so.
In the first place, we are not confronted with a regulatory program identical to that at issue in the Second Circuit. New York, like California, but unlike Massachusetts, imposed fleet average requirements to determine the mix of vehicles sold in the state each year. The Second Circuit determined that the leadtime provision was “best read” with an industry-wide commencement date because splitting the year would “unduly complicate the fleet averaging plan.”
MVMA,
17 F.3d at 535. Manufacturers would be unable to buy and sell emissions credits to meet the requirements because some of them would have to comply with 1995 standards, but others would not.
Id.
We agree with the Second Circuit that fleet averaging might be more complicated in the first year that California-type standards are effective in a § 177 state, but we discount the significance of that consideration. Fleet averaging for emissions programs is a concept devised by California, not Congress. Although the Second Circuit found fleet averaging to be the “crux” of the LEV plan,
id.,
neither party in this case has argued that under § 177, states must adopt fleet average requirements.
Accordingly, the extent to which a split model year interpretation unduly complicates the administration of fleet averaging is not a pertinent consideration.
Furthermore, we do not agree with the Second Circuit’s characterization of EPA’s definition as having been “newly minted” for litigation. EPA did not develop its interpretation during litigation. Rather, the agency issued Advisory Circular 6B in 1987, while New York and Massachusetts adopted Cali
fornia’s requirements in 1992. And in a letter dated March 8, 1991, to Congressman John Dingell (D.Mich.), the EPA Administrator cited Advisory Circular 6B for the premise that “a state adopting California emissions standards may apply these standards to any engine family whose production period begins on a date which is beyond two years past the date that the standards were adopted_” It is significant that EPA’s interpretation did not spring from a litigator’s self-spun argument, but arose prior to litigation, and was expressed by the Administrator in a letter to a member of Congress from Michigan.
See Federal Labor Relations Auth. v. United States Dep’t of Navy,
941 F.2d 49, 59 (1st Cir.1991) (deferring to agency interpretation first announced in ami-cus brief and later adopted as “official” agency position by agency director in unpublished letter);
cf. Martin v. Occupational Safety and Health Review Comm’n,
499 U.S. 144, 156, 111 S.Ct. 1171, 1179, 113 L.Ed.2d 117 (1991) (“Our decisions indicate that agency ‘litigating positions’ are not entitled to deference when they are merely appellate counsel’s
‘post hoc
rationalizations’ for agency action, advanced for the first time in the reviewing court.”).
Based on the statutory requirement that “model year” be determined by EPA regulations, the Second Circuit held that Congress intended that EPA would promulgate a regulation defining “model year” under § 177.
MVMA,
17 F.3d at 535. We disagree. We find that a regulatory definition predating § 177 satisfies the statute. Congress’s use of the passive voice indicates that an existing regulatory definition would suffice.
Compare
§ 177, 42 U.S.C. § 7507 (“as determined by regulations of the Administrator”)
with, e.g., id.
§ 7521(a)(1) (“the Administrator shall” by regulation prescribe federal auto emission requirements). In 1970, Congress passed 42 U.S.C. § 7521(b)(3)(A), which defines “model year,” for the purposes of the federal emissions control program, as the “calendar year,” or “the manufacturer’s annual production period (as determined by the [EPA] Administrator) which includes January 1 of such calendar year_” The regulatory definition of model year in effect when § 177 was enacted tracked that definition: “ ‘Model year’ means [the calendar year, or] the manufacturer’s annual production period (as determined by the Administrator) which includes January 1 of such calendar year_” 40 C.F.R. § 86.-082-2.
We also reject the Second Circuit’s finding that Congress could not have contemplated that the leadtime provision might apply on an engine-family basis.
MVMA,
17 F.3d at 535. We note first that what Congress “contemplated” is of limited relevance, given that EPA was expressly authorized to define when the model year commences. Moreover, since 1972, EPA has issued advisory circulars describing how to determine the model year “for any specific model within an engine family.”
E.g.,
Advisory Circular 6A, at 2 (Sept. 1, 1972). And while EPA has never implemented a split model year in the federal emissions control program, we do not place great weight on this. There are relatively few leadtime provisions in the Clean Air Act emissions control program. Because states with § 177 programs are, by definition, encountering significant air pollution problems, and because Congress expressly delegated to EPA the power to define model year under § 177, EPA may identify policy considerations allowing it to construe the leadtime provisions in the federal program differently from § 177.
See Comite pro Rescate de la Salud v. Puerto Rico Aqueduct and Sewer Auth.,
888 F.2d 180, 187 (1st Cir.1989) (“[W]here the
reason
for the court’s ‘deference’ reflects its belief that Congress, in effect,
delegated
to the agency a degree of interpretive power, it does not seem odd to find the agency interpreting the same words somewhat differently as they apply to different parts of the statute in order better to permit that statute to fulfill its basic congres-sionally determined purposes.” (emphasis in original)),
cert. denied,
494 U.S. 1029, 110 S.Ct. 1476, 108 L.Ed.2d 613 (1990).
On the other hand, one might argue that a court owes EPA’s interpretation no deference because the statute requires EPA to define “model year” by “regulation,” while EPA’s definition is found not in a regulation, but in a policy statement (Advisory Circular
6B).
See MVMA,
17 F.3d at 535 (“Section 177 charges the EPA with the single, narrow responsibility to issue ‘regulations’ in order to define the commencement of a model year under § 177. The EPA Advisory Circular ... is not a ‘regulation’ for § 177 purposes and was not promulgated specifically to implement this provision_”). Plaintiffs failed to make such an argument to the district court and compounded that error by omitting the point from their opening brief.
See McCoy v. MIT,
950 F.2d 13, 22 (1st Cir.1991) (“It is hornbook law that theories not raised squarely in the district court cannot be surfaced for the first time on appeal.”), cer
t. denied,
— U.S. -, 112 S.Ct. 1939, 118 L.Ed.2d 545 (1992);
see also Frazier v. Bailey,
957 F.2d 920, 932 n. 14 (1st Cir.1992) (arguments raised only in reply brief are insufficient to preserve claim on appeal). Until filing their reply brief in this court, plaintiffs failed to assert that no EPA definition of model year existed for the purposes of § 177, and in fact cited Advisory Circular 6B and 40 C.F.R. § 86.082-2 to the district court for the premise that the model year began on January 2, 1994.
See, e.g.,
Plaintiffs Mem. of Law in Support of Mot. for S.J., at 44; First Amended Complaint ¶ 55 (“As defined by EPA’s regulations, the 1995 model year commences as early as January 2, 1994.
See
40 C.F.R. § 86-082-2 (1992); EPA Office of Mobile Sources Circular 6B (1987).”).
We have recognized an exception to the raise-or-waive rule where the argument surfacing for the first time on appeal is “ ‘so compelling as virtually to insure appellant’s success,’ ” and a “ ‘gross miscarriage of justice’” would result from our failure to address it.
Johnston v. Holiday Inns, Inc.,
595 F.2d 890, 894 (1st Cir.1979) (citations omitted);
accord United States v. Slade,
980 F.2d 27, 31 (1st Cir.1992). The argument here is not so compelling as to assure plaintiffs’ success. EPA’s interpretation of § 177 would be entitled to some weight, where EPA administers the federal emissions program and is charged with evaluating whether state plans for meeting the NAAQS are consistent with the Act.
See
42 U.S.C. § 7410(k)(3).
Furthermore, plaintiffs do not contend that our failure to consider the argument would cause a gross miscarriage of justice. Nor could they so contend. In the first place, this is an interlocutory appeal; plaintiffs may raise the argument in the district court before issues pertaining to the 1995 requirements become moot, because the model year for any vehicle lasts until December 31,1995. In addition, this is not a case in which an appellant might lose her home,
see United States v. One Urban Lot,
885 F.2d 994, 1001-02 (1st Cir.1989), or a prisoner might remain incarcerated,
see United States v. La Guardia,
902 F.2d 1010, 1013 (1st Cir.1990), if we deem the issue waived. And though the question before us, concerning the earliest date vehicles outside California might be subject to California-type emissions standards, is certainly one of interest to the public, the degree of public interest pales in contrast with that involved when the federal government’s right to prosecute suspected criminals is at issue,
e.g., United States v. Krynicki,
689 F.2d 289, 292 (1st Cir.1982). These cases show the gulf that exists between the prospective harm here and the type of harm that permits serious consideration of relaxing the raise-or-waive rule, within the reviewing court’s discretion. Accordingly, we find the argument waived for the purposes of this appeal.
Assuming, therefore, that the regulatory definition of model year required by § 177 is embodied in Advisory Circular 6B, we next inquire whether EPA’s interpretation is arbitrary, capricious, or manifestly contrary to the statute.
Chevron U.S.A., Inc. v. Natural Resources Defense Council,
467 U.S. 837, 843-44, 104 S.Ct. 2778, 2782, 81 L.Ed.2d 694 (1984). Such deference is due because Congress explicitly delegated to EPA, the agency responsible for administering the federal emissions program, the task of defining model year under § 177.
Plaintiffs argue that applying the leadtime requirement to individual models or engine families contradicts Congress’s intent made manifest by the statute’s use of the terms “commencement” and “model year” in the singular. We disagree. At best, the statutory language is ambiguous with respect to whether the leadtime requirement might apply on an industry-wide or engine-family-specific basis.
See
42 U.S.C. § 7507 (State “may adopt and enforce [standards] for any model year ... if— ... California and such State adopt such standards at least two years before commencement of such model year_”). An examination of other lead-time provisions enacted in 1977 for the Act’s federal emissions program does not clarify the issue because those provisions generally pertain to heavy duty engines, whose model year commencement date, according to the record, is always January 1 of the calendar year. Moreover, those provisions could be read with either an industry-wide model year commencement date, or separate dates for different engine families.
E.g.,
42 U.S.C. § 7521(a)(3)(E)(ii) (1988) (“No such changed standard shall apply for any model year before the model year four years after the model year during which regulations containing such changed standard are promulgated.”) (repealed in 1990).
Moreover, the legislative history of § 177 is generally unenlightening.
Congress clearly enacted the leadtime provision for the manufacturers’ benefit. H.R.Rep. No. 294, 95th Cong., 1st Sess. 310, U.S.Code Cong. & Admin.News, 1977, p. 1077 (1977) (“Manufacturers are not only assured of identity of standards and test procedures; they are also assured adequate lead time.”);
see also MVMA,
17 F.3d at 535. Although plaintiffs would prefer that all 1995 cars be subject to the same regulatory requirements, that is not necessarily the import of that statement of legislative intent. EPA’s interpretation grants every manufacturer two years to develop emissions controls and to devise marketing and distribution strategies for any new vehicle or engine family subject to California-type standards. There is no inherent conflict between EPA’s interpretation and Congress’s intent.
Plaintiffs maintain that EPA’s interpretation does not reflect a reasonable policy determination because it would cause “enormous competitive and practical problems,” in that California-type requirements would apply to some 1995 vehicles, while the remainder would be subject to federal standards. Appellants’ Br. at 46. According to plaintiffs, this split would cause dealer and consumer confusion, the disruption of vehicle distribution systems, and competitive disadvantages for some dealers and manufacturers. DEP argues, however, that these concerns are overstated, given the widespread use of computerized inventory controls. Also, on the other side of the balance is the state’s interest in applying California requirements to some models as soon as possible. Any vehicle subject to regulatory controls will be subject to those controls for the vehicle’s useful life. Conversely, vehicles escaping the controls may travel over Massachusetts highways for years emitting pollutants in excess of California standards. Whether EPA’s interpretation imposes greater costs than benefits is a policy determination. ‘When Congress, through express delegation or the introduction of an interpretive gap in the statutory structure, has delegated policy making authority to an administrative agency, the extent of judicial review of the agency’s policy determinations is limit
ed.”
Pauly v. Bethenergy Mines, Inc.,
501 U.S. 680, 696-98, 111 S.Ct. 2524, 2534, 115 L.Ed.2d 604 (1991). We will reject the agency’s interpretation only if it is arbitrary or illegal. It is neither. Accordingly, based on the assumption that Advisory Circular 6B provides a regulatory definition of “model year” for the purposes of § 177, we conclude that the leadtime requirement was satisfied.
The likelihood of success on the merits is a predicate to the issuance of a preliminary injunction. Plaintiffs failed to establish such a likelihood. Moreover, plaintiffs “have not persuaded us that the lower court overlooked pertinent factors, focused on inappropriate factors, or made a serious error in weighing and balancing the relevant concerns.”
Weaver,
984 F.2d at 14. Therefore, we hold that the district court did not abuse its discretion in refusing to enjoin the 1995 standards.
IV.
COSTS AND FEES
DEP argues that it is entitled to costs and attorney’s fees. Prevailing parties are normally entitled to costs. Fed.R.App.P. 39; 9 James W. Moore et al.,
Federal Practice
¶ 239.02[1], at 39-6 to -7 (2d ed. 1994).
And costs are routinely available whenever this court dismisses an appeal, even if the appellant moved for dismissal.
See Waldrop v. Department of Air Force,
688 F.2d 36, 37 (7th Cir.1982).
On the other hand, DEP’s argument for attorney’s fees must be rejected. DEP seeks reimbursement for legal fees incurred in responding to the appeal on the claims that were dismissed pursuant to Rule 42(b). Neither Rule 42(b) nor Rule 39 provides authority for routine awards of attorney’s fees as a condition of voluntary dismissal.
Waldrop,
688 F.2d at 37-39. While fees may be awarded if an appellant has filed a frivolous appeal or has acted in bad faith,
see Cruz v. Savage,
896 F.2d 626, 631-32, 635 (1st Cir.1990), we find no evidence of such conduct here. We are unpersuaded by DEP’s attempt to characterize the weeks that transpired between the issuance of the Second Circuit opinion and the motion for partial dismissal as evidence
of
plaintiffs’ vexatiousness. It takes time to evaluate a new opinion, and to confer with the client on an appropriate strategy.
V.
CONCLUSION
For the foregoing reasons, we grant the motion for partial dismissal and affirm the district court’s decision not to enjoin the 1995 requirements based on plaintiffs’ leadtime claim. Costs to DEP.
It is so ordered.