DANTRAN, INC., et al., Plaintiffs, Appellants, v. U.S. DEPARTMENT OF LABOR, Defendant, Appellee.
The Secretary of Labor (the Secretary) accused Dantran, Inc., and its principal, Robert C. Holmes (collectively, the plaintiffs), of having violated certain provisions of the McNamara-O'Hara Service Contract Act of 1965, 41 U.S.C. §§ 351-358 (the Act), and the regulations thereunder. She wished to debar the plaintiffs, that is, to place them on a list of contractors with whom no government entity may transact business, for a period of three years. The district court, echoing a determination of the Labor Department's Administrative Review Board (the ARB), authorized debarment. We reverse.
For over a decade, the United States Postal Service routinely awarded contracts to Dantran for hauling mail between various sites in Maine, Vermont, New Hampshire, and Massachusetts. During this period, Dantran operated profitably; in its heyday, the company employed approximately 40 persons and generated annual revenues in the $2,000,000 range. The tectonic plates shifted in mid-1991, when the Secretary, acting on a compliance officer's conclusion that the company's practices violated certain regulations dealing with, among other things, employee fringe benefit payments, “froze” funds owed to Dantran by the Postal Service.
If the Secretary's earlier investigation of Dantran provided any baseline for comparison, the results of the 1991 probe must have come as something of a shock. In 1989, the Secretary had sent a compliance officer to inquire into Dantran's payroll practices. On that occasion, the investigator, George Rioux, uncovered no irregularities. Rioux's final report specifically noted that there were no problems with Dantran's fringe benefit payment practices.
When the Wage and Hour Division returned in 1991, Dantran's practices had not changed at all from 1989. The Secretary's outlook apparently had: shortly after the new compliance officer, Scott Wilkinson, began his review, he informed the plaintiffs that two of Dantran's routine practices-paying employees on a monthly basis and capping fringe benefits at 40 hours per week regardless of the number of contracts, or hours per contract, an employee actually worked-violated the Act.
In response to Wilkinson's admonitions, the plaintiffs promptly devised a plan to inaugurate semi-monthly wage payments and commenced negotiations to identify the amounts due in respect to the cross-crediting of fringe benefits. Wilkinson nonetheless recommended that the Secretary freeze some $20,000 owed to Dantran by the Postal Service to satisfy his estimate of what Dantran owed its employees by virtue of cross-crediting. The Secretary obliged. The timing could not have been worse. Cf. Benjamin Franklin, Poor Richard's Almanac (1758) (explaining how for want of a nail, the kingdom was lost). Without the withheld funds, Dantran could not cover insurance premiums on its fleet of trucks and, unable to keep the uninsured trucks in service, suspended operations. This, in turn precipitated a further withholding of funds, to the tune of some $60,000, without which Dantran could not meet its July 1991 payroll.
Dantran and the Secretary eventually settled all wage-related matters. The settlement totaled roughly $67,000, ($40,000 of which consisted of the wages Dantran had been unable to pay in July 1991). Despite the fact that the settlement made Dantran's work force whole, Wilkinson's final report pressed for debarment “because of the size of the violations and the fact that the firm was investigated once before.” The Secretary acquiesced and, in her complaint, attempted to justify so extreme a sanction on the basis of cross-crediting and an alleged failure adequately to maintain records. The case was tried before an Administrative Law Judge (ALJ). During the hearing, the Secretary raised two new issues: the frequency of payment of wages and Dantran's failure to pay its employees in July 1991 (when the Secretary froze its revenue stream).
In the end, the ALJ concluded that Holmes's testimony was credible 1 and that neither he nor Dantran should be debarred. In his view, the case boiled down to cross-crediting and frequency of wage payments. He determined that Dantran's practices in these respects did not transgress the regulations. As an alternative holding, the ALJ further concluded that Dantran had continued the challenged practices in reasonable reliance on the results of the Secretary's earlier investigation and that the case therefore displayed unusual circumstances sufficient to warrant relief from debarment. See 41 U.S.C. § 354(a). These circumstances included, in addition to the fact that the plaintiffs had been misled by the Secretary's earlier investigation and by statements of various Postal Service employees, three additional facts: (1) Dantran had not acted culpably, willfully, or deliberately to violate the law; (2) it had an excellent history of compliance with the wage and hour laws; and (3) the plaintiffs had fully cooperated with the Secretary's inquiry.
The Secretary appealed the ALJ's ruling to the ARB, which reversed. It found that the cross-crediting and frequency of payment practices violated the regulations and that Dantran had been culpable in disregarding the law because, during the 1989 indagation, Rioux had given Holmes a copy of the regulations (which should have alerted Dantran to the illegality of its actions). The ARB deemed this finding of culpable disregard dispositive on the question of debarment.
The plaintiffs sought judicial review of the ensuing debarment order. In a brief, unpublished memorandum, the district court rejected their plea. This appeal followed. Because the ARB predicated its debarment order on two of Dantran's practices-cross-crediting and making monthly wage payments-we discuss each practice.
With certain exceptions not relevant here, the Act provides that every service contract entered into by the United States “shall contain” provisions specifying the fringe benefits which those employees of the contractor who perform the work will receive. See 41 U.S.C. § 351(a)(2). The Act lists the types of fringe benefits that must be included in such a package, including items such as health insurance, life insurance, and retirement benefits. See id. Instead of prescribing a single method for the delivery of these benefits, the Act permits employers to choose the manner in which they wish to satisfy their statutory obligation. Exercising this latitude, many employers opt to discharge parts of it by making payments in cash to the affected employee (or to a union trust fund on his behalf) in a sum equivalent to the value of certain benefits owed.
The Act leaves room for the operation of the collective bargaining process in valuing fringe benefits. See id. Often, however, service contractors abide by general per-hour valuations determined by the Secretary, based on the Secretary's assessment of prevailing fringe benefit arrangements applicable to similarly situated employees in a particular locality. See id.; see also 29 C.F.R. § 4.51. Thus, for example, if truck drivers in rural Vermont typically earn fringe benefits worth $2.80 per hour, the Secretary might well ask a service contractor to pay that amount to its truck drivers who operate principally in rural Vermont. The particular wage/benefit terms applicable to a specific service contract between an employer and a government agency typically are written into that contract, and so it was with the contracts entered into between Dantran and the Postal Service.
The dispute in this case concerns not the values assigned to fringe benefit payments (i.e., the per-hour rates), but the total number of hours per week for which a service contractor must make payments to its employees. Relying on the regulations-the statute is silent on the matter-the Secretary argues, and the ARB agreed, that when employees work on multiple contracts for a single employer, they are entitled to fringe benefits corresponding to every hour they work on each particular contract in a given week, up to a maximum of 40 hours worth of fringe benefits per contract. In other words, under the Secretary's reading of the regulations, fringe benefit determinations turn not on the total number of hours worked per week, but on the number of different contracts to which an employee is assigned.
To illustrate, assume that a service contractor has three separate mail-hauling contracts with the Postal Service, and that in a given week worker A spends 25 hours on contract X, 20 hours on contract Y, and 10 hours on contract Z. According to the Secretary, worker A must receive an incremental payment equal to 55 hours worth of fringe benefits, notwithstanding that worker B, who likewise toiled for 55 hours that week but spent it all in carrying out contract X, will only receive a payment equal to 40 hours worth of fringe benefits. In contrast, Dantran's interpretation is not contract-specific. On its understanding, both A and B would receive incremental payments in lieu of fringe benefits equal to the rate times 40 hours. It follows, then, that if the Secretary's reading of the regulation is correct, Dantran's use of cross-crediting constituted a violation. Giving due weight to the language and structure of the regulations, we find the Secretary's gloss insupportable.
The regulations' general provision on fringe benefits states, in pertinent part, that “every employee performing on a covered contract must be furnished the fringe benefits required” by the particular contract “for all hours spent working on that contract up to a maximum of 40 hours per week and 2,080 (i.e., 52 weeks of 40 hours each) per year, as these are the typical number of nonovertime hours of work in a week, and in a year, respectively.” 29 C.F.R. § 4.172. Although at first glance one might read the 40-hour cap on benefits to refer to “that contract,” which would imply that the 40-hour maximum applies independently to each contract on which an employee toils, the whole of the proviso belies this reading. The sentence we have quoted explains that the reason why the regulations have capped fringe benefits at “a maximum of 40 hours per week” is that this figure represents “the typical number of nonovertime hours of work in a week.” This is a clear indication that the Secretary intended to make the number of nonovertime hours worked in a week her criterion for determining the amount of fringe benefit payments due. Because the Secretary's reading of this provision mandates fringe benefit payments for overtime hours in addition to nonovertime hours, it contravenes the explicit text of her own regulation and is therefore untenable. See Thomas Jefferson Univ. v. Shalala, 512 U.S. 504, 512, 114 S.Ct. 2381, 129 L.Ed.2d 405 (1994) (noting that an agency's interpretation of a regulation is not given controlling weight when it is inconsistent with, or contradicts, the plain language of the regulation).
The Secretary attempts to parry this thrust by citing the next sentence in section 4.172, which reads: “Since the Act's fringe benefit requirements are applicable on a contract-by-contract basis, employees performing on more than one contract subject to the Act must be furnished the full amount of fringe benefits to which they are entitled under each contract and applicable wage determination.” The Secretary's position involves some circularity: she strives to export the “contract-by-contract basis” language from the later sentence into the earlier in order to establish that the 40-hour maximum refers to each and every contract, and then attempts to import that understanding back into the later sentence to argue that the phrase “to which they are entitled” must refer to 40 hours per week per contract.
We decline this invitation to distort the plain meaning of the earlier sentence by linguistic legerdemain. To adopt the Secretary's construction would eviscerate the regulation's expression of its rationale for establishing a 40-hour maximum, namely, that such a figure represents the typical nonovertime work week. Moreover, such a course also would entail abandonment of the accepted rule that all words and phrases in a statute or regulation should be given effect. See Walters v. Metropolitan Educ. Enterps., Inc., 519 U.S. 202, 117 S.Ct. 660, 664, 136 L.Ed.2d 644 (1997); McIntosh v. Antonino, 71 F.3d 29, 35 (1st Cir.1995); United States v. Ven-Fuel, Inc., 758 F.2d 741, 751-52 (1st Cir.1985).
We hasten to add that recognizing an absolute 40-hour weekly maximum is perfectly consistent with the language of the second sentence. The “contract-by-contract” language in that sentence most probably is meant to clarify that employers must give credit to employees for work done on each contract, perhaps to avoid situations in which an employer might seek to shortchange employees by assigning them to work less than 40 hours on each of several contracts, and then limiting the corresponding fringe benefit payment to work done on a single contract. On this reading, the next passage (mentioning the “full amount of fringe benefits to which [employees] are entitled under each contract”) simply refers to a number less than or equal to the 40-hour maximum.
Other, related sections of the regulations also call the Secretary's proposed interpretation of section 4.172 into question. Consider the regulation that addresses vacation pay. This section purports to synthesize the intent of section 4.172. It states: “As set forth in § 4.172, unless specified otherwise in an applicable fringe benefit determination, service employees must be furnished the required amount of fringe benefits for all hours paid for up to a maximum of 40 hours per week and 2,080 hours per year.” 29 C.F.R. § 4.173(e)(2) (emphasis supplied). In our estimation, this language confirms that 40 hours per week operates as an aggregate cap on an employee's entitlement to fringe benefit payments, regardless of the number of contracts involved.2
Perhaps the most telling sign of the agency's intent appears in an example that the regulations supply to illuminate their effect. The example follows almost immediately after section 4.175(a)(1), see supra note 2, and states that if an employee works 32 hours in one week and receives 8 hours of holiday pay, then he is “entitled to the maximum of 40 hours of health and welfare and/or pension payments in that workweek.” 29 C.F.R. § 4.175(a)(1)(ii). The insertion of the definite article (“the”) preceding the phrase “maximum of 40 hours” itself strongly suggests that the regulations intend to set an overall ceiling of 40 hours per week on fringe benefit payments. Even more telling is the example's next sentence, which explains: “If the employee works more than 32 hours and also received 8 hours of holiday pay, the employee is still only entitled to the maximum of 40 hours of health and welfare and/or pension payments.” Id. This statement is completely at odds with the Secretary's rendition of the regulations. If fringe benefit payments are only subject to per-contract ceilings, as the Secretary insists, then the hypothetical worker described in the example would be entitled not only to 8 hours worth of benefits attributable to his holiday pay, but also to at least 40 hours of benefits for his work-and more if he labored on multiple contracts.
We are not unmindful of our wonted obligation to defer to the Secretary's construction of her legislative rules. See Martin v. OSHRC, 499 U.S. 144, 150-51, 111 S.Ct. 1171, 113 L.Ed.2d 117 (1991); P. Gioioso & Sons v. OSHRC, 115 F.3d 100, 107 (1st Cir.1997). Here, however, when the regulations are read as a whole, the conclusion, already dictated by plain language, becomes inexorable: employees who work for a single service contractor are only entitled to fringe benefit payments up to a maximum of 40 hours per week, regardless of the number of contracts on which they labor. This obvious inconsistency between what the regulations say and what the Secretary says they say eliminates any need for judicial deference. See, e.g., Thomas Jefferson Univ., 512 U.S. at 512, 114 S.Ct. 2381.
In all events, courts should be reluctant to rubber-stamp an agency's interpretation of its regulations when that interpretation has no plausible link to the goals of the regulatory scheme and would lead to absurd results. See P. Gioioso & Sons, 115 F.3d at 107. In this instance, the Secretary's interpretation creates anomalous situations in which employees doing the same kind of work for the same hours will receive different levels of fringe benefit payments merely because the employer happens to assign them to work within the parameters of one, two, or more different contracts.3 At oral argument, we asked the Secretary's counsel directly if the Secretary could provide us with any policy rationale that would justify so quixotic an outcome. Counsel was unable to identify any way in which the Secretary's position plausibly advances (or even jibes with) the purposes of the wage and hour regulations.
Federal courts long have recognized that this statutory and regulatory scheme comprises “remedial labor legislation.” Midwest Maint. & Constr. Co. v. Vela, 621 F.2d 1046, 1050 (10th Cir.1980). If labor concerns are the focal point of this scheme-and we firmly believe that they are-we can perceive no direct relation, at any level of generality, between the Secretary's interpretation of the rules and the purposes that Congress intended to serve. If anything, the Secretary's proffered construction runs at cross-purposes with the Act, for it creates disparities in the treatment of employees who do the same work for the same employer for the same amounts of time. See supra note 3. Consequently, we are unable to sustain the Secretary's position.
III. MONTHLY WAGE PAYMENTS
We detect no flaw in the Secretary's construction of the frequency of payment requirement. Although the Act itself does not prescribe the length of requisite pay periods, the debarment sanction is not limited to statutory violations simpliciter. The Act is written in general terms, and Congress has conferred upon the Secretary broad power to implement its provisions by, among other things, fashioning legislative rules. See 41 U.S.C. § 353(a). Once duly promulgated, such rules become part of the warp and woof of the Act's enforcement scheme, see id., and, if reasonably faithful to the statutory language and intent, constitute binding law.
So it is here: the Secretary promulgated a regulation, 29 C.F.R. § 4.165(b), which states bluntly that “[a] pay period longer than semimonthly is not recognized as appropriate for service employees and wage payments at greater intervals will not be considered as constituting proper payments in compliance with the Act.” In wording the regulation, the Secretary made it crystal clear that monthly pay periods will not do. The regulation is valid and the plaintiffs are bound by its terms.4
The plaintiffs have a fallback position. Brandishing the clean bill of health that they received in the report of the 1989 wage-and-hour investigation, they contend that they reasonably relied on that report's conclusions, and, therefore, that the government should be estopped from pursing an action based on practices (like the monthly payment of wages) that drew no criticism at that time. They stress that Rioux (the first compliance officer) knew of their monthly payment ritual and argue that his silence on the matter suggested to them that they were in compliance. Had he questioned the practice then, they quickly would have taken corrective action (as they did when the second compliance officer, Wilkinson, raised the issue in 1991). To reinforce this point, the plaintiffs also note that the Postal Service paid them on a monthly basis, was on notice that Dantran paid its employees on the same schedule (indeed, Dantran regularly submitted copies of its payroll records to the Postal Service), and tacitly approved the regime.
Estoppel against the government is a concept more frequently discussed than applied. While the Supreme Court has never definitively ruled out the possibility of an estoppel against the government, it consistently has emphasized the difficulties that such a concept entails. See, e.g., OPM v. Richmond, 496 U.S. 414, 419-20, 423, 110 S.Ct. 2465, 110 L.Ed.2d 387 (1990) (plurality op.); Heckler v. Community Health Servs., 467 U.S. 51, 60-61, 104 S.Ct. 2218, 81 L.Ed.2d 42 (1984). If estoppel against the government possesses any viability (a matter on which we take no view), the phenomenon occurs only in the most extreme circumstances. This case does not qualify.
The plaintiffs argue that estoppel is available here because they invoke it as a shield rather than a sword. This distinction apparently has its roots in OPM, wherein the Court restricted its holding to cases involving demands made upon the public fisc. 496 U.S. at 426-28, 110 S.Ct. 2465. But such commendable chariness does not signify that the Court has ever approved-or hinted at approving-a sword/shield distinction. Because blanket adherence to the distinction would conflict with fundamental tenets that underlie estoppel jurisprudence, we reject it.
Using estoppel as a shield implies nothing less than frustrating the government's authority to enforce valid laws. We cannot in good conscience accept a broad rule that prevents the sovereign from enforcing valid laws for no better reason than that a government official has performed his enforcement duties negligently. It does not overstate the case to say that such a rule would risk embroiling the judiciary in the Executive Branch's duty faithfully to execute the law and thereby would raise separation of powers concerns. See United States v. Marine Shale Processors, 81 F.3d 1329, 1348 (5th Cir.1996).
The most obvious manner in which a sword/shield dichotomy would raise separation of powers concerns is by placing the sovereign in a position where, in order to recognize estoppel relief, it would itself have to violate the law. Such a result is untenable and, not surprisingly, numerous cases have held that estoppel must fail under such circumstances. See OPM, 496 U.S. at 430, 110 S.Ct. 2465; Federal Crop Ins. Corp. v. Merrill, 332 U.S. 380, 385-86, 68 S.Ct. 1, 92 L.Ed. 10 (1947); Utah Power & Light Co. v. United States, 243 U.S. 389, 408-09, 37 S.Ct. 387, 61 L.Ed. 791 (1917); see also FDIC v. Hulsey, 22 F.3d 1472, 1489 (10th Cir.1994) (positing that estoppel, if available at all, would only succeed if it would “not frustrate the purpose of the statutes expressing the will of Congress”). This type of holding extends to legal requirements fixed by duly promulgated regulations. See Schweiker v. Hansen, 450 U.S. 785, 790, 101 S.Ct. 1468, 67 L.Ed.2d 685 (1981) (per curiam); Federal Crop Ins., 332 U.S. at 384-85, 68 S.Ct. 1.
In the instant case, the Act's enforcement scheme requires the Secretary to initiate debarment proceedings when she determines that a government contractor has violated the Act or the regulations. Were we to adopt the plaintiffs' suggested sword/shield distinction, we would frustrate the operation of this enforcement mechanism without so much as even inquiring into the equities of the particular case. We are not prepared to embark on so poorly conceived an itinerary.
Our concerns would be no different if the Act's enforcement scheme lay entirely within the Secretary's discretion. There, too, estoppel-even “shield-type” estoppel-would insinuate the Judicial Branch into the Executive Branch's exclusive preserve and would carry a serious potential of undermining the Executive Branch's policies and priority setting in enforcing the laws-a result that is no more palatable than frustrating the congressional will. See Marine Shale Processors, 81 F.3d at 1348-49; Hulsey, 22 F.3d at 1489.
Beyond the attempted sword/shield distinction, the plaintiffs' estoppel claim is insubstantial. It is firmly settled that a party seeking to raise estoppel against the sovereign must, at the very least, demonstrate that government agents have been guilty of affirmative misconduct. See Drozd v. INS, 155 F.3d 81, 90 (2d Cir.1998); Linkous v. United States, 142 F.3d 271, 277-78 (5th Cir.1998); Frillz, Inc. v. Lader, 104 F.3d 515, 518 (1st Cir.), cert. denied, 522 U.S. 813, 118 S.Ct. 59, 139 L.Ed.2d 22 (1997); see also OPM, 496 U.S. at 421-22, 110 S.Ct. 2465; see generally Kenneth Culp Davis & Richard J. Pierce, Jr., 2 Administrative Law Treatise § 13.1, at 232 (3d ed.1994). In this case, no estoppel lies because the record reflects no affirmative governmental misconduct.
It is common ground that affirmative misconduct requires something more than simple negligence, see United States v. Hemmen, 51 F.3d 883, 892 (9th Cir.1995); Kennedy v. United States, 965 F.2d 413, 421 (7th Cir.1992); Azizi v. Thornburgh, 908 F.2d 1130, 1136 (2d Cir.1990), and the plaintiffs' proof does not cross that threshold. The plaintiffs base their estoppel initiative on Rioux's conduct and, to a lesser extent, that of various Postal Service plenipotentiaries. But, though Rioux's final report disclaims any violations of the Act, no one argues that he made that statement with an intent to mislead the plaintiffs about their responsibilities. So, too, the comments attributed to various officials of the Postal Service. In a nutshell, there is not the slightest whiff of affirmative misconduct.
In a related vein, if a statute or regulation clearly limns a party's legal obligations, the party cannot justifiably rely for estoppel purposes on a government agent's representation that the law provides to the contrary. See Federal Crop Ins., 332 U.S. at 383-84, 68 S.Ct. 1; Marine Shale Processors, 81 F.3d at 1349-50 (collecting cases). This is so because the government speaks most authoritatively through its official policymaking machinery, not through individuals, even if the individuals occupy responsible agency positions. See Federal Crop Ins., 332 U.S. at 384, 68 S.Ct. 1; Irving v. United States, 162 F.3d 154, 166 (1st Cir.1998) (en banc). Here, even if the plaintiffs relied on Rioux's report and/or the statements of postal officials to conclude that their regimen did not violate the Secretary's frequency of payment requirement, the clear language of the applicable regulation would render such reliance unreasonable.
IV. RELIEF FROM DEBARMENT
We grapple next with an issue of some complexity. Both the statute and the regulations provide that the existence of “unusual circumstances” may forestall the imposition of a debarment sanction. See 41 U.S.C. § 354(a); 29 C.F.R. § 4.188(a). Although the Act does not elaborate, the Secretary has established that the existence of “unusual circumstances” in a given case depends on the absence of aggravating factors and the presence of mitigating factors. See 29 C.F.R. § 4.188(b)(3)(i)-(ii). The plaintiffs assert that their case fits within these confines.
A first, potentially conclusive, step in the pavane involves aggravation. Under the regulations, if aggravating factors inhere, a government contractor cannot be saved from debarment. See id. § 4.188(b)(3)(ii). The regulations describe in some detail what constitutes aggravation. They speak in terms of “deliberate” or “willful” conduct, “culpable neglect,” and “culpable disregard,” id. § 4.188(b)(3)(i), and thus evince a design to ensure blacklisting of those who have defied the law with a degree of impunity or premeditation. What the regulations mean by the term “culpable” is not spelled out, except to stipulate that “falsification of records” (an evil not present in this case) qualifies as “culpable failure to comply with recordkeeping requirements.” Id. If this latter example is intended to serve as a guide, culpability must require more than simple negligence or a mere failure to ascertain whether one's practices coincide with the law's demands.
The ARB concluded that the plaintiffs' conduct satisfied the standard of culpability. This conclusion relied, in part, on the plaintiffs' practice of cross-crediting fringe benefits and, in part, on the plaintiffs' violation of the frequency of payment regulation. The first ground does not hold water given our determination that the Act allows cross-crediting. See supra Part II. Thus, the ARB's finding of aggravation must stand or fall on the second ground.
The ARB found the violation relating to monthly wage payments culpable due to the clarity of the prohibitory regulation. But this assumes that the enforcement scheme mandates debarment whenever an employer violates an unambiguous regulation. This assumption is unfounded.
The regulations speak of “culpable neglect to ascertain” or “culpable disregard” of whether one is in violation of the Act as being criteria for finding factors in aggravation. To hold that this language renders culpable every violation of an unambiguous regulation would, for all practical purposes, turn the debarment provision into a strict liability regime. The very invocation of a culpability standard, however, is a sure sign that strict liability is not what the Secretary intended. This view is explicitly confirmed by the Secretary's incorporation into the regulations of a portion of the Act's legislative history which states that “[t]he authority [to relieve from blacklisting] was intended to be used in situations where the violation was a minor one, or an inadvertent one, or one in which disbarment ․ would have been wholly disproportionate to the offense.” 29 C.F.R. § 4.188(b)(2) (alterations in the original). The ARB's strict liability regime drains section 4.188(b)(2) of meaning: after all, a violation of an unambiguous regulation can be minor, inadvertent, wholly disproportionate to a proposed debarment, or all of the above. Thus, simply stating that an employer violated an unambiguous regulation, without more, cannot justify debarment.
The ARB's reliance on a different portion of the same regulation does not affect our conclusion. Section 4.188(b)(1) states, in pertinent part, that unusual circumstances do not include instances “such as negligent or willful disregard of the contract requirements and of the Act and regulations, including a contractor's plea of ignorance of the Act's requirements where the obligation to comply with the Act is plain from the contract.” Drawing on this language and on the fact that the plaintiffs were in possession of a copy of the regulations, the ARB reasoned that debarment was warranted. We are not convinced.
Fairly read, this language contemplates an automatic finding of culpability only when the law's requirements are obvious on the face of the contract. Under such circumstances, a contractor's disregard of legal requirements legitimately can be considered willful or grossly negligent, and thus manifest the species of culpability delineated by the regulations. Here, however, none of Dantran's contracts with the Postal Service provided notice of the Secretary's interpretation of her frequency of payment rule. The reverse is closer to the truth, for the contracts themselves provided for monthly payments to Dantran and thus appeared to confirm Dantran's historic payroll practices. In this case, then, it is not enough to point to possession of the regulations, and rest. There must be affirmative evidence of culpable conduct.
This brings us to the second reason upon which the ARB premised its determination that Dantran's monthly payment protocol warranted debarment. In the last analysis, this is less a reason than a self-fulfilling prophecy. The ARB describes this reason as “ample evidence” of culpable conduct, and cites two purported facts to prove ampleness: first, the plaintiffs' longstanding practice of monthly wage payments; and second, their attitude toward the first compliance officer when he allegedly raised questions anent this practice. These two “facts” telescope into one, for the significance of the first “fact” depends entirely on the validity of the assumption that underlies the second. If no questions were raised about monthly payments during the first investigation, then the plaintiffs scarcely can be criticized for continuing the established praxis.
The assumption underlying the second “fact” rests solely upon the following testimony from Rioux's direct examination:
Q․ What about the payments on a monthly basis? How were they paying people? If you know.
A. I think they were paying on a monthly basis back then. And I think we discussed that. I'm not sure exactly when, but ․
Q. Okay, but you did discuss it with him. what were the nature-what did you tell him regarding the payment on a monthly basis?
A. Then this is going by memory, but it's a final conference, initial conference, in between. We did talk about it. I said the regulations call for payment on a semi-monthly, bi-weekly basis. And he said, “the post office pays me on a monthly basis When they pay me on a bi-weekly basis, I'll pay them on a bi-weekly basis.”
According to the ARB, this exchange demonstrates that Dantran's subsequent monthly payment of employees knowingly violated the law. We conclude that the ARB, in superimposing its interpretation of this testimony upon the decisional calculus, failed to accord due deference to the ALJ's findings of fact.
An odd standard of review obtains here. In cases arising under the Act, courts do not employ the “substantial evidence” test that most frequently accompanies judicial review of final administrative orders. See, e.g., 5 U.S.C. § 706(2)(E) (Administrative Procedure Act); see also Universal Camera Corp. v. NLRB, 340 U.S. 474, 490-91, 71 S.Ct. 456, 95 L.Ed. 456 (1951). Instead, the Service Contract Act, 41 U.S.C. § 353(a), incorporates the hearing provisions of the Walsh-Healey Act, 41 U.S.C. § 38-39, and thereby adopts a standard of review under which findings of fact made by the Secretary's authorized representative upon notice and hearing are conclusive if supported by a preponderance of the evidence. See Vigilantes, Inc. v. Administrator of Wage & Hour Div., 968 F.2d 1412, 1418 (1st Cir.1992) (explicating standard of judicial review for cases under the Act).
This standard presents certain interpretive difficulties. Asking an appellate tribunal to review a factfinder's conclusions for a “preponderance of the evidence” gives a familiar phrase a new twist. In its normal iteration, the preponderance of the evidence standard, like “clear and convincing” and “beyond a reasonable doubt,” establishes a quantum of proof to be measured by the factfinder, not a standard for error-detection. When used to describe appellate review, however, the phrase is at best an awkward locution, for it connotes nothing about the degree of probability of error required before a reviewing court may set aside a factual determination. If an appellate tribunal were to apply the preponderance of the evidence standard literally, it would mire itself in a factfinding exercise inconsistent with the accepted appellate role. We believe that Congress, in selecting the preponderance of the evidence phraseology, must have contemplated some other course.
The Secretary posits that, under the Act, the preponderance standard should be treated as synonymous with the substantial evidence standard. We think not. Congress appears deliberately to have refrained from employing the conventional substantial evidence rule.5 Courts ordinarily should presume that Congress legislates with knowledge of the legal standards prevailing in administrative law. See, e.g., Commissioner v. Keystone Consol. Indus., Inc., 508 U.S. 152, 159, 113 S.Ct. 2006, 124 L.Ed.2d 71 (1993). Consequently, we are unprepared to say that Congress's election to use a seemingly different rule in this instance should be thwarted. See Haas v. IRS, 48 F.3d 1153, 1156-57 (11th Cir.1995); cf. BFP v. Resolution Trust Corp., 511 U.S. 531, 537, 114 S.Ct. 1757, 128 L.Ed.2d 556 (1994) (“It is generally presumed that Congress acts intentionally and purposely when it includes particular language in one section of a statute but omits it in another, and that presumption is even stronger when the omission entails the replacement of standard legal terminology with a neologism.”) (citation and internal quotation marks omitted).
Rejecting the Secretary's position establishes what the phrase (“preponderance of the evidence”) does not mean, but leaves unanswered the question of what it does mean. We are aided in this inquiry by the fact that the Act's standard of appellate review, though curious, is not unique: Congress used comparable language to describe judicial review of factfinding in arbitration proceedings under the Multiemployer Pension Plan Amendments Act of 1980 (MPPAA), Pub.L. 96-364, 94 Stat. 1208. In relevant part, that statute's judicial review provision, 29 U.S.C. § 1401(c), states that “there shall be a presumption, rebuttable only by a clear preponderance of the evidence, that the findings of fact made by the arbitrator were correct.” In construing this provision, the Seventh Circuit concluded that preponderance of the evidence, when articulated as a standard of review, operates much the same as the “clearly erroneous” standard. See Jos. Schlitz Brewing Co. v. Milwaukee Brewery Workers' Pension Plan, 3 F.3d 994, 998-99 (7th Cir.1993); Chicago Truck Drivers Pension Fund v. Louis Zahn Drug Co., 890 F.2d 1405, 1411 (7th Cir.1989).6 The Eleventh Circuit reached an identical conclusion in respect to the precise provision at issue here. See Amcor, Inc. v. Brock, 780 F.2d 897, 899 (11th Cir.1986). Among its other virtues, such an approach imports a certain symmetry, as courts regularly review factfinding done pursuant to a preponderance of the evidence standard for clear error. See, e.g., Foster-Miller, Inc. v. Babcock & Wilcox Canada, 46 F.3d 138, 147 (1st Cir.1995); Fed.R.Civ.P. 52(a). We therefore hold that the Service Contract Act's preponderance of the evidence standard requires that we review factual findings for clear error.
This leaves a further question as to whether deference is due to the ALJ or the ARB. The Secretary advocates the latter position. In many settings, that viewpoint would have force. As a default rule, the Administrative Procedure Act establishes that an agency enjoys plenary powers of review over a hearing officer's factual and legal conclusions. See 5 U.S.C. § 557(b); see also 2 Davis & Pierce, supra § 11.2, at 178-79. Nevertheless, agencies may elect to deviate from this default rule. See Vercillo v. CFTC, 147 F.3d 548, 553 (7th Cir.1998); Chen v. GAO, 821 F.2d 732, 737 & n. 6 (D.C.Cir.1987); Mullen v. Bowen, 800 F.2d 535, 541 & n. 8 (6th Cir.1986). Here, the statute and the regulations, read together, effectuate such a departure.
The Service Contract Act provides, in relevant part, that the Secretary or her authorized representative “shall make findings of fact after notice and hearing, which findings shall be conclusive upon all agencies of the United States, and if supported by the preponderance of the evidence, shall be conclusive in any court of the United States.” 41 U.S.C. § 39. We think it is significant that the final clause refers back to the findings of fact that are made “after notice and hearing.” Under the current administrative structure, those findings of fact are made by the hearing officer (i.e., the ALJ), not the ARB.
The regulations further confirm our intuition. They refer to the ARB as an “appellate body,” 29 C.F.R. § 8.1(d), and empower it to “modify or set aside” an ALJ's findings of fact “only when it determines that those findings are not supported by a preponderance of the evidence,” id. § 8.9(b). Given that the phrase “preponderance of the evidence” is here couched as a standard of appellate review, we believe that the regulations contemplate deference to the hearing officer qua factfinder and require the ARB (like a reviewing court) to accept an ALJ's factfinding absent clear error.
Further evidence of the Secretary's intent to limit the ARB's function to appellate review is found in her refusal to confer the authority to receive evidence upon the ARB. See id. § 8.1(d). If additional evidence is needed, the ARB must remand to the hearing officer. See id. Courts have held that the authority to receive additional evidence customarily signals the power to reweigh the evidence. See, e.g., Carlisle Area Sch. v. Scott P., 62 F.3d 520, 528 (3d Cir.1995). We believe that the converse also is true: the absence of such authority signals that a reviewing body's power is limited to error-correction.
Such a structure of agency review, in which deference is due to the hearing officer's findings of fact rather than to the factual determinations of the agency's reviewing body, is unusual, but not unprecedented. It mirrors, for example the Secretary's administration of black lung benefits under the Federal Coal Mine Health and Safety Act, Pub.L. No. 91-173, Title IV, 83 Stat. 792 (Dec. 30, 1969) (codified, as amended, at 30 U.S.C. §§ 901-945). The applicable regulations create a framework in which the agency's reviewing body, the Benefits Review Board (BRB), is enjoined to employ a substantial evidence standard of review-and, thus, functions under precisely the same constraints as does a reviewing court. See Consolidation Coal Co. v. McMahon, 77 F.3d 898, 901 (6th Cir.1996); Taylor v. Alabama By-Products Corp., 862 F.2d 1529, 1532-33 (11th Cir.1989). Consequently, as long as the ALJ's findings of fact satisfy the stipulated standard, they are conclusive upon both the BRB and the courts.
Based on these insights, we conclude that the ARB is an appellate tribunal which reviews an ALJ's findings of fact only for clear error.7 Accord Amcor, 780 F.2d at 899; see also American Waste Removal Co. v. Donovan, 748 F.2d 1406, 1408-09 (10th Cir.1984) (“The findings of fact made by an ALJ in proceedings pursuant to the Service Contract Act are conclusive in any court of the United States if supported by a preponderance of the evidence.”). It is against this backdrop that we assess the ARB's debarment decision in this matter.
On the crucial point-whether Rioux had raised the monthly payment question with Dantran during the first investigation-the ARB condemned the ALJ for “erroneously ignor[ing]” Rioux's testimony. But the ALJ's memorandum opinion makes it crystal clear that he did not “ignore” the snippet of testimony that the ARB plucked from the record. The ALJ explicitly considered the evidence and declined to give it weight, instead crediting Holmes's testimony that Rioux made no such point. Having reviewed the entire record, we are persuaded that this finding was not clearly erroneous-especially since the ALJ, not the ARB, had the opportunity to observe the witnesses' demeanor at first hand.
To be sure, Rioux testified, albeit somewhat tentatively, that he had once mentioned to Holmes that the monthly payment practice violated the regulations. His final report, however, did not contain a word to that effect. On the contrary, it stated that, based on employee interviews and perscrutation of payroll records, the plaintiffs' payroll practices were “in compliance with all the provisions of the [Act].” Even were the ALJ obliged to credit Rioux's testimony that he warned the plaintiffs about making monthly wage payments-a dubious supposition, considering a hearing officer's latitude in making credibility calls, see, e.g., Aguilar-Solis v. INS, ---F.3d ----, ---- (1st Cir.1999) [No. 98-1481, slip op. at 11]-such a statement could only have been made informally and, as Rioux himself testified, not in the closing conference. Given that the compliance officer's final report exonerated the plaintiffs up and down the line, the ALJ's finding that Dantran had a reasonable, good-faith belief throughout the ensuing period that its wage-payment practices conformed with the Act's requirements was not clearly erroneous.
In this case, moreover, the ARB's stance is further weakened because, although it rejected the ALJ's finding of fact, it never explained why that finding was mistaken. The ARB pointed to no extrinsic evidence to support its conclusion anent the plaintiff's lack of good faith-and the major piece of relevant extrinsic evidence (Rioux's final report) supports the ALJ's credibility determination. On appellate review, courts are entitled to expect, at a minimum, that an agency which rejects an ALJ's factfinding will provide a rational exposition of how other facts or circumstances justify such a course of action. See International Bhd. of Teamsters v. NLRB, 587 F.2d 1176, 1181 (D.C.Cir.1978); Local No. 441, IBEW v. NLRB, 510 F.2d 1274, 1276 (D.C.Cir.1975). There is no hint of such an analysis in the ARB's opinion.
We have said enough on this score. The short of it is that, gauged by the proper standard of review, the ARB had no legally sufficient reason for upsetting the ALJ's findings of fact (particularly those that relied on credibility assessments).8
At the risk of redundancy, we pause to repastinate ground previously plowed. The ARB ordered debarment because of the presence of aggravating factors. It rested this order on two practices that it deemed culpable: (i) cross-crediting of fringe benefits, and (ii) payment of wages on a monthly, rather than semi-monthly, basis. As we have seen, however, the ARB premised the first of these conclusions on an incorrect view of the law (its belief that cross-crediting violated the regulations), and it premised the second on an erroneous rejection of the ALJ's factfinding. Thus, the ARB's sole remaining rationale for debarment collapses because its foundation is porous.
We agree with our dissenting brother that, when a reviewing court discovers a serious infirmity in agency decisionmaking, the ordinary course is to remand. See, e.g., Pension Benefit Guar. Corp. v. LTV Corp., 496 U.S. 633, 654, 110 S.Ct. 2668, 110 L.Ed.2d 579 (1990); Baystate Altern. Staffing, Inc. v. Herman, 163 F.3d 668, 679 (1st Cir.1998). But such a course is not essential if remand will amount to no more than an empty exercise. See Fergiste v. INS, 138 F.3d 14, 20-21 (1st Cir.1998); Empire Co. v. OSHRC, 136 F.3d 873, 877 (1st Cir.1998); Cotton Petroleum Corp. v. United States Dep't of Interior, 870 F.2d 1515, 1528-29 (10th Cir.1989). Two circumstances bring this principle into play. First, the ALJ found that no aggravating factors were extant, and the record reveals no plausible basis for a contrary finding. Second, mitigating factors abound.
In respect to mitigation, the regulations suggest that factors such as compliance history, cooperation during the investigation, and prompt repayment of sums owed bear heavily on whether relief from debarment should be granted. See 29 C.F.R. § 4.188(b)(3)(ii). Beyond these factors, the agency must consider items such as the results of previous investigations, the extent to which the disputed issues involved uncertain legal questions, the impact of debarment on innocent employees, and the seriousness of the violations. See id. Here, remand would serve no useful purpose from a factfinding perspective, for the ALJ already has made specific findings of historical fact that resolve virtually all of the enumerated factors in the plaintiffs' favor.9 Because these findings easily survive clear-error review, the statute and the regulations make them conclusive upon the ARB. Consequently, there is no point in remanding on this account. See, e.g., Freeman United Coal Min. Co. v. Anderson, 973 F.2d 514, 518 (7th Cir.1992) (declining, in black lung benefits case, to remand when the ALJ had made relevant findings of fact and the court had an obligation to apply the same standard of review as the agency).
The only remaining question is whether we should remand so that the ARB can review the mitigating factors to determine the supportability of the ALJ's balancing (i.e., his application of legal standards to the facts as found). Were evidence of aggravation present, or were mitigation evidence lacking, or were the relevant proof meager, we would not hesitate to remand for the ARB's reconsideration. Here, however, there is no reasonable doubt about the balancing equation's overall equilibrium. Although the Act grants the Secretary latitude in considering whether to recommend relief from debarment, she has cabined that discretion by enumerating the specific factors (and the balancing methodology) upon which she will rely to determine the existence of unusual circumstances. In this case, the findings as to mitigation are so potent that solving the balancing equation in any manner contrary to that which the ALJ reached would constitute an abuse of discretion. Furthermore, while we do not regard the Secretary's list of enumerated factors as exhaustive, neither the ARB's opinion nor the Secretary's brief so much as hints at any other datum in the record that might alter the outcome of the mitigation equation.
In the last analysis, the Secretary's only plausible argument for remand emphasizes that the plaintiffs' transgressions did not involve legal issues of “doubtful certainty.” To the extent that this argument refers to cross-crediting, it is simply wrong. See supra Part II. To the extent that it refers to the frequency of payment issue, the argument has some bite, but for two reasons we do not believe, as a matter of law, that the Secretary's “legal certainty” refrain can prevail.
For one thing, the fact that the Secretary's own representative failed to advise the plaintiffs during the 1989 investigation that their wage-payment practices were impermissible undermines her position. As we have said, Rioux's final report left the plaintiffs with the justifiable impression that the monthly payment of wages met the Secretary's criteria. This impression was bolstered both before and after that investigation by the conduct of various Postal Service officials, who contracted to pay Dantran monthly, accepted the plaintiffs' submission of payroll schedules that clearly showed monthly payments without evincing the slightest concern, and assured the plaintiffs (tacitly, at least) that they were handling employee payments in an appropriate manner. While these facts do not warrant estoppel against the government, see supra Part III, they nevertheless form a significant part of the equitable overlay that the Act and the regulations take into account in determining whether relief from a very harsh penalty is warranted.10
For another thing, the legal certainty associated with the payment frequency provision, without more, cannot overcome a record that reeks of mitigation. This solitary item cannot trump the numerous indicia of mitigation contained in the record (most of which satisfy precisely the specific mitigating criteria limned by the regulations). Thus, given the absence of aggravating circumstances, the prevalence of mitigating factors, and the relatively insubstantial nature of the violation, debarment would be a punishment totally out of proportion to the offense (and, therefore, contrary to the regulations). See 29 C.F.R. § 4.188(b)(2).
We summarize succinctly. The customary rule, as our dissenting brother says, favors remand when a court sets aside an agency determination. But the rule also allows some flexibility. Courts should not indulge in wasteful wheel-spinning. Thus, in the rare case in which the facts admit of only one plausible legal conclusion, an inquiring court serves the ends of justice by a frank acknowledgment that remand promises to be an exercise in futility. See Gatson v. Bowen, 838 F.2d 442, 450 (10th Cir.1988) (declining to remand after concluding that agency had misapplied the law where, regardless of any further findings, the record would support only one legal conclusion); Brock v. L.R. Willson & Sons, Inc., 773 F.2d 1377, 1389 n. 12 (D.C.Cir.1985) (concluding that the court of appeals could rule on a question involving application of law to fact, without remanding to the agency, as long as only one conclusion would be supportable); Donovan ex. rel. Anderson v. Stafford Constr. Co., 732 F.2d 954, 961 (D.C.Cir.1984) (same); see also Donovan v. Capital City Excavating Co., 712 F.2d 1008, 1010 (6th Cir.1983); Usery v. Marquette Cement Mfg. Co., 568 F.2d 902, 910-11 (2d Cir.1977). Because this is such a case-after all, debarment proceedings under the Act are all-or-nothing propositions, see Federal Food Serv., Inc. v. Donovan, 658 F.2d 830, 834 (D.C.Cir.1981); no intermediate penalties are available, so if debarment is not in order, no practical necessity exists for resurrecting the administrative proceeding-remand is not obligatory.
We need go no further. Although the plaintiffs did violate the Act in one particular, they have made a compelling case for relief from debarment. In view of that conclusion, and because remand is not an appealing option, we rescind the debarment order, reverse the decision of the district court, and return the case to that court for the entry of an appropriate judgment.
With some reluctance, I undertake a critique of Judge Selya's monumental analysis of this tangled case. But in several respects I think a different approach may be required.
Before reaching the more troubling question, I generally approve of the direction taken by the majority's dissection of the fringe benefit provisions. However, I do not come away from this examination entirely convinced that the Department's reading of its regulations is wrong and Dantran's is right. As far as it goes, the majority's analysis is persuasive. That analysis does not, however, fully account for the use of terms like “contract-by-contract,” “employees performing on more than one contract,” “on each contract” and the like. The regulations, particularly the latter pertinent sentence of 29 C.F.R. § 4.172,1 emphasize a need to calculate fringe benefits as separately attributable to specific contracts. It is not a complete answer to suggest, as does the majority, that this language is designed merely to mandate accounting for all of the hours worked on the various contracts in order to make up the allowed forty hours. If this were all that was involved, why would the regulations specify that the calculation be made “contract-by-contract?” The emphasis is too insistent, it seems to me, merely to convey the meaning suggested by the majority.
I believe that the language that I have quoted above does tend to support the Department's position. However, reading this language together with the other language which the majority opinion correctly analyzes as supporting Dantran's view,2 the only fair conclusion is that these regulations are self-contradictory and too vague and ambiguous to be enforced. One reason for this is that the Department has not seen fit to explain to us the rationale for its reading of the rule. One would think that, for example, fifty hours worked under several contracts should be rewarded with the same fringe benefits as fifty hours worked under a single contract, as Dantran has interpreted the regulations. It may be, giving maximum ambit to one's imagination, that somehow two contracts provide more funds to cover fringe benefits than does a single contract. We can hardly be expected to speculate that this is the case, however. The best we can do is leave the subject scratching our heads and concluding that the regulations are indecipherable. If that is the case, they are unenforceable. They cannot furnish any basis for an order of debarment.
On the other matter in dispute-the frequency of wage payments-I part company in a significant way with the majority. The record here does not reflect any good reason why an employer, exercising normal prudence, is entitled to assume, without checking, that monthly payment of wages is an available option or that the applicable regulations would endorse this practice. This is not an esoteric matter, like the methodology of furnishing fringe benefits. Instead, it is easily understood and likely to be of immediate concern to employees. It is something about which a prudent employer would take reasonable steps to acquaint himself-even to the point of actually looking at the regulations.
In that connection, did Dantran make any effort between 1986 (when it apparently began monthly payments) and 1991 to ensure compliance with the frequency of payment provisions? It appears not. Holmes testified that he spoke with his Post Office contacts about the propriety of cross-crediting fringe benefits, but he never mentioned asking how often he was required to pay his employees. His wife, who kept Dantran's books and ran its payroll, was similarly remiss. The majority does not even note this major omission but apparently adopts Dantran's thesis that if the contract specifies monthly payments to Dantran by the government, Dantran is entitled to infer that Dantran's employees need be paid only monthly. See ante at 63. This seems specious to me. Although Dantran apparently did not keep large cash reserves, it ought to have better access to working capital than its employees. Dantran is in a better position to adjust for temporary mismatches of revenue and expense than are its employees to wait a full month for their pay. Contractors and merchants the world around receive progress payments, remittances on invoices and other payments from customers on a variety of schedules, but I am not aware that these schedules determine how often they are expected or required to pay their own employees. This inference, for which Dantran argues persistently, is simply a non-sequitur. And I know no reason why government contractors would have some special dispensation to pay their employees less frequently than businessmen in the economy at large.
I am therefore not nearly as indulgent of Dantran's frequency of pay practices as is the majority. On the other hand, I am certainly not convinced that debarment is appropriate. I believe the matter of frequency of wage payments should be remanded to the Board for separate consideration, the fringe benefit matter having dropped out of the case.
Remand would clearly not be an empty exercise. Notwithstanding the majority's apparent effort to consign the Board to irrelevance, the Board does-and should-retain some discretion to determine whether debarment is warranted. The plain language of 41 U.S.C. § 39-“findings of fact ․ shall be conclusive upon all agencies of the United States, and if supported by the preponderance of the evidence, shall be conclusive in any court of the United States”-suggests that reviewing bodies may exercise what amounts to de novo review of the record. If the finder of fact and the reviewing authority are bound by the same standard in establishing the facts (preponderance of the evidence), the logic of the situation is that review is essentially de novo. First Circuit case law apparently supports such a literal reading. See Vigilantes, Inc. v. Administrator of Wage & Hour Div., 968 F.2d 1412, 1418 (1st Cir.1992) (“We review the Secretary's findings to determine whether a preponderance of the evidence supports his conclusion that [the appellants] should be barred from bidding on government contracts.”). The Eleventh Circuit case cited by the majority, Amcor, Inc. v. Brock, 780 F.2d 897, 899 (11th Cir.1986), at best only stirs up already murky waters by mentioning “clear error” and “preponderance of the evidence” in consecutive sentences.
In addition, the majority's reliance on interpretations of the MPPAA is misplaced, since that statute frames the standard of review more conventionally: “there shall be a presumption, rebuttable only by a clear preponderance of the evidence, that the findings of fact made by the arbitrator were correct,” 29 U.S.C. § 1401(c) (emphasis added). This language simply establishes a presumption in favor of the factfinder (something like the clear error standard). This is quite different than the language of 41 U.S.C. § 39 (governing this case) which establishes no such presumption but instead instructs the reviewing body to accept findings only if they are supported by a preponderance of the evidence.
Moreover, even though the Secretary's own regulations appear to vest the Board with essentially appellate powers, appellate bodies need defer, if they are to defer at all, only to findings of historical fact; how the law applies to the facts is certainly within the de novo reviewing powers of the appellate body. The majority offers no reasons why we would defer to the ALJ's application of the law to the facts rather than accept the same application by the Board.
I therefore disagree with the majority about the degree of deference in factual matters the Board owes to the ALJ. But these differences are not crucial in resolving the question of debarment, since relief from debarment does not issue automatically from determinations of fact. Instead, as prescribed by the regulations, the absence of aggravating circumstances, the presence of mitigating elements and a sensitive balancing of related factors (potentially either aggravating or mitigating) determines whether a case presents truly unusual circumstances. An adjudicator must address, for example, a contractor's efforts to ensure compliance and a contractor's culpability, if any, in failing to make the necessary efforts. None of these determinations simply requires findings of fact; each requires a judgment about how the facts relate to the legal standards set out in the regulations. Ultimately, there must be a discretionary judgment about the relative weight to be accorded each factor.
Again, with respect to Dantran's efforts at compliance, the majority is persuaded that inspector Rioux's “clean bill of health” provides some sort of “equitable overlay,” see ante at 65, excusing Dantran's failure to inquire about or read the applicable regulations. I do not find this convincing. At the same time, the majority goes to considerable lengths to reject Dantran's estoppel argument. It follows from this emphatic denial of estoppel that any reliance by Dantran on Rioux's report to justify its own breach of regulations is unreasonable. In that context, it is not clear why fairness requires us to credit Dantran's mistaken beliefs.
In my view, equitable considerations of whatever sort are better left in the first instance to the Board, precisely because the regulations mandate a factor-specific balancing with accompanying judgments about the relative weights of the factors. The present case illustrates the potential for competing inferences and varying weights to be attached to findings of historical fact. Thus, even if the Secretary must review the ALJ's fact-finding with deference, she must still make an independent determination whether the facts add up to circumstances unusual enough for Dantran to avoid debarment. I would permit the Secretary or her delegate, the Board, to undertake this balancing anew, consistent with this opinion, of course. To that extent, I respectfully dissent.
1. In his rescript, the ALJ noted that in each instance in which testimony conflicted, he accepted the version of events offered by Holmes because he found Holmes's accounts to be more “credible, persuasive and probative.”
2. We do not find 29 C.F.R. § 4.175, entitled “Meeting requirements for health, welfare and/or pension benefits,” sufficient to rehabilitate the Secretary's position. This section provides that fringe benefit payments “are due for all hours paid for ․ up to a maximum of 40 hours per week and 2,080 hours per year on each contract.” Id. § 4.175(a)(1). Although this language caps benefit entitlement to 40 hours per week on any particular contract, it in no way implies preclusion of an absolute maximum entitlement of 40 hours per week, for the former limit legitimately can be considered a subset of the latter.
3. Take, for example, workers C, D and E. All are truck drivers whose work involves hauling mail for the same employer in the same locale. Worker C works 60 hours performing tasks in furtherance of contract X. Under the Secretary's interpretation, he receives 40 hours worth of fringe benefit payments. In the same week, worker D also works 60 hours-48 in furtherance of contract X and 12 in furtherance of contract Y. Under the Secretary's interpretation, he receives 52 hours worth of fringe benefit payments. That week, C and D's colleague, worker E, also puts in 60 hours-20 hours apiece in furtherance of contracts X, Y, and Z, respectively. Under the Secretary's interpretation, he receives 60 hours worth of fringe benefit payments.
4. We find unpersuasive the suggestion made by the plaintiffs at oral argument that we should overlook this regulation because it exalts form over substance. In mounting this offensive, the plaintiffs draw upon 29 C.F.R. § 4.165(a)(1), a regulation that permits employers to pay employees their wages for a particular pay period as late as the end of the following pay period. Taken as a whole, however, the regulations unmistakably treat a semi-monthly payment as an independent criterion for determining the propriety of compensation practices. Hence, regardless of whether an employer seeks to avail itself of the payment delay provision of section 4.165(a), it must abide by the separate and distinct semi-monthly payment requirement of section 4.165(b).
5. There is nothing unprecedented about such a switch. See 2 Davis & Pierce, supra § 11.2, at 194-99 (describing legislative departures from the substantial evidence standard).
6. In the course of its analysis, the Seventh Circuit remarked, as have we, on the problems that such a provision presents because it employs burden of proof language to limn a standard of appellate review. See Jos. Schlitz Brewing, 3 F.3d at 998-99; see also Concrete Pipe & Prods. of Cal., Inc. v. Construction Laborers Pension Trust, 508 U.S. 602, 622-25, 113 S.Ct. 2264, 124 L.Ed.2d 539 (1993) (commenting upon the dissonance resulting from the MPPAA's confusion of burdens of proof with standards of review). We agree with the Seventh Circuit that infelicitous statutory language does not relieve courts of their obligation to decipher congressional intent and thereby give coherent meaning to positive law.
7. The original regulations specifically stated that an ALJ's findings of fact were to be reviewed under the clearly erroneous standard. See Saavedra v. Donovan, 700 F.2d 496, 498 (9th Cir.1983) (citing former 29 C.F.R. § 6.14). Effective March 21, 1984, the Secretary promulgated new regulations to govern Service Contract Act actions. See 49 Fed.Reg. 10626 (codified at 29 C.F.R. Part 6) & 49 Fed.Reg. 10636 (codified at 29 C.F.R. Part 8). These rules, which are currently in effect, established a new review board (the ARB). While earlier drafts of the rules failed to clarify the board's standard of review, the Secretary, in response to comments, revised sections 8.1(d) and 8.9(b) “to provide that the [ARB] shall base its decision on the preponderance of the evidence, the statutory standard under the Service Contract Act.” Id. at 10636. Accordingly, there is no indication that the Secretary intended her standard to be interpreted differently than the statutory standard.
8. We note that, even in instances in which courts defer to the agency as opposed to the hearing officer, judicial scrutiny becomes more exacting when the agency overturns a hearing examiner's credibility-based findings of fact. See, Kimm v. Department of Treasury, 61 F.3d 888, 892 (Fed.Cir.1995); Aylett v. Secretary of HUD, 54 F.3d 1560, 1566 (10th Cir.1995). NLRB v. Matouk Indus., Inc., 582 F.2d 125, 128 (1st Cir.1978).
9. The ALJ found that the plaintiffs attempted to comply with the regulations in good faith; that they cooperated fully with the Secretary's investigation; that they promptly settled their account and changed their monthly payment practice once the matter was brought to their attention; that nothing in their past compliance history reflected adversely on them; and that, in all events, the alleged violations were not especially serious.
10. We must recall that the regulatory scheme with which we are dealing is designed to debar those whose conduct is culpable and to excuse those whose actions invite leniency. As a general matter, therefore, reasonable good-faith reliance on the official assessments of a responsible government functionary removes a government contractor from the former category and places it within the latter. Cf. OPM, 496 U.S. at 428-29, 110 S.Ct. 2465 (discussing instances in which Congress has acted to provide appropriate relief for individuals who have relied on government misrepresentations to their detriment).
1. “Since the Act's fringe benefit requirements are applicable on a contract-by-contract basis, employees performing on more than one contract subject to the Act must be furnished the full amount of fringe benefits to which they are entitled under each contract and applicable wage determination.” 29 C.F.R. § 4.172.
2. “[E]very employee performing on a covered contract must be furnished the fringe benefits required by that determination for all hours spent working on that contract up to a maximum of 40 hours per week and 2,080 (i.e. 52 weeks of 40 hours each) per year, as these are the typical number of nonovertime hours of work in a week, and in a year, respectively.” 29 C.F.R. § 4.172.
SELYA, Circuit Judge.