This case is before us on the application of the National Labor Relations Board (the "NLRB") for enforcement of its order against Cal-Maine Farms, Inc. The NLRB's order issued on April 30, 1992. We have jurisdiction under § 10(e) of the National Labor Relations Act, 29 U.S.C. §§ 151 et seq., 160(e). After a thorough review of the record, we enter judgment enforcing the NLRB's order.

Cal-Maine Farms, Inc., is a large commercial agricultural operation that produces and processes hen-laid eggs in its plants located in various southern states. In a NLRB election conducted on March 30, 1988, employees of Cal-Maine Farms' 4000-acre Edwards, Mississippi plant voted 60-4 in favor of having the United Food and Commercial Workers International Union, Local 1529, AFL-CIO-CLC ("the union") represent employees. Following the NLRB's certification of the union shortly thereafter, Cal-Maine's management refused to recognize or bargain with the union. The management claimed it was justified in refusing to bargain on the ground that the workers at the Edwards plant were not "employees" within the meaning of the National Labor Relations Act ("the Act") and instead were "agricultural laborers." "Agricultural laborers" are explicitly excluded from the coverage of the Act. See 29 U.S.C. § 152(3).1

In May 1988, the union filed a charge with the NLRB stating that Cal-Maine had failed to recognize the union or bargain with it in good faith, as required by the Act. General Counsel for the NLRB agreed with the union, and a complaint issued in June 1988. After an administrative hearing in May 1989, the administrative law judge ("the ALJ") found that Cal-Maine violated §§ 8(a)(1) & (5) of the Act. At the time of that hearing, the rule governing the exemption of agricultural workers had been set forth by the NLRB in DeCoster Egg Farms, 223 N.L.R.B. 884, 1976 WL 6879 (1976). The so-called "single-egg test" of DeCoster "limit[ed] the exemption to those processors who deal exclusively with their own goods." Id. (emphasis added). That is, the procurement of even a single egg produced from outside sources would preclude classification of a commercial farming operation's workers as "agricultural laborers." In light of DeCoster, the ALJ defined the issue for adjudication at the May 1989 hearing as "whether [Cal-Maine's] employees at the Edwards egg packing plant processed only eggs produced at that facility after April 1, 1988, or also processed [any] eggs produced elsewhere." The ALJ found that outside eggs had been processed at Cal-Maine's Edwards plant and thus held that Cal-Maine's management had violated the Act by refusing to recognize or bargain in good faith with the union.

While Cal-Maine's appeal of the ALJ's initial decision was pending with the NLRB, the agency issued its decision in Camsco Produce Co., 297 N.L.R.B. 905, 1990 WL 122306 (Mar. 15, 1990), which overruled DeCoster to the extent it was inconsistent with the formulation of the rule announced in Camsco. Camsco held that the Act's exemption for "agricultural laborers" would be based on whether the employees in question "regularly" handle "any" amount of the products of outside producers.2 Thus, the "single egg test" of DeCoster was replaced by Camsco 's "regularity" standard. Camsco also required that the party seeking exemption from the Act--in this case, Cal-Maine--has the burden to establish that outside produce is not regularly handled by the employees seeking representation. Camsco, 297 N.L.R.B. 905. The NLRB subsequently remanded the case to the ALJ to consider the impact of the new rule in Camsco on the result in this case.

The ALJ issued his supplemental decision after a supplemental hearing in October 1990. In that decision, the ALJ reaffirmed his original position. On appeal to the NLRB, the ALJ's decision was affirmed by a vote of three to one. See Cal-Maine Farms, Inc., 307 NLRB No. 66, 1992 WL 101249, 1992 NLRB LEXIS 609 (April 30, 1992). The NLRB then applied to this court for a judgment enforcing its order.

It is the NLRB's "special duty" to apply the National Labor Relation Act's exemption for agricultural laborers "to varying fact patterns." Bayside Enterprises, Inc. v. NLRB, 429 U.S. 298, 304, 97 S. Ct. 576, 581, 50 L. Ed. 2d 494 (1977). In performing that duty, the NLRB is charged with construing the Act--including its incorporation of the term "agricultural laborer" as used in the Fair Labor Standards Act--"liberally in favor of the workers" for whose protection those laws were designed, and that any exemption from the terms of those laws must be "narrowly construed." Wirtz v. Ti Ti Peat Humus Company, Inc., 373 F.2d 209, 212 (4th Cir.), cert. denied, 389 U.S. 834, 88 S. Ct. 37, 19 L. Ed. 2d 94 (1967); see also NLRB v. Security Guard Service Inc., 384 F.2d 143, 147 (5th Cir.1967) (recognizing "the standard reluctance to apply [a statutory] exception broadly").

The NLRB's "expert" construction of the agricultural-laborer exemption is entitled to deference on review, because it presents a question that is "particularly unsuitable" for a reviewing court. NLRB v. Design Sciences, 573 F.2d 1103, 1104 (9th Cir.1978); see also Bayside Enterprises, 429 U.S. at 304 n. 14, 97 S. Ct. at 581 n. 14 (where the NLRB has construed the agricultural-laborer exemption, a reviewing court has a "limited" function and must give "appropriate weight" to the NLRB's judgment). Accordingly, the NLRB's determination that particular workers are statutory employees and not agricultural laborers must be upheld " 'if it has warrant in the record and a reasonable basis in law.' " Bayside Enterprises, 429 U.S. at 304 n. 14, 97 S. Ct. at 581 n. 14 (quoting NLRB v. Hearst Publications, Inc., 322 U.S. 111, 126, 64 S. Ct. 851, 858, 88 L. Ed. 1170 (1944)); see also NLRB v. Design Sciences, 573 F.2d at 1104.

This court reviews the NLRB's factual determinations under the well-known "substantial evidence" standard announced by the Supreme Court in Universal Camera Corp. v. NLRB, 340 U.S. 474, 71 S. Ct. 456, 95 L. Ed. 456 (1951). The Court defined substantial evidence as "more than a scintilla. It means such relevant evidence as a reasonable mind would accept as adequate to support a conclusion." Id. at 477, 71 S.Ct. at 459. Substantial evidence "must be enough to justify, if the trial went to a jury, a refusal to direct a verdict when the conclusion sought to be drawn from it is one of fact for the jury." Id. A reviewing court must consider the totality of evidence in the record, including "that which fairly detracts from the [NLRB's] decision." Universal Camera, 340 U.S. at 488, 71 S.Ct. at 464.

However, in determining whether the NLRB's factual findings are warranted by the record, this court will not "pass on the credibility of witnesses or reweigh the evidence." Helena Laboratories Corp. v. NLRB, 557 F.2d 1183, 1187 (5th Cir.1977). Indeed, where a case turns on witness credibility, this court will accord special deference to the NLRB's credibility findings and will overturn them "only in the most unusual of circumstances." Centre Property Management v. NLRB, 807 F.2d 1264, 1268 (5th Cir.1987); see also NLRB v. Ryder/P.I.E. Nationwide, Inc., 810 F.2d 502, 507 (5th Cir.1987) ("great deference" accorded to NLRB's credibility findings); NLRB v. Florida Medical Center, Inc., 576 F.2d 666, 671 (5th Cir.1978) (credibility findings will be upheld unless "self-contradictory").

There is no dispute among the parties that, as a general principle of labor law, the employees of a farming operation that processes agricultural products exclusively grown on the premises of the farm itself would not be subject to the terms of the Act. Rather, the parties dispute whether there is "substantial evidence," see Universal Camera, supra, to support the NLRB's finding that Cal-Maine "regularly" procured "any" eggs from outside sources after April 1, 1988.3 As a collateral matter, the parties are in dispute whether the NLRB, in keeping with its prior decision in Camsco Produce Company, Inc., 297 N.L.R.B. 905, 1990 WL 122306 (1990), made sufficient findings in the instant case regarding the "regularity" of Cal-Maine's procurement of any such outside eggs.

After a thorough review of the voluminous record, we conclude both that there is substantial evidence supporting the NLRB's decision and that the agency did make proper findings regarding the "regularity" of the outside eggs handled by Cal-Maine employees. Thus, the NLRB was correct in finding that Cal-Maine violated the provisions of the Act that require employees to recognize and bargain in good faith with a NLRB-certified union. Our review here will summarize the primary pieces of evidence relied upon by the ALJ and the NLRB.4 We note at the outset that the burden of proof in this case was on Cal-Maine to establish by a preponderance of the evidence that it did not regularly procure any amount of outside eggs.

Two of the General Counsel's witnesses testified that they observed distinctive types of dollies5 and shipping stickers on the dollies which indicated that outside eggs were received after April 1, 1988, the date that Cal-Maine claims that such shipments ceased. Employees Daisy Bishop and Virginia Foster testified that they saw dollies and stickers on dollies from other Cal-Maine plants arrive at the packing plant in December 1988 and January and March 1989. In addition, employee Larry Bishop testified that he saw a tractor-trailer arrive with a shipment of unprocessed outside eggs at the packing plant in early 1989. The ALJ credited these three witnesses' testimony.

Cal-Maine has challenged these three employees' testimony on various grounds. With respect to the two employees who claimed to have seen dollies and stickers from outside sources, Cal-Maine argues that the presence of dollies and stickers from other locations may be explained: According to Cal-Maine's witnesses at the administrative hearing, Cal-Maine regularly intermingles empty dollies from its various plants; furthermore, many of the stickers from other Cal-Maine plants allegedly were left on such intermingled dollies from previous shipments and did not indicate that egg shipments actually came to the Edwards plant from other Cal-Maine facilities after April 1, 1988. Although the ALJ appeared to agree that empty dollies may have been intermingled, the ALJ discredited Cal-Maine's explanation for the allegedly out-dated stickers by finding, based on competent evidence, that Cal-Maine had a policy of removing out-dated stickers and that the presence of the stickers on more than one occasion was not adequately explained by Cal-Maine's claim that the stickers were simply left on the dollies inadvertently.

As for Larry Bishop's testimony about the truck, the ALJ noted that, based on undisputed testimony, an eighteen-wheeler truck would not ordinarily be used to move eggs within Cal-Maine's Edwards plant and that such vehicles were used only to make outside deliveries, such as those that regularly occurred before April 1, 1988. One of Cal-Maine's managers attempted to explain the presence of the truck by claiming that on one occasion eggs had been moved within the Edwards complex via an eighteen-wheeler. The ALJ discredited that witness' explanation on the grounds that the same witness had previously claimed that eighteen-wheelers were not used to move eggs produced within the Edwards complex and only "remembered" the alleged single instance in order to rebut Larry Bishop's testimony.

The NLRB affirmed the ALJ's findings. Our review of the relevant testimony leaves us in agreement with the NLRB. Although this evidence, by itself, would not establish that Cal-Maine regularly procured outside eggs, it is probative evidence that may be considered along with other evidence, discussed infra.

In his testimony at the hearing, Cal-Maine's shipping and receiving supervisor, Robert Turner, flatly denied that outside eggs had been received at Edwards after April 1, 1988. The ALJ noted, however, that Turner also stated that the packing plant received outside eggs until around the time that a certain record-keeping change was made, in November 1988. Furthermore, when asked about Cal-Maine's policy of removing out-dated dolly stickers by employees, Turner replied: "I don't have a lot of time to go pulling stickers off of 28 dollies or eight dollies or whatever is on the load. If it is an outside load, you would have at least 43 or 45 dollies on there." According to undisputed evidence, loads of 43 or 45 dollies necessarily would be from outside of the Edwards plant, delivered on eighteen-wheeler trucks, as opposed to the smaller vehicles used to make internal shipments of eggs within the Edwards complex. The ALJ found that Turner's present-tense statement was in effect an admission that the Edwards plant had recently been receiving outside shipments of eggs.

Cal-Maine argues that the ALJ misinterpreted Turner's statements which the ALJ considered to be evidence that the Edwards plant was in fact receiving outside shipments after April 1, 1988. We have examined Turner's statements in context and agree that they are by no means conclusive admissions by him that the Edwards plant continued to receive outside shipments after April 1, 1988. At the same time, however, we do not agree with Cal-Maine that Turner's statements have no probative value. In particular, his use of the present tense in his statements about "twenty-eight," "forty-three," and "forty-five" dollies raises a legitimate question about whether such outside shipments were still being received around the time of the 1989 hearing. Again, we note that such evidence by itself is not sufficient to support the ALJ and NLRB's findings. However, it is probative evidence that was properly considered by the NLRB along with other evidence.

In addition to its processing plant located in Edwards, Mississippi, Cal-Maine also has an office in Mendenhall, Mississippi, where Cal-Maine contracts with local egg farmers and receives deliveries from them that are shipped to Cal-Maine's various processing plants. Cal-Maine does not dispute that until April 1, 1988, eggs procured from outside sources in Mendenhall were transported to the Edwards plant. At the original hearing in May 1989, Cal-Maine introduced the Mendenhall facility's shipping log through December 31, 1988, purporting to show no shipments to the Edwards plant after March 31, 1988. In addition, Cal-Maine introduced partial receiving logs for the Edwards plant, which did not record any Mendenhall receipts after April 1, 1988. One of Cal-Maine's witnesses, packing plant manager James King, stated that he could not recall how Mendenhall receipts were recorded, but speculated that they were listed in a separate "outside" log, which was not offered into evidence.

After the ALJ noted in its initial decision that the records introduced were incomplete and therefore unreliable, Cal-Maine, at the second hearing, introduced one page, dated April 1, 1988, purporting to be part of the "outside log." The document contained no references to egg receipts from outside sources. The ALJ again discredited the purported log and the NLRB agreed, holding that "the proffer of the record from only one date does not inspire confidence that the records submitted in this proceeding are reliable evidence that all outside deliveries ceased on April 1, 1988." We have examined the various records and concur with the conclusions reached by the ALJ and NLRB.

Cal-Maine presented a witness who stated he was a truck foreman at the Mendenhall shipping and receiving office. He flatly denied that eggs from Mendenhall were delivered to the Edwards plant after April 1, 1988. The ALJ did not credit his testimony because the witness was found to be "evasive" and, thus, unreliable. Furthermore, the ALJ noted that, although the Mendenhall truck foreman was one of six or seven drivers who could have made such deliveries, Cal-Maine called none of the others to testify. The NLRB refused to disturb the credibility finding of the ALJ. In view of the tremendous deference that we afford to the trier-of-fact on credibility matters, nor will we.

E. Cal-Maine's Correlation Between Production Records and "Run Sheets"

At the hearing, Cal-Maine introduced an analytical document purporting to correlate egg-receipt log entries with "run sheets," i.e., weekly records that track opening inventory, the number of eggs laid, and the number actually processed. Through this document, Cal-Maine sought to show that hens in the Edwards facility laid enough eggs to account for the processing totals at the packing plant. The ALJ found, however, that the analysis revealed significant discrepancies, both "overages" and shortages, as it tracked each week's beginning inventory, production, and processing totals. The ALJ found that these discrepancies "create further doubts that [Cal-Maine's] documentary evidence accurately reflects egg receipts at the Edwards farm."

The NLRB agreed with the ALJ that, if anything, Cal-Maine's documentation actually detracted from, rather than supported, Cal-Maine's position. The NLRB refused to consider Cal-Maine's attempt to explain the discrepancies, noting that Cal-Maine's "attempt in its brief to explain these discrepancies ... was provided only in the brief, and not based on record evidence [and, accordingly, that] there was no opportunity for the parties or the judge to explore its premises and methodology. We therefore decline to rely on [Cal-Maine's] discussion of its 'correlation.' "

On appeal to this court, Cal-Maine once again attacks the ALJ's interpretation of the production records and run sheets and argues that the ALJ's finding that the document showed significant "overages" of eggs on fifteen different occasions was unfounded. For essentially the same reasons offered by the NLRB, we refuse to accept Cal-Maine's belated explanation and instead accept the ALJ's interpretation of the document. Moreover, we observe that Cal-Maine, in its original brief to this court, does not attack the ALJ's interpretation of the records and only does so for the first time in its reply brief. See Reply Brief for the Respondent, at p. 9. As this court has repeatedly held, we will not review arguments raised for the first time in a reply brief. See, e.g., United States v. Clinical Leasing Service, Inc., 982 F.2d 900, 902 n. 4 (5th Cir.1992).

F. Conversation Between Supervisor Meyers and Employees Bishop and Foster

At the hearing, two employees of Cal-Maine, Daisy Bishop and Virginia Foster, each claimed that their supervisor, William C. Meyers, told them in December 1988 that the management at Cal-Maine had "already bought some [outside] eggs and [was] going to buy more," so as to assure that there was a sufficient supply of eggs available for processing in response to the increased demand during Christmas. The ALJ credited the two employees' testimony and discredited the supervisor's claim to the contrary. Again, the NLRB refused to disturb the ALJ's credibility findings.

On appeal to this court, Cal-Maine argues that the ALJ and NLRB erred in giving any weight to the two employees' claims because their testimony was hearsay.6 To begin with, as the NLRB observes, Cal-Maine did not object to the two employees' testimony on hearsay grounds at the administrative hearing. Thus, their hearsay objection is waived. Even if the claim were not waived, we would find that it is non-hearsay under the party-opponent admission exception to the hearsay rule. See FED.R.EVID. 801(d)(2)(D) (admission of employer's "agent or servant concerning a matter within the scope of the agency or employment" admissible as non-hearsay). The statements made by Cal-Maine's supervisor are highly probative evidence supporting the NLRB's finding that outside eggs were procured well after April 1, 1988.

2) Is the NLRB's decision based on "substantial evidence"?

Cal-Maine argues that the agency's finding that outside eggs were "regularly" procured by Cal-Maine is not based on substantial evidence. We disagree. The agency's decision was based on numerous pieces of evidence that, in the aggregate, undoubtedly support its finding. Particularly probative is the testimony of Daisy Bishop and Virginia Foster, who testified not only that they had observed various dollies and stickers from some of Cal-Maine's other plants, but also that a leading member of Cal-Maine's management had admitted to them each that Cal-Maine was procuring eggs from outside sources in December of 1988, many months after Cal-Maine allegedly ceased procuring outside eggs. We also are impressed by Cal-Maine's own documentary evidence offered that shows substantial weekly "overages" of eggs in the Edwards plant's logs after April 1, 1988.

Cal-Maine argues that the ALJ and NLRB simply ignored the voluminous documentary evidence purporting to show that no eggs were shipped to Cal-Maine from any outside sources after April 1, 1988. We agree with the ALJ that such evidence, while probative, is hardly conclusive. The various documents--which constitute numerous volumes of the appellate record--were admitted by the ALJ under the "business records" exception to the hearsay rule. See FED.R.EVID. 803(6). We observe that the commentary to the pertinent rule of the Federal Rules of Evidence discusses many problems associated with evidence of business records, noting, inter alia, that "hesitation must be experienced in admitting everything which is observed and recorded in the course of regularly conducted activity." Id. (Advisory Committee Notes). And although the fact "that records might be self-serving has not been a ground for exclusion" of business records, id., the potential certainly is a permissible matter for a trier-of-fact to consider in deciding how much weight to afford the business records of a party offered in support of its position. This is precisely what the ALJ did in the instant case. Moreover, in view of the various pieces of evidence that belie what Cal-Maine's business records purport to show--including aspects of the records themselves, see supra discussion--we refuse to disturb the agency's decision that Cal-Maine's documentary evidence should be given little weight.

Finally, we must address Cal-Maine's claim that neither the NLRB nor the ALJ made sufficient findings regarding "regularity," as required by the NLRB's decision in Camsco. We disagree. Although the ALJ and NLRB did not make specific findings regarding precisely when and precisely how many outside eggs were handled by Cal-Maine after April 1, 1988, we do not read Camsco as requiring such precision. We believe that the ALJ and NLRB made sufficient findings regarding "regularity" in finding, inter alia:

i) that there were "overages" of eggs in Cal-Maine's records, which were documented over the course of many months;7

ii) that two employees each observed dollies and shipping stickers indicating that outside eggs had been shipped in late 1988 and early 1989;8

iii) that a third employee observed at least one eighteen-wheeler unload "outside" eggs at Cal-Maine's Edwards plant around Easter of 1989;9 and

iv) that a Cal-Maine supervisor admitted to two employees that the company had been acquiring "outside" eggs in December of 1988.

In sum, we believe that the record contains sufficient evidence to support the NLRB's finding that Cal-Maine "regularly" procured eggs from outside sources after April 1, 1988. Thus, the NLRB was correct in holding that Cal-Maine has violated §§ 8(a)(1) & (5) of the National Labor Relations Act by refusing to recognize and bargain in good faith with the NLRB-certified union. Accordingly, we enter judgment enforcing the order of the National Labor Relations Board in Cal-Maine Farms, Inc., No. 15-CA-10588.