|UNITED STATES DEPARTMENT OF AGRICULTURE,|
(Agency No. D-99-0010)
Stephen M. Reilly, Senior Counsel (James Michael Kelly, Deputy General Counsel, and Margaret M. Breinholt, Assistant General Counsel, with him on the brief), Office of the General Counsel, United States Department of Agriculture, Washington, D.C., for Respondent.
Excel, a corporation based in Wichita, Kansas, is estimated to be the fourth or fifth largest hog slaughterer in the United States. ROA, Vol. V, Doc. 155 at 13, 82. Excel purchases hogs from numerous hog producers using one of two methods. First, Excel purchases some hogs on a "spot" market basis, meaning that it negotiates a specific price for a specific lot of hogs. Id. at 13. Second, Excel purchases other hogs through short- and long-term contracts with hog producers, pursuant to which the producers agree to sell a given number of hogs to Excel for a set base price. Id.
Most of the hogs purchased by Excel fall within its "carcass merit" program. Id. Under the carcass merit program, hog producers deliver hogs to Excel's buying stations where the hogs are placed into a holding pen, tattooed for identification, given a lot number, weighed, and inspected. Id. at 13-14. The hogs are then transported to one of Excel's three slaughtering facilities (located in Illinois, Iowa, and Missouri). There, the hogs are "killed, bled, eviscerated, de-haired, washed, and inspected . . . ." Id. at 14. Afterwards, the carcasses are evaluated for their "estimated percentage of lean (red) meat." Id. Because hogs with a high percent of lean meat have a higher market value than hogs with a low percent of lean meat, Excel "applies th[ese] percentage figure[s] to a pricing table called the 'lean percent matrix' to determine whether the hog producer receives a discount for the carcass a deduction from the base price or a premium an addition to the base price." Id.
Some of the producers who supply hogs to Excel also sell to other packers. Id. at 20. Generally speaking, these producers sell "trial lots" to various packers, including Excel, to determine where they can obtain the best price. Id. Because USDA no longer has in place an official grading system for hogs, id. at 16, "[a]ll packers appear to base the prices they pay for hogs on base price, lean percent, and a matrix . . . ." Id. at 20. However, no industry standard exists for estimating lean percent and it is generally impractical for slaughterers to dissect and examine each carcass for fat and lean meat percentages. Id. at 14. Thus, slaughterers use a variety of less accurate, but more practical, methods of estimating lean percent. Id. The result is that each packer "has a slightly different grading program," i.e., "[t]hey use slightly different means of getting to the same point for the end value." Id. at 20.
Excel had used the "Fat-O-Meat'er" method for estimating lean percent for approximately ten years. Id. at 14. "The Fat-O-Meat'er," which was developed in Denmark from a study of European hogs, "is a hand-held device with a probe that is inserted in the carcass." Id. "A light measures the difference between the loin-eye and back fat depth." Id. "A regression formula or equation embedded in the Fat-O-Meat'er, commonly referred to as the 'Danish formula' . . , then uses this measurement to estimate the lean percent of the carcass." Id. at 14-15. The device has been approved for use by the USDA and is used by approximately thirty-two packers in the United States. Id. at 15. It is unclear, however, how many of these packers rely solely on the Danish formula to estimate lean percent. Id.
After Excel determined the lean percent and weight of each carcass, those figures were applied to Excel's "Lean Value Matrix" to determine the "meat PX factor." Aplee. Br. at 12. The matrix generated a higher "meat PX factor" for standard-sized carcasses (163 to 206 pounds) with a higher lean percent. Conversely, the matrix produced a lower "meat PX factor" for non-standard-sized carcasses (greater than 206 pounds or less than 163 pounds) and for carcasses with a lower lean percent. Id. To determine the exact price to be paid for a particular carcass, Excel multiplied the "meat base" (i.e., the price per hundred weight quoted to the producer) by the "meat PX factor." Id.
The producers from whom Excel purchased hogs on a carcass merit basis were aware that Excel used the Fat-O-Meat'er to estimate lean percent and that the lean percentage figure was used by Excel to determine the price paid for each carcass. Generally speaking, however, Excel did not inform producers of the details of the formula utilized for estimating lean percent.
In 1997, Excel decided to switch from the Danish formula for estimating lean percent to "a formula developed by Purdue University and promoted by the National Pork Producers Council," i.e. "the Purdue formula." Id. at 17. "The Purdue formula uses hot carcass weight as a variable with the Danish formula to estimate lean percent . . . ." Id. In contrast to the Danish formula, which was estimated to be 72-73 percent accurate, the Purdue formula was estimated to be approximately 90 percent accurate. Id.
At the time it adopted the Purdue formula, Excel knew that the "change could affect the price it paid for hogs," and thus "considered the" change's "economic effect on hog producers . . . ." Id. Excel "concluded, based on a study of 1.5 million hogs, that there would be only a 'minimal impact' on hog producers . . . ." Id. at 17-18. In turn, Excel "decided not to tell hog producers about the change in the formula because, while it was not a secret, company officials believed that the formula, like the process methods and technology it used, was not a factor that interested hog producers or formed a basis for whether they sold hogs to" Excel. Id. at 18. "Another consideration was the corporate belief that hog producers who received more because of a change to a more accurate formula would be unhappy because they had been selling in the past under an inaccurate formula, while hog producers who received less because of the change would be upset . . . ." Id.
Although Excel concluded that none of its written contracts with hog producers required it to provide notification of the formula change, Excel nevertheless notified Tyson Foods, the main supplier of hogs for Excel's Missouri facility, of the formula change. Id. at 19. Tyson objected to the change. Id. In turn, Excel agreed not to use the Purdue formula to estimate the lean percent of Tyson's hogs. Id.
Excel implemented the formula change at its Iowa and Illinois slaughtering facilities in October 1997, and at its Missouri slaughtering facility (for all non-Tyson hogs) in April 1998. Id. at 20. Following implementation of the formula change, some hog producers noticed a difference in the prices they were receiving from Excel for hogs. Id. at 21. Some hog producers began asking managers at Excel's slaughtering facilities about the matter. Id. In response, Excel told these producers about the formula change. Id.
In April 1998, the Grain Inspection, Packers and Stockyards Administration (GIPSA), a division of the USDA, "initiated what appears to have been a routine investigation of [Excel's] use of the Fat-O-Meat'er." Id. at 22. During the course of this audit, GIPSA "found the prices that hog producers should have been paid using the Danish formula were not those that appeared on the kill sheets." Id. at 23. Excel then informed GIPSA that it had changed the formula for estimating lean percent. Id. As a result of the 1998 audit, GIPSA decided that Excel's "failure to disclose its change of the formula to hog producers prior to the purchase of hogs from those producers" was a violation of the P&S Act and one of its implementing regulations. Id. at 25. Excel was informed of the alleged violation in June 1998. Id. In July 1998, Excel "sent a letter to hog producers notifying them that the formula had changed . . . ." Id. Excel "also adjusted the matrix so that hog producers received the same price under the Purdue formula as they would have received had [Excel] used the Danish formula." Id.
On April 9, 1999, the Deputy Administrator of GIPSA instituted a disciplinary administrative proceeding against Excel by filing a complaint and notice of hearing. The complaint alleged that, between October 23, 1997, and June 1, 1998, Excel violated § 202(a) of the P&S Act, 7 U.S.C. § 192(a), and § 201.99 of the Act's implementing regulations, 9 C.F.R. § 201.99, by failing to make known to hog producers a change in the formula used by Excel to estimate lean percent in hogs that it purchased, which in turn changed the price paid by Excel for hogs. The complaint further alleged that, as a result of the change in formula, Excel paid hog producers approximately $1,839,000 less for approximately 19,942 lots of hogs than it would have paid if it had not changed the formula.(1)
USDA's Chief Administrative Law Judge (ALJ) conducted hearings on July 18-21, July 25-28, September 23-27, 2000, and March 27-29, 2001. On February 7, 2002, the Chief ALJ issued a Decision and Order finding that, as alleged in the complaint, Excel failed to notify hog producers of its changed formula for estimating lean percent and that such failure violated § 202(a) of the P&S Act, 7 U.S.C. § 192(a), and § 201.99 of the implementing regulations, 9 C.F.R. § 201.99. The Chief ALJ ordered Excel to cease and desist from failing to notify livestock sellers of any change in the formula used to estimate lean percent and further ordered Excel to submit to arbitration with the hog producers with whom they had not yet resolved the matter and who received less money for hogs sold to Excel between October 1997 and July 1998 under the revised formula than they would have received under the old formula. The Chief ALJ refused GIPSA's request, however, to impose a monetary sanction against Excel.
Excel and GIPSA each sought review of the Chief ALJ's decision by the Secretary of the USDA. On January 30, 2003, the Judicial Officer (JO) issued a decision and order on behalf of the USDA addressing the challenges to the Chief ALJ's order. The JO affirmed the decision that Excel violated the P&S Act and the implementing regulation by failing to make known to all hog producers its change in the formula used to estimate lean percent in hogs. The JO dismissed the arbitration requirement and modified the cease and desist order. The JO agreed with the Chief ALJ that a monetary sanction was not appropriate.
Both sides unsuccessfully sought reconsideration of the JO's decision and order. Excel has since filed a petition for review with this court.
Our jurisdiction to review a final order issued by the USDA in a disciplinary action brought under the P&S Act arises under 28 U.S.C. § 2342(2). We review such final orders under the Administrative Procedure Act's ("APA") arbitrary and capricious standard. See JSG Trading Corp. v. USDA, 176 F.3d 536, 541 (D.C.Cir. 1999). "That is, we will uphold the JO's decision unless we find it to be arbitrary, capricious, an abuse of discretion, not in accordance with law, or unsupported by substantial evidence." Id. (citing 5 U.S.C. § 706(2)(A), (E)).
Before addressing Excel's specific arguments on appeal, we begin by briefly outlining the statute and regulation that the JO determined Excel had violated. Section 202 of the P&S Act, 7 U.S.C. § 192, entitled "Unlawful practices enumerated," provides in pertinent part as follows:
It shall be unlawful for any packer or swine contractor with respect to livestock, meats, meat food products, or livestock products in unmanufactured form, or for any live poultry dealer with respect to live poultry, to:
(a) Engage in or use any unfair, unjustly discriminatory, or deceptive practice or device . . . .
7 U.S.C. § 192(a).
In turn, the USDA has promulgated regulations implementing the provisions of the P&S Act. Specifically, 9 C.F.R. § 201.99, entitled "Purchase of livestock by packers on a carcass grade, carcass weight, or carcass grade and weight basis," provides, in pertinent part, as follows:
(a) Each packer purchasing livestock on a carcass grade, carcass weight, or carcass grade and weight basis shall, prior to such purchase, make known to the seller, or to his duly authorized agent, the details of the purchase contract. Such details shall include, when applicable, expected date and place of slaughter, carcass price, condemnation terms, description of the carcass trim, grading to be used, accounting, and any special conditions.
* * *
(e) * * * If settlement and final payment are based upon any grades other than official USDA grades, such other grades shall be set forth in detailed written specifications which shall be made available to the seller or his duly authorized agent. * * *
9 C.F.R. § 201.99(a) and (e) (italics added).
Applying the statute and the regulation to the established facts, the JO concluded that a violation of both the regulation and the statute had occurred. In particular, the JO noted that "[t]he record [wa]s clear that all parties considered the Fat-O-Meat'er to be a form of grading." ROA, Vol. V, Doc. 155 at 41. In turn, the JO concluded that "[t]he formula [Excel] used to estimate lean percent was also a part of the 'grading' within the meaning of section 201.99 of the Regulations . . . as it was an element of [Excel's] carcass evaluation process." Id. The JO further concluded that, because "[s]ection 201.99 of the Regulations . . . explicitly provides that packers purchasing livestock on a carcass merit basis must make known to the seller the grading to be used prior to purchase," Excel violated that provision by failing to inform hog producers of its change in formula for determining lean percent. Id. In addition, the JO concluded that the violation had a direct impact on the hog producers who sold hogs to Excel. According to the JO, "the purpose of section 201.99 of the Regulations . . . is to provide some basic level of similarity to allow sellers to evaluate different purchase offers," id. at 42 (internal quotations omitted), and Excel deprived hog producers of this opportunity by failing to disclose its change in formula. More specifically, the JO stated: "Had hog producers been alerted to the change, they could have shopped their hogs to other packers to determine if they could obtain a better price for their hogs than [Excel's] price under its changed formula." Id. Ultimately, the JO concluded that Excel "violated section 202(a) of the [P&S] Act and section 201.99(a) of the Regulations . . . when it failed to make known to hog producers that it was changing the formula to estimate lean percent, prior to purchasing hogs on a carcass merit basis from those producers." Id. at 83.
In its appeal, Excel contends the "JO erred when he ruled that Excel violated the law by changing the lean percent equation without prior notice" because "the USDA never met its burden to demonstrate that there was substantial evidence for this finding." Aplt. Br. at 15. In support of this contention, Excel argues that (1) "the JO never cited to a single court case or prior agency decision that provides any precedential support," (2) "the JO did not and could not rely on any expert testimony because GIPSA provided none," and (3) "GIPSA failed to introduce any survey of hog producers that producers believed that Excel had committed an unfair or deceptive practice or that these producers cared that Excel had changed the lean percent equation without disclosing the change to producers." Id.
By raising these arguments, Excel is clearly attempting to reframe the nature of the JO's decision. Generally speaking, it is true that an agency's decision must be supported by "substantial evidence." Trimmer v. United States Dept. of Labor, 174 F.3d 1098, 1102 (10th Cir. 1999). Here, however, the JO expressly noted in his decision and order that "[t]he salient facts [of the case] [we]re not in dispute." DO at 27. In particular, the JO noted that "all parties considered the Fat-O-Meat'er to be a form of grading," id. at 41, and "[t]he parties [we]re in agreement that [Excel] did not tell all hog producers when it changed the formula to estimate lean percent and did not disclose details of the formula to all hog producers." Id. Thus, the JO's decision ultimately was based on whether those established facts constituted a violation of § 201.99 (and, in turn, § 202(a) of the P&S Act). In other words, the JO's decision was based on his interpretation of § 201.99 and his application of that interpretation to the uncontroverted facts.
The absence of any true factual disputes is further highlighted by carefully examining Excel's specific arguments. As noted, Excel first complains that "the JO never cited to a single court case or prior agency decision that provides any precedential support" for his decision. Aplt. Br. at 15. Obviously, however, prior court cases or agency decisions are not "evidence" that would support or refute the JO's decision. Second, Excel complains that "the JO did not and could not rely on any expert testimony because GIPSA provided none . . . ." Id. It is unclear, however, why any such expert testimony was necessary. To the contrary, the resolution of the USDA's complaint against Excel required the JO only to apply the provisions of § 201.99 to the uncontroverted facts developed during the evidentiary hearing. Lastly, Excel complains that "GIPSA failed to introduce any survey of hog producers that producers believed that Excel had committed an unfair or deceptive practice or that these producers cared that Excel had changed the lean percent equation without disclosing the change to producers." Again, no such evidence was necessary to support the JO's conclusion. Indeed, the JO rejected this identical argument in his decision and order:
Finally, I find [Excel's] argument that, when hog producers learned about the formula change, they did not care that the change had been made or that [Excel] failed to inform them about the formula change, irrelevant to the issue of whether [Excel] violated the Packers and Stockyards Act. [Excel] cites no authority supporting its contention that the feelings of hog producers have a bearing on whether [Excel] engaged in an unfair or deceptive practice under section 202(a) of the Packers and Stockyards Act . . , and I cannot find authority which supports [Excel's] contention. The determination as to whether [Excel] violated section 202(a) of the Packers and Stockyards Act . . . is made by the administrative law judge, the judicial officer, and ultimately, the courts. The determination is not based on how livestock producers, who the Packers and Stockyards Act is designed to protect, view [Excel's] actions. Moreover, the record does not support [Excel's] assertion that hog producers did not care about [Excel's] change in the formula to estimate lean percent or [Excel's] failure to inform them about the formula change . . . .
DO at 65-66.
For these reasons, we conclude there is no merit to Excel's assertion that the JO's decision was not supported by substantial evidence.
In its opening appellate brief, Excel asserts a host of arguments concerning why, in its view, it did not violate the P&S Act by "[c]hang[ing] an [e]quation [u]sed to [e]stimate [l]ean [p]ercent . . . ." Aplt. Br. at ii. In particular, Excel argues that (1) no prior decisions existed holding that an undisclosed equation change was violative of the P&S Act, (2) USDA does not have carte blanche authority to prohibit whatever practices it wants to stop, (3) practices are not violative where they are required by the exigencies of the business and are justified by business standards, (4) none of its contracts with hog producers required it to notify producers before implementing an equation change, (5) hog producers did not care about the equation change, (6) its failure to disclose the formula change did not impede competition or hog producers' choices, and (7) there was no evidence it acted with wrongful intent.
At the outset, it is clear that Excel's arguments do not relate to whether Excel violated § 201.99(a) of the regulations implementing the P&S Act. As discussed in greater detail below, the JO concluded that Excel violated § 201.99(a) by failing to disclose to hog producers the change in formula. In other words, contrary to Excel's arguments, the conduct at issue that violated the regulation was Excel's failure to disclose its change in formula to producers, and not the mere change in formula itself. Further, the JO's focus was on the requirements of the implementing regulation. After first concluding Excel violated that regulation, the JO in turn necessarily concluded that Excel also violated the P&S Act. Thus, the critical focus in this case is on the language of the regulation and its applicability to Excel's conduct.
In any event, it is apparent that none of the specific arguments asserted by Excel have merit. First, Excel has cited no authority, and we have found none, holding that the USDA is precluded from finding a violation in this case simply because it has not previously found a similar violation in the past. Indeed, such a rule would be nonsensical, for it would effectively preclude the USDA from applying the P&S Act and its implementing regulations to new techniques and tools utilized by slaughterers for grading livestock carcasses. Second, although the USDA does not have "carte blanche authority" to prohibit whatever practices it wants to stop, it is clear that Congress granted the USDA authority to implement and enforce the P&S Act. And, as noted, the critical issue in this case is whether Excel's failure to disclose its formula change to hog producers violated the USDA's implementing regulation. Third, and relatedly, it is clear that Congress and the USDA are the arbiters of what practices will impede competition. Thus, contrary to Excel's assertion, the fact that a particular act is "required by the exigencies of the business," or is not violative of a contractual obligation, has no impact on whether that act is violative of the P&S Act and the implementing regulations. Indeed, in the instant case, the USDA concluded that Excel's failure to disclose its formula change was violative of § 201.99(a) of the implementing regulations, even though Excel did not have a contractual obligation to disclose that change to hog producers and was otherwise justified in changing its formula to better estimate the lean percent of hog carcasses.
Fourth, Excel is incorrect when it suggests that hog producers did not care about the equation change. Indeed, the JO specifically found that some hog producers did care about the equation change, and that finding appears to be adequately supported by the record on appeal. DO at 65-66. In any event, nothing in the P&S Act or the implementing regulations provides that a violation thereof hinges on the opinions of the persons affected by the practice at issue. Although Excel cites to Ferguson v. United States Department of Agriculture, 911 F.2d 1273, 1281-82 (9th Cir. 1990), in support of its assertion that customers' opinions are critical, a review of Ferguson undercuts Excel's arguments. To begin with, Ferguson involved a different type of violation (incorrect invoicing), and thus a different provision of the P&S Act (7 U.S.C. § 213(a)), than is at issue here. Further, although the court in Ferguson did consider the testimony of customers, that testimony had no effect on the conclusion that a violation of the P&S Act had occurred; rather, the customer testimony was considered solely for purposes of determining whether the sanction imposed was proper.(2) Id. at 1282-83.
Fifth, Excel contends its actions did not impede competition or hog producers' choices. The JO, however, specifically concluded otherwise:
Hog producers can compare prices and choose to continue to sell to [Excel] or sell to [Excel's] competitors. However, [Excel] impeded that choice when it made an unannounced change in the formula. [Excel] thereby altered the price it offered hog producers without the hog producers knowing that the price structure had changed. Had hog producers been alerted to the change, they could have shopped their hogs to other packers to determine if they could obtain a better price for their hogs than [Excel's] price under its changed formula. [Excel's] failure to notify hog producers of the change in the formula to estimate lean percent impeded competition. As [GIPSA] states, the purpose of section 201.99 of the Regulations . . . "is to provide some basic level of similarity to allow sellers to evaluate different purchase offers" (Complainant's Post-Hearing Brief at 91). The assessment of harm to hog producers of the change would therefore have been whatever higher market price they might have been able to obtain from [Excel's] competitors. Therefore, I find [Excel's] violation of section 201.99(a) of the Regulations . . . grave.
DO at 57. Although Excel attempts to undercut these conclusions (e.g., by arguing that other packers did not inform hog produces about their equations to estimate lean percent), a review of the record on appeal demonstrates that they are reasonable inferences drawn from the evidence presented to the JO.
Lastly, Excel is simply wrong in asserting that, "to show an impediment to competition, GIPSA would have had to show Excel acted with wrongful intent." Aplt. Br. at 32. Nothing in the language of § 192(a) of the P&S Act or § 201.99(a) of the regulations requires a showing of wrongful intent. To the contrary, the focus is solely on the acts committed or omitted.
Excel next directly challenges the JO's interpretation of § 201.99(a). Specifically, Excel contends that, contrary to the conclusion reached by the JO, its failure to notify hog producers of the change in formula did not violate § 201.99(a). According to Excel, the "regulation does not mention: (1) lean percent; (2) equations; or (3) a change to either of them." Aplt. Br. at 34. Indeed, Excel contends that the key phrase in the regulation, i.e., "grading to be used," is ambiguous and thus it is unclear whether or not the actual formula employed by Excel in determining lean percent fell within the scope of this phrase. To support its assertion of ambiguity, Excel contends that, prior to the complaint being filed against it, the USDA never consistently or clearly interpreted § 201.99(a) in a manner that would have given Excel notice that it had to disclose to hog producers the change in formula. Excel also contends the JO failed to offer a sound explanation of the interplay between § 201.99(a) and § 201.99(e). Lastly, Excel contends that USDA has effectively sought "to rewrite the regulation in this proceeding to fit conduct that is simply not covered." Aplt. Br. at 44.
In determining whether the USDA (through the JO) committed any errors of law in interpreting § 201.99, we owe "substantial deference" to the USDA's interpretation of that regulation. Thomas Jefferson Univ. v. Shalala, 512 U.S. 504, 512 (1994). That is because the USDA has been charged by Congress with administering the P&S Act, see 7 U.S.C. § 228 (outlining the authority of the Secretary of the USDA with regard to the P&S Act), and § 201.99 is one of the regulations intended by the USDA to implement the P&S Act. See generally Mainstream Marketing Serv., Inc. v. FTC, 358 F.3d 1228, 1236 (10th Cir. 2004) (noting "that the courts owe deference to a federal agency's interpretation of a statute it administers"). Our "task is not to decide which among several competing interpretations best serves the regulatory purpose." Thomas Jefferson, 512 U.S. at 512. "Rather, the agency's interpretation must be given controlling weight unless it is plainly erroneous or inconsistent with the regulation." Id. (internal quotations omitted). "In other words," we "must defer to the Secretary's interpretation unless an alternative reading is compelled by the regulation's plain language or by other indications of the Secretary's intent at the time of the regulation's promulgation." Id. (internal quotations omitted).
The JO in this case interpreted § 201.99(a) in the following manner. First, the JO concluded that "[s]ection 201.99(a) . . . provides that each packer purchasing livestock on a carcass merit basis shall, prior to the purchase, make known to the seller the details of the purchase contract." DO at 67. Second, the JO concluded that "[t]he regulation [i.e., § 201.99(a)] explicitly provides that those details include the 'grading to be used.'" Id. Citing Merriam Webster's Collegiate Dictionary, the JO concluded that the term "grade" "[g]enerally . . . refers to quality and [the term] 'grading' is an action or process of sorting (hogs) into categories according to quality." Id. at 67 and n.28. Applying that definition to the circumstances before him, the JO concluded that "a formula to estimate lean percent is part of the grading process." Id. at 68. Thus, the JO concluded that "[t]he Fat-O-Meat'er and the formula and the change in the formula [we]re all 'grading to be used' within the meaning of" § 201.99(a). Id. at 82. In sum, the JO concluded that § 201.99(a) requires a packer such as Excel, prior to the purchase of a hog carcass, to make known to the seller the formula used in estimating the lean percent of the carcass and to make known any changes in that formula.
Excel asserts, and we agree, that the key phrase in § 201.99(a), i.e., "grading to be used," is ambiguous. In his decision and order, the JO noted the word "grading" is defined in the dictionary to mean "[t]he action or process of sorting . . . into grades according to quality." Oxford English Dictionary Online (2004). In turn, the word "grade" is defined, in pertinent part, as "[a] degree of comparative quality or value," "[a] class of things, constituted by having the same quality or value." Id. Thus, the phrase "grading to be used," as employed in § 201.99(a), clearly appears to refer to the process a particular packer will employ for sorting livestock carcasses into grades or classes according to quality. Nevertheless, the phrase is ambiguous in that it could reasonably be construed in one of at least two ways under the circumstances presented here: (1) to require Excel merely to inform hog producers that it grades carcasses according to lean percent, or (2) to require Excel not only inform hog producers that it grades carcasses according to lean percent, but also to inform hog producers that it uses a particular mathematic formula, programmed into the Fat-O-Meat'er, to estimate lean percent, and to inform hog producers when and if it implements a change in that formula.(3)
Importantly, we must "defer to both formal and informal agency interpretations of an ambiguous regulation unless those interpretations are 'plainly erroneous or inconsistent with the regulation.'" Soltane v. U.S. Dept. of Justice, 381 F.3d 143, 148 (3d Cir. 2004) (quoting Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414 (1945)); see Auer v. Robbins, 519 U.S. 452, 461 (1997) (holding that an agency's interpretation of its own regulation is entitled to deference). Here, the JO concluded that "[t]he Fat-O-Meat'er and the formula and the change in the formula [we]re all 'grading to be used' within the meaning of" § 201.99(a).(4) DO at 82. In our view, this conclusion is neither plainly erroneous nor inconsistent with the language of the regulation. Indeed, interpreting the phrase "grading to be used" to require revelation of the specific formula utilized to estimate lean percent appears to us to be entirely reasonable. Thus, we are bound to uphold the JO's interpretation.
Excel complains that the JO failed to rationally explain the interplay between §§ 201.99(a) and (e). As previously noted, § 201.99(e) provides, in pertinent part: "If settlement and final payment are based upon any grades other than official USDA grades, such other grades shall be set forth in detailed written specifications which shall be made available to the seller or his duly authorized agent." In Excel's view, the JO's interpretation of § 201.99(a) renders superfluous the language of § 201.99(e). We find it unnecessary to address Excel's arguments on this point, however, because there is no indication in the record on appeal that Excel presented these arguments to the JO. Thus, we consider the arguments waived. See United States v. L.A. Tucker Truck Lines, 344 U.S. 33, 37 (1952) ("Simple fairness . . . requires as a general rule that courts should not topple over administrative decisions unless the administrative body not only has erred but has erred against objection made at the time appropriate under its practice.").
Lastly, Excel argues that the USDA, through the JO, has effectively rewritten § 201.99(a) to encompass conduct that is otherwise not encompassed by its plain language. We disagree. As discussed above, the phrase "grading to be used," as employed in § 201.99(a), can reasonably be interpreted in at least two ways. Simply because the JO adopted one of those interpretations does not mean that the JO effectively rewrote the regulation. In other words, the JO's interpretation cannot be considered to be so far afield of the regulation's text as to "create de facto a new regulation." Christensen v. Harris County, 529 U.S. 576, 588 (2000).
Based upon his finding that Excel violated the P&S Act and the implementing regulation, the JO included the following cease and desist order in his decision and order:
Respondent, its agents and employees, directly or indirectly through any corporate or other device, in connection with its purchases of livestock on a carcass merit basis, shall cease and desist from:
(a) Failing to make known to sellers, or their duly authorized agents, prior to purchasing livestock, the factors that affect Respondent's estimation of lean percent, including, but not limited to, any change in the formula used to estimate lean percent; and
(b) Failing to make known to sellers, or their duly authorized agents, prior to purchasing livestock, the details of the purchase contract, including, when applicable, the expected date and place of slaughter, carcass price, condemnation terms, description of the carcass trim, grading to be used, accounting, and any special conditions.
ROA, Vol. V, Doc. 155 at 83. On appeal, Excel challenges the cease and desist order, arguing it (a) was imposed without fair notice, (b) should expire after no longer than three years, (c) is vague, overbroad and otherwise improper, and (d) places Excel at a competitive disadvantage. For the reasons discussed below, we reject all but Excel's assertion that the cease and desist order was overly broad.
a) Fair notice
Broadly speaking, "the requirement of notice" is "[e]ngrained in our concept of due process . . . ." Lambert v. People of State of California, 355 U.S. 225, 228 (1957). "Notice is required before property interests are disturbed" and "before penalties are assessed." Id. In short, "[n]otice is required in a myriad of situations where a penalty or forfeiture might be suffered for mere failure to act." Id. In the context of agency proceedings, an agency "may fail to give sufficient fair notice to justify a penalty if the regulation [at issue] is so ambiguous that a regulated party cannot be expected to arrive at the correct interpretation using standard tools of legal interpretation, must therefore look to the agency for guidance, and the agency failed to articulate its interpretation before imposing a penalty." United States v. Lachman, 387 F.3d 42, 57 (1st Cir. 2004).
Here, however, there is no indication in the record, and indeed no assertion by Excel, that the JO's cease and desist order infringed upon any of Excel's protected liberty or property interests. In other words, there is no basis for concluding that the JO's cease and desist order amounts to a penalty. Cf. Exxel/Atmos, Inc. v. NLRB, 28 F.3d 1243, 1248 (D.C. Cir. 1994) ("Cease and desist orders are remedial; they require only that the employer 'conform his conduct to the norms set forth in the Act.'"); Carpenter Sprinkler Corp. v. NLRB, 605 F.2d 60, 67 (2d Cir. 1979) (noting that cease and desist order was "clearly remedial" rather than punitive); Benrus Watch Co. v. FTC, 352 F.2d 313, 322 (8th Cir. 1965) ("Cease and desist orders are not punitive . . . ."). Thus, we reject Excel's "fair notice" arguments.
b) Duration of cease and desist order
Excel argues that the cease and desist order, however it is written, should expire after no longer than three years pursuant to 28 U.S.C. § 530D(a)(1)(C)(ii).(5) The JO addressed this precise argument in his order rejecting Excel's petition for reconsideration. We agree with the JO that the statute cited by Excel does not apply here because the parties did not settle or compromise this proceeding. Rather, the record makes clear that the proceeding was resolved by the JO only after the parties fully litigated the issues.
c) Vague and overbroad
Excel argues that the cease and desist order is unduly vague and overbroad. In particular, Excel notes that the cease and desist order covers its purchase of all "livestock," rather than just hogs, and requires disclosure of all "factors that affects [its] estimation of lean percent, including, but not limited to, any change in the formula used to estimate lean percent." Aplt. Br. at 52. According to Excel, this language goes beyond the violation found by the JO and beyond the requirements of § 201.99(a) as interpreted by the JO. Thus, Excel argues, there is "no way [it] can possibly know what is required" by the cease and desist order. Id. at 53.
Generally speaking, we must uphold an agency's cease and desist order so long as "the remedy selected" bears a "reasonable relation to the unlawful practices found to exist." FTC v. Colgate-Palmolive Co., 380 U.S. 374, 394-95 (1965); see generally NLRB v. Express Publ'g Co., 312 U.S. 426, 435 (1941) (noting that a federal court may "restrain acts which are of the same type or class as unlawful acts which the court has found to have been committed or whose commission in the future, unless enjoined, may fairly be anticipated from the defendant's conduct in the past."). We may, however, "narrow [an agency's] orders . . . by deleting those portions for which a reasonable relationship to the offending conduct is lacking." ITT Continental Baking Co. v. FTC, 532 F.2d 207, 220-21 (2d Cir. 1976) (modifying cease and desist order issued by Federal Trade Commission); see Encyclopaedia Brittanica, Inc. v. FTC, 605 F.2d 964, 970 (7th Cir. 1979) (same).
Here, we agree with Excel that portions of the cease and desist order fail to bear a reasonable relationship to the conduct which the JO found had violated the regulation and statute at issue. As noted, the primary violative conduct identified by the JO was Excel's failure, in connection with its purchase of hogs, to disclose to sellers the change in the formula used to estimate lean percent. The cease and desist order, however, unreasonably exceeds the scope of this violation in three respects. First, the cease and desist order broadly refers to "purchases of livestock," even though it is uncontroverted that Excel's violation was limited to the purchase of hogs. Second, the cease and desist order prohibits Excel from "[f]ailing to make known" not only "any change in the formula used to estimate lean percent," but virtually all "the factors that affect [its] estimation of lean percent . . . ." Third, the cease and desist order broadly prohibits Excel from "[f]ailing to make known to sellers, or their duly authorized agents, prior to purchasing livestock, the details of the purchase contract, including, when applicable, the expected date and place of slaughter, carcass price, condemnation terms, description of the carcass trim, grading to be used, accounting, and any special conditions." Although this language generally tracks the requirements of 9 C.F.R. § 201.99(a), there is simply no evidence in this case that Excel failed to comply with those requirements, other than with respect to the formula used in the Fat-O-Meat'er for estimating lean percent and the change in that formula. In sum, we conclude the burdens imposed on Excel by these three aspects of the JO's cease and desist order are not justified by the violation the JO found.
To narrow the cease and desist order to reflect and address the violation found by the JO, (1) the reference to "livestock" in the opening sentence of the order is changed to "hogs," (2) the language of paragraph (a) is changed to refer solely to "any change in the formula used to estimate lean percent," and (3) paragraph (b) is deleted entirely. As modified, the cease and desist order will now read as follows:
Respondent, its agents and employees, directly or indirectly through any corporate or other device, in connection with its purchases of hogs on a carcass merit basis, shall cease and desist from failing to make known to sellers, or their duly authorized agents, prior to purchasing livestock, any change in the formula used to estimate lean percent.
d) Competitive disadvantage
Finally, Excel argues that the cease and desist order places it at a competitive disadvantage because a violation of the order will subject it and its employees, but not its competitors, to criminal prosecution. Aplt. Br. at 46. Excel further argues that this "threat of criminal sanctions could lead to Excel employees leaving Excel to work for packers who are not subject to such penalties." Id. Ultimately, Excel argues, these factors could "impact [its] decision to stay in the pork business." Id. at 47.
Having modified the cease and desist order to tailor it to the specific violation found by the JO, we conclude there is no merit to Excel's arguments. Simply put, the requirements imposed by the modified cease and desist order are narrow and clear. Moreover, by reason of the USDA's action against Excel, Excel's competitors are on notice that they are also subject to the same regulatory requirements. Thus, we fail to see how compliance with the modified cease and desist order could reasonably place Excel at a competitive disadvantage.
The petition for review is GRANTED for the sole purpose of modifying the Judicial Officer's decision and order in accordance with this opinion. As so modified, the decision and order is enforced.
1. When Excel responded that it had refunded to producers $3,093,581.00 (including 5.85% interest) as the difference between the price it paid under the Purdue formula and the Danish formula, the complaint was amended to allege an underpayment to producers of $635,345.52. Id. at 26.
2. In a related point, Excel complains that the Chief ALJ precluded Excel from calling six producer witnesses (the Chief ALJ apparently ruled that only four of Excel's producer witnesses could testify, and that the remaining six would merely provide cumulative testimony). This is clearly a red herring that has no impact on the propriety of the JO's decision.
3. The phrase "grading to be used" could also arguably be interpreted to require Excel to either (a) reveal only that it uses a mathematic formula programmed into the Fat-O-Meat'er for purposes of estimating lean percent, or (b) reveal the precise details of that mathematic formula, as well as all the details of its matrix.
4. The uncontroverted facts of this case readily establish that, because the USDA had no official grades in place for hog carcasses, Excel adopted and used its own grading system for hog carcasses which focused primarily on lean percent. The uncontroverted facts further establish that Excel's calculation of lean percent was based on a mathematic formula programmed into the Fat-O-Meat'er. More specifically, the uncontroverted facts indicate that Excel physically employed the Fat-O-Meat'er and its embedded mathematic formula to estimate the lean percent of each hog carcass, and that the lean percent estimate, along with the carcass's overall weight, effectively resulted in a grade on Excel's matrix.
5. The statute cited by Excel, entitled "Report on enforcement of laws," provides in pertinent part as follows:
(1) In general.The Attorney General shall submit to the Congress a report of any instance in which the Attorney General or any officer of the Department of Justice
* * *
(C) approves . . . the settlement or compromise . . . of any claim, suit, or other action
* * *
(ii) by the United States (including any agency or instrumentality thereof) pursuant to an agreement, consent decree, or order . . . that provides injunctive or other nonmonetary relief that exceeds, or is likely to exceed, 3 years in duration . . . .
28 U.S.C. § 530D(a)(1)(C)(ii).