Deaf Smith County Grain Processors, Inc. v. Glickman

162 F.3d 1206, 333 U.S. App. D.C. 299, 1998 U.S. App. LEXIS 32517, 1998 WL 898358

• Court: Court of Appeals for the D.C. Circuit
• Decided: December 29, 1998
• Docket: 98-5021
• Status: Published

Opinion

Opinion for the Court filed by Chief Judge HARRY T. EDWARDS.

HARRY T. EDWARDS, Chief Judge:

In this appeal, Deaf Smith County Grain Processors, Inc. (“Deaf Smith” or “appellant”) contests the District Court’s grant of summary judgment in favor of the United States Department of Agriculture (“USDA” or “appellee”) on Deaf Smith’s claim that it has not received the farm subsidy and disaster relief payments to which it is allegedly entitled under USDA regulations. Our review in this ease is identical to that of the District Court; that is, we uphold the agency action so long as it was not arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law. We agree with the District Court that the USDA acted reasonably in denying Deaf Smith the payments it requests. Accordingly, we affirm the grant of summary judgment in favor of the agency.

I. BACKGROUND

A. Regulatory Background

The USDA administers the nation’s major agricultural commodity programs through a wholly-owned Government corporation called the Commodity Credit Corporation (“CCC”). See 15 U.S.C. § 714c (1994). The day-to-day operations of the CCC are performed by local chapters of the Farm Service Agency (“FSA”), formerly known as the Agricultural Stabilization and Conservation Service (“ASCS”). Two CCC programs are at issue in this case: the Production Adjustment Program (“PAP”) and the Disaster Assistance Program (“DAP”).

The goal of the Production Adjustment Program is to reduce the total nationwide acreage of farmland devoted to the production of certain crops. To achieve this goal, the CCC pays farmers to limit their production of those crops. In order to participate in the PAP, and thereby receive payments from the CCC, farmers enter into annual contracts with the CCC, pursuant to which they agree to limit their cropland acreage and devote a certain amount of their land to conservation uses approved by the USDA. See 7 C.F.R. § 1421.5 (1998). In exchange, the CCC provides price support in the form of “deficiency” payments and nonrecourse loans. The amount of these payments is determined by the amount of farmland that the farmer agrees to withhold from crop harvest. A farm’s eligible acreage for farming a particular crop is called the crop acreage base (“CAB”). Each year, the USDA determines a percentage of the CAB on which farmers participating in the PAP are restricted from growing crops, and calculates its deficiency payments accordingly. As a result, the higher a particular farmer’s CAB, the greater the payments are from the CCC.

The Disaster Assistance Program provides relief for farmers whose crops have been destroyed by natural disasters. When struck by a natural disaster, a farmer may file an application for disaster credit with the CCC. The USDA, through the local branch of the FSA, then establishes benchmark crop “yields” and “rates” in order to calculate a particular farmer’s disaster payments. See 7 C.F.R. § 1477.5 (1988). The purpose of the “yields” and “rates” is to approximate the value of the crops lost by estimating what the farmer’s crop would have been absent the disaster. A percentage of the estimated value of the lost crops is then paid to the farmer in the form of disaster relief.

B. Factual Background

Appellant is a large farming enterprise that does business in the southwestern United States. During the 1980s, appellant participated in both the PAP and the DAP, in connection with its farms in Colfax County and Union County, New Mexico. Appellant’s two claims are essentially unrelated to each other.

1. Appellant’s PAP Claim

Appellant bought the land at issue in this ease in 1986. In 1982, however, the Colfax County ASCS erroneously designated 3005.1 acres of this land “non-cropland,” thus significantly diminishing the land’s CAB for the purposes of calculating payments under the PAP. The parties now agree that this 3005.1 acres was cropland, and was therefore potentially “eligible” for consideration in calculating the land’s total CAB. Appellant concedes that when it bought the land, it knew that the CCC’s “data on Appellant’s cropland acreage was much less and inconsistent with the cropland actually on the farm.” Brief for Appellant at 5.

When appellant bought the land in 1986, and every year thereafter through 1989, it entered into a PAP contract with the CCC. Each time it entered into a new annual contract, the CCC provided notice of the CAB calculated for the land, and informed appellant of its right to appeal this calculation within fifteen days of receipt of the notice. It is undisputed that appellant never formally appealed the CAB figure assigned to its land, even though it knew that the figure was inaccurate. Appellant did complain orally to local ASCS officials about the erroneous CAB. Appellant claims, however, that it was discouraged from filing formal appeals, because the local officials represented that the documents needed to clarify the mistake had been inadvertently destroyed. See Brief for Appellant at 5-8. Appellant eventually filed a Freedom of Information Act (“FOIA”) request for the relevant documents. The documents had not been destroyed, and they vindicated appellant’s assertion that the 3005.1 acres was, indeed, cropland.

Appellant argues that it is entitled to retroactive payments for the years 1986 through 1989 to reflect the payments that should have been made had the proper CAB been used during those years. Appellee responds that appellant was aware of its actual cropland acreage when it entered into the contracts with the CCC each year, and that both parties fully performed in accordance with those contracts. Appellee contends that these contracts “cannot be performed retroactively,” and that the court should not construct “implied in law” contracts to reflect hypothetical contracts that the parties might have entered into had the proper CAB been used. Brief for Appellee at 24. Furthermore, appellee points out that appellant has not established that the disputed acres were in fact devoted to conservation during 1986 through 1989; since the Colfax County ASCS had designated the land “non-cropland,” those 3005.1 acres were theoretically “freed up ... for other uses.” Id. at 21.

2. Appellant’s DAP Claim

Appellant farmed wheat grass in New Mexico from 1986 to 1989. When wheat grass farmers in that region experienced severe droughts in 1988 and 1989, Congress activated the DAP to provide relief for those farmers. The USDA established wheat grass yields and payment rates to calculate the disaster payments for wheat grass farmers in New Mexico. Appellant applied for disaster relief and was awarded payments based on those yields and rates.

Appellant contends, however, that the selected yields and rates were too low, precluding the CCC from realistically approximating the value of the crops destroyed by the droughts. Specifically, appellant challenges the yield selected because the USDA relied upon data that was admittedly seventeen to thirty-one years old, see Appendix for Appellant (“A.A.”) 73, and failed to use data from Colorado farms that arguably would have better reflected appellant’s crop growth po tential, see id. at 39. Appellant also challenges the payment rate selected, primarily on the basis of affidavits from other farming enterprises, which suggest that the USDA’s rate was lower than rates that were used elsewhere. See, e.g., id. at 79.

Appellee responds that the use of the disputed data was necessary and reasonable given the fact that wheat grass “was not a common crop in New Mexico” and no better data was readily available. Brief for Appel-lee at 25. Indeed, appellant points to no other data for New Mexico. Furthermore, according to appellee, it was proper for the USDA to use the New Mexico data, as opposed to data collected in Colorado, because appellant’s farm is, in fact, in New Mexico. As for the rate selected, appellee maintains that as long as the rate selected is reasonable, appellant “may not complain merely because there are other rates which also may be reasonable.” Id. at 31.

C. Procedural History

1.The Office of Inspector General Report

At the USDA’s request, the USDA Office of Inspector General (“OIG”) reviewed appellant’s PAP and DAP complaints. The OIG issued its report on May 3, 1990. See A.A. 37-42. With respect to appellant’s PAP claim, the OIG concluded that the Colfax County ASCS had indeed “understated” appellant’s cropland, resulting in an “inadvertent[]” miscalculation of appellant's CAB. Id. at 41. The OIG suggested that, in response to this error, “[a]pplicable program payments should ... be corrected as necessary.” Id. at 42. With respect to appellant’s DAP claim, the OIG “recognize[d] the Disaster Program was implemented under tight time constraints and difficult conditions,” but nevertheless “question[ed] whether it was proper to use 17- to 31-year-old data.” Id. at 40. The OIG also noted that the “Colorado yields may more closely reflect [appellant’s] growth potential.” Id. As for the rate selected, the OIG found “no basis for questioning the [USDA’s] payment rate since it is similar to Colorado’s ■... rate and was based on rates obtained from New Mexico seed companies.” Id.

Despite the statements favorable to appellant in the OIG Report, the CCC refused to make the requested payments. See Brief for Appellant at 13.

2. The National Appeals Division Decision

Appellant then sought relief through the USDA’s administrative procedures, which culminated in a March 1995 decision of the National Appeals Division (“NAD”) of the USDA. See Determination of the National Appeals Division, Appeal of Clifford Skiles, Jr. (Mar. 13, 1995) (“NAD Decision”), reprinted in A.A. 5.

The NAD noted the existence of, and considered, the OIG Report, see NAD Decision at 1, reprinted in A.A. 5, but found in favor of appellee on both the PAP and the DAP claims. With respect to the PAP claim, the NAD held that the “contracts in question cannot be revised and program payments cannot be issued that exceed what was actually earned under the contracts signed by Appellant and approved by [the] CCC.” Id. at 7, reprinted in A.A. 11. With respect to the DAP claim, the NAD agreed that the data used in calculating the yield was “somewhat aged”; however, the NAD found that the data used was “the best data available at that time and reflects the cropping conditions of the geographical location where the crop was grown.” Id. at 8, reprinted in A.A. 12. Finally, as for the payment rate, the NAD found no “conclusive evidence that the rate was improperly established or that it did not reflect actual market conditions at that time.” Id.

3. The District Court Opinion

Appellant appealed the NAD decision to the District Court. Appellant’s complaint in the District Court requested a total of approximately $95,000 in damages. See Complaint ¶ 25. In a memorandum decision, the court granted appellee’s motion for summary judgment on both claims. See Deaf Smith County Grain Processors, Inc. v. Glickman, Civ. Action No. 95-1214 (D.D.C. Dee. 29, 1997) (“Order”), reprinted in A.A. 1.

With respect to the PAP claim, the court noted that appellant had been aware of the “trae acreage of its farmland” when it purchased the land and when it entered into the contracts each year with the CCC. Order at 2, reprinted in A.A. 2. Furthermore, the court held, appellant “would not be able to certify that [from 1986 to 1989] it planted or set aside additional acres and/or otherwise complied with the additional contract requirements.” Id. at 2-3, reprinted in A.A. 2-3. With respect to the DAP claim, the court held that the “administrative record in this case clearly demonstrates that defendant’s actions in establishing the yields in question were reasonable and that defendant had a rational basis in its selection of the yields.” Id. at 2, reprinted in A.A. 2. Likewise, the court found no unreasonable conduct in the selection of the payment rate. See id.

II. Analysis

A. Jurisdiction

It has long been settled that the federal Government may be sued in federal court only if Congress has waived sovereign immunity for the lawsuit. See Cohens v. Virginia, 19 U.S. (6 Wheat.) 264, 411-12, 5 L.Ed. 257 (1821); First Va. Bank v. Randolph, 110 F.3d 75, 77 (D.C.Cir.1997). The District Court’s jurisdiction over this case is founded on Congress’s waiver of sovereign immunity in 7 U.S.C. § 6999, which states, in its entirety:

A final determination of the [National Appeals] Division shall be reviewable and enforceable by any United States district court of competent jurisdiction in accordance with chapter 7 of Title 5.

Were it not for § 6999’s specific reference to the judicial review provisions of the Administrative Procedure Act (“APA”), 5 U.S.C. §§ 701-706, there would be no doubt that § 6999 is an effective waiver of the Government’s sovereign immunity in cases involving challenges to a decision of the NAD, and that the District Court properly exercised jurisdiction over this case. See Randolph, 110 F.3d at 77-78 (recognizing district court jurisdiction where there is an express waiver of sovereign immunity coupled with an express grant of jurisdiction). However, because § 6999 states that final determinations of the NAD shall be reviewed in the district court “in accordance with” the APA, there is some question whether the limits on the APA’s waiver of sovereign immunity apply to this case. This is significant, because the APA exempts from its waiver of sovereign immunity claims seeking “money damages” from the Government, see 5 U.S.C. § 702, and claims for which an “adequate remedy” is available elsewhere, see id. § 704. See Transohio Sav. Bank v. Director, Office of Thrift Supervision, 967 F.2d 598, 607 (D.C.Cir.1992).

It is not clear whether this particular suit for monetary relief from the federal Government can be characterized as a claim for “money damages” within the meaning of § 702 of the APA. Compare Kidwell v. Department of the Army, 56 F.3d 279, 284 (D.C.Cir.1995) (noting that “a claim is subject to the Tucker Act and its jurisdictional consequences if, in whole or in part, it explicitly or ‘in essence’ seeks more than $10,000 in monetary relief from the federal government” and explaining that “ ‘^jurisdiction under the Tucker Act cannot be avoided by ... disguising a money claim’ as a claim requesting a form of equitable relief’) (citation omitted), with Esch v. Yeutter, 876 F.2d 976, 984 (D.C.Cir.1989) (holding that challenge to USDA’s decision to suspend farmers from participation in subsidy programs was not an action for “money damages” under Bowen v. Massachusetts, 487 U.S. 879, 108 S.Ct. 2722, 101 L.Ed.2d 749 (1988), and was therefore not subject to Tucker Act jurisdiction). If § 6999 is read to incorporate the APA’s “money damages” exception to sovereign immunity, and if appellant’s claim is one for “money damages,” the District Court is not a court of “competent” jurisdiction, because sovereign immunity has not been waived. As a result, the District Court would not have jurisdiction over this suit.

Moreover, if § 6999 is read to incorporate the judicial review provisions of the APA, then it would appear that the Tucker Act— which waives sovereign immunity and provides the United States Court of Federal Claims (“Claims Court”) with jurisdiction over certain claims for monetary relief from the federal Government, see 28 U.S.C. § 1491 — provides an “adequate remedy” in the Claims Court, which would preclude district court jurisdiction under § 704 of the APA. This, too, would militate against the District Court asserting jurisdiction over this lawsuit.

We must therefore determine whether the “in accordance with” language of § 6999 suggests that Congress intended the district court to have jurisdiction over appeals from the NAD only to the extent that such appeals would be permitted under the APA. If so, jurisdiction over this case may properly lie in the Claims Court, depending upon the resolution of the complicated APA and Tucker Act issues just described. We need not decide these APA and Tucker Act issues, however, because we are confident that — as both parties to this lawsuit contend — Congress intended § 6999’s reference to the APA to govern only the procedural and scope of review aspects of any appeal from the NAD. It did not intend for the APA to govern the jurisdiction of the district court over these kinds of claims. Accordingly, we hold that § 6999 provides the district court with jurisdiction over all final determinations of the NAD.

In reaching this conclusion, we have little to guide us but our common sense. The parties claim that the “plain language” of § 6999 is controlling. See, e.g., Appellee’s Response to Court’s Order at 6. We disagree. The language of § 6999, standing alone, is subject to several plausible interpretations. The judicial review provisions of the APA referenced by § 6999 contain both scope of revieiv limitations, see, e.g., 5 U.S.C. § 706(2)(A) (establishing the “arbitrary and capricious” standard of review), and substantive limitations upon judicial review of agency action, see, e.g., id. §§ 702, 704 (establishing the “money damages” and “adequate remedy” exceptions to the APA’s waiver of sovereign immunity). To say, as § 6999 does, that final determinations of the NAD are reviewable by the district court “in accordance with” the judicial review provisions of the APA is to say one of three things: (i) that the NAD determinations are reviewable to the extent allowed by §§ 702 and 704 of the APA; (ii) that NAD determinations are reviewable under the procedures set forth in § 706 of the APA; or (iii) that NAD determinations are reviewable pursuant to all of the judicial review provisions of the APA. If anything, the language supports the third option, given that § 6999’s reference to “chapter 7 of Title 5” is unbounded.

The parties both assert that to adopt this third interpretation would be to render § 6999 meaningless. Appellant’s theory is that if § 6999 provides jurisdiction in the district court only for claims that would be reviewable under the APA, the district court would be effectively, divested of jurisdiction over any NAD appeals, because all such appeals would fall within the exceptions to the APA’s grant of sovereign immunity. According to appellant, “all appeals from NAD determinations could be held ‘in essence’ to involve claims of entitlement to monetary relief from the federal government ... and thus implicate Tucker Act jurisdiction.” Appellant’s Supplemental Brief at 6. Appellee disagrees with that contention. See Appel-lee’s Response to Court’s Order at 9 (asserting that appellant’s DAP claim “would not meet the narrow view of what is or is not ‘money damages’ ”). Nevertheless, appellee agrees with appellant that an interpretation that subjects § 6999 to the APA’s limits on sovereign immunity would violate the “basic canon of statutory construction” that “a proper interpretation of [a] legislative provision must give all of the words of the provision meaning.” Id. at 5.

Appellee’s reasoning does not make sense in this context. If it is true that some appeals from the NAD would not involve claims for “money damages” and also would not be subject to the Tucker Act — and we suspect that this is the case, cf. Esch, 876 F.2d at 984-85 (holding that lawsuit filed prior to the creation of the NAD was neither for “money damages” nor subject to the Tucker Act)— then § 6999 may plausibly be read to provide jurisdiction in district court only for those appeals from the NAD that do not run afoul of the APA’s “money damages” and “adequate remedy” exceptions. Such a reading would not render § 6999 meaningless. Nev ertheless, we agree that the statute is ambiguous. Therefore, because the language of § 6999 provides no solution to the jurisdictional problem in this case, we turn to the history of § 6999 — legislative and otherwise — in an effort to ascertain Congress’s intent.

Unfortunately, the little legislative history that exists for § 6999 is as ambiguous as the statute itself. Section 209 of the Senate version of the Department of Agriculture Reorganization Act of 1994 contained a provision that was virtually identical to the provision that eventually became § 6999. See S.1970, 103d Cong. § 209 (1994). The committee report accompanying the bill contained the following discussion of § 209:

This section provides that a final determination of the [National Appeals] Division can be appealed to a U.S. District Court. Analysis of which issues are subject to judicial review shall conform with provisions of the Administrative Procedure Act (APA).

S.Rep. No. 103-241, at 15 (1994). The parties make much of this language. They claim that it unambiguously supports the proposition that Congress “referred to the APA in the judicial review provision not to specify the location of the review but rather to specify the methodology of the review to be conducted.” Appellee’s Response to Court’s Order at 7; see also Appellant’s Supplemental Brief at 5.

Although the first sentence of the quoted paragraph is quite straightforward, we do not agree that the meaning of the second sentence is readily apparent. It is certainly far from clear that an “[a]nalysis of which issues are subject to judicial review” would involve only the APA’s procedural and scope of review provisions, such as § 706, and not, say, § 704, which expressly describes the agency actions that are “subject to judicial review.” Because the first sentence contains no limitation on the right of appeal to the district court, we are of the view that the committee’s statement, read as a whole, is ambiguous. However, we do not accept the parties’ contention that it positively supports their position.

Armed only with this single sentence of legislative history, the parties urge us to consider as well “the clearest indication of legislative intent,” which is that the pre- § 6999 state of affairs was marked by numerous “thorny and frustrating” jurisdictional disputes, Appellee’s Response to Court’s Order at 7-8, and that Congress enacted § 6999 in order to remove the confusion surrounding this issue. Appellee contends that “[i]t does not make sense that Congress intended, in a statute that was designed to improve the conditions involved in the review of farm program decisions, to leave jurisdictional gridlock in place.” Appellee’s Response to Court’s Order at 9-10. Although the lack of legislative history to support the proposition is somewhat troubling, we do agree that, given the history of the NAD, Congress most likely intended § 6999 to vest jurisdiction over appeals from the NAD in one, and only one, forum — the district court.

Prior to 1994, claims such as appellant’s were heard in an administrative tribunal known as the National Appeals Division, which is not to be confused with the present-day NAD. See 7 .U.S.C. § 1433e(c) (Supp. Ill 1991) (repealed 1994). Final decisions of that National Appeals Division' — generally referred to as the “old NAD” — were, by statute, “reviewable by a United States court of competent jurisdiction.” Id. § 1433e(d). As a practical matter, § 1433e(d) gave litigants (usually farmers seeking benefits under particular USDA programs) the choice of filing their claims in either the Claims Court or in district court. See Christopher R. Kelley & John S. Harbison, A Guide to the ASCS Administrative Appeal Process and to the Judicial Review of ASCS Decisions, 36 S.D. L.Rev. 435, 439 (1991). Apparently, due to the perception that review in district court was generally more favorable to farmers than review in the Claims Court, farmers often attempted to draft their complaints to avoid the Tucker Act. See Alexander J. Pires, Jr., Why the U.S. Claims Court is Not a Viable Venue for Farmers: The U.S. Claims Court’s Handling of Agricultural Cases, 1980-1990, 15 U. Aek. Little Rock L.J. 223, 259 (1993) (suggesting techniques for avoiding Claims Court jurisdiction, including such advice as “do not plead a breach of contract count” and “plead for equitable relief’). The USDA, for its part, frequently challenged the district court’s jurisdiction on the ground that the farmers’ claims were actually for “money damages.” See, e.g., Justice v. Lyng, 716 F.Supp. 1567, 1568-69 (D.Ariz.1988); Alan R. Malasky et al., Resolving Federal Farm, Program Disputes: Recent Developments, 19 Wm. Mitchell L.Rev. 283, 327 & nn.361-62 (1993). It appears that courts could not agree on the appropriate venue for the farmers’ claims; some were heard in the Claims Court, see, e.g., Wardlaw Farms, Inc. v. United States, 32 Fed. Cl. 475 (1994), while others were heard in the district court, see, e.g., Justice. In short, there was “considerable confusion” over this jurisdictional question. Kelley & Harbison, 36 S.D. L.Rev. at 440; see also Malasky et al., 19 Wm. Mitchell L.Rev. at 326 (referring to this jurisdictional question as a “recurring problem[]” in USDA litigation).

Against this backdrop, Congress created the “new” NAD in 1994, and concurrently enacted the judicial review provision at issue in this case. Significantly, § 6999 differs from the judicial review provision of the old NAD, in that it vests authority to review NAD decisions in “any United States district court of competent jurisdiction” (emphasis added). By specifying that jurisdiction would lie in the district court, Congress was arguably signaling its intent to remove the specter of the Tucker Act — and the Claims Court — from the judicial review of NAD decisions. Commentators appear to have assumed that § 6999 “finally puts to rest the Government’s ... argument that all actions seeking judicial review of USDA determinations denying farm program benefits are really actions for ‘money damages’ which must be brought in the [Claims Court] under the Tucker Act.” Alan R. Malasky & William E. Penn, USDA Reorganization — Fact or Fiction?, 25 U. Mem. L.Rev. 1161, 1183 (1995). Moreover, although certainly not dispositive, we find quite relevant the fact that it is now “the position of the United States that 7 U.S.C. § 6999 provides district courts with jurisdiction to conduct judicial review of all determinations made by the NAD.” Appel-lee’s Response to Court’s Order at 2. After all, it was the United States that routinely challenged the district courts’ jurisdiction over these claims prior to the creation of the new NAD, so the Government’s present position is at least noteworthy.

We are convinced that a contrary reading of § 6999 — one founded on the view that the district court had jurisdiction over only those appeals from the NAD that were not claims for “money damages” — would result in perhaps even greater “jurisdictional gridlock” than existed prior to the passage of § 6999. For example, if we determined that appellant’s PAP claim was one for “money damages,” but its DAP claim was not, jurisdiction over the NAD decision at issue in this case could potentially lie in both the District Court and the Claims Court. We agree with the parties that it is highly unlikely that, given the jurisdictional confusion and uncertainty that reigned prior to the creation of the new NAD, Congress intended this result. In the end, although the evidence is not overwhelming and the language of the statute is far from unambiguous, it appears that the purpose of § 6999 was to simplify appeals from the NAD by placing jurisdiction over them solely in the district court. Accordingly, the District Court properly exercised jurisdiction over this case, and we may now address the merits.

B. Standard of Revieio

As we have just articulated, § 6999 mandates that the District Court review, as it has, the final determination of the NAD under the APA’s “arbitrary and capricious” standard of review. On appeal from the District Court, we review the NAD’s decision de novo, see Dr. Pepper/Severv-Up Cos. v. FTC, 991 F.2d 859, 862 (D.C.Cir.1993), and will uphold it unless we find it to be “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A); see also Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 41-44, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983).

C. The PAP Claim

Appellant contends that it is entitled to retroactive payments under the PAP, in or der to compensate it for the erroneous designation of 3005.1 acres of its land as “noncrop-land.” Appellee admits that the Colfax County ASCS incorrectly reduced the amount of cropland it attributed to appellant. See Brief for Appellee at 20-21. However, it is undisputed that appellant nevertheless knowingly entered into contracts with the CCC each year to receive PAP payments based on the erroneous amount of cropland that was attributed to the farm. It is also undisputed that the payments made under the contracts were correct pursuant to the terms of those contracts. Although appellant would have been entitled to additional payments had the contracts accurately reflected the “eligible” cropland, the fact remains that the contracts did not so reflect that cropland. And, significantly, appellant never formally appealed the CABs that had been attributed to its farm, despite adequate notice of its ability to do so. See NAD Decision at 7, reprinted in A.A. 11 (“The administrative record contains no evidence that Appellant timely submitted an .appeal concerning the CABs established for the farm during the specified time period for appeals to be filed.”).

By neglecting to formally appeal the CABs, appellant failed to exhaust its administrative remedies. Its action, at least with respect to this claim, is therefore barred. See 7 U.S.C. § 6912(e) (1994) (“[A] person shall exhaust all administrative appeal procedures established by the Secretary [of Agriculture] or required by law before the person may bring an action in a court of competent jurisdiction against ... the Secretary”). Appellant protests that it “orally contested the accuracy of [the CABs] ... and that Appellee discouraged any formal appeal by erroneously telling Appellant that the history records needed to appeal the cropland acreage and CAB[s] had probably been destroyed.” Brief for Appellant at 19. Appellant’s claim' — that it failed to avail itself of the procedures available to contest the assigned CAB because it was induced not to by low-level USDA officials — is, at bottom, an equitable estoppel claim against the Government. Under clear precedent of the Supreme Court, we are constrained to reject such a claim.

In Office of Personnel Management v. Richmond, 496 U.S. 414, 415-16, 110 S.Ct. 2465, 110 L.Ed.2d 387 (1990), the Court emphatically dismissed the notion that “erroneous oral and written advice given by a Government employee to a benefits claimant” would “give rise to estoppel against the Government and so entitle the claimant to a monetary payment not otherwise permitted by law.” The Court noted that the “whole history and practice with respect to claims against the United States reveals the impossibility of an estoppel claim for money in violation of a statute,” id. at 430, 110 S.Ct. 2465, and went on to say that “[t]o open the door to estoppel claims would only invite endless litigation over both real and imagined claims of misinformation by disgruntled citizens, imposing an unpredictable drain on the public fisc,” id. at 433, 110 S.Ct. 2465. Appellant in this case seeks to avoid the exhaustion requirement on the ground that USDA officials erroneously advised him of the futility of pursuing his administrative remedies. His claim, although not labeled as such, falls squarely within the category of equitable es-toppel claims against the Government that the Court precluded in Richmond.

In any event, even if appellant’s claim were not barred for failure to exhaust, we would be unable to grant appellant the relief it seeks. As the District Court noted, the record does not establish what appellant did with the land that was erroneously designated “non-cropland.” See Order at 2-3, reprinted in A.A. 2-3. Thus, even if we could reconstruct the contracts between appellant and the CCC — a question that we need not address — appellant would still have the burden of establishing not only that it did not farm the disputed acres during the relevant period, but also that it “complied with the additional contract requirements,” such as setting aside a portion of the land for certain USDA-approved conservation uses. See id. at 3, reprinted in A.A. 3. Appellant has not met its burden of proof on these points.

In sum, the NAD’s decision to deny appellant’s appeal on the PAP claim was reasonable. The NAD was not compelled to follow the recommendations of the OIG Report. The question at issue here is whether the agency’s decision making was arbitrary and capricious, and we hold that it was not.

D. The DAP Claim

With respect to appellant’s claim under the DAP, we agree with the District Court that there is nothing in the record to suggest that the agency acted arbitrarily or capriciously when it set the yields and rates for wheat grass payments. See Order at 1-2, reprinted in A.A. 1-2.

The USDA arguably could have selected yields and rates that were more favorable to appellant. It may even be true that the USDA failed to use the most accurate data that it could have found in setting the yield. For example, according to the OIG Report, USDA officials were told when they received the New Mexico data that data from Colorado might “better reflect[]” the actual crop growth in the area surrounding appellant’s farm. A.A. 39. However, given the “tight time constraints and difficult conditions” under which the local USDA officials were operating, id. at 40, it was not unreasonable for the officials to rely on the New Mexico data, given that appellant’s farm was located in that state. See North Carolina v. FERC, 112 F.3d 1175, 1190 (D.C.Cir.1997) (stating that although estimates agency used to select growth rates might have been less reasonable than other available data, “the fact that these estimates were less ‘reasonable’ does not necessarily make them unreasonable or arbitrary”); National Wildlife Fed’n v. Burford, 871 F.2d 849, 855 (9th Cir.1989) (noting that under the arbitrary and capricious standard, “[t]he action ... need be only a reasonable, not the best or most reasonable, decision”). Furthermore, we are unconvinced that the agency’s use of this data was unreasonable merely because it was, in the words of the NAD. “somewhat aged.” NAD Decision at 8, reprinted in A.A. 12. Appellant’s talismanic repetition of the data’s age, see, e.g., Brief for Appellant at 21, does nothing to illuminate why data of this type does not maintain its accuracy over time.

As for the USDA’s payment rate, even the OIG Report found “no basis for questioning” it, because it was so similar to other rates in the area, and was based on rates obtained from New Mexico seed companies. A.A. 40. It is not surprising that appellant can point to other rates in existence at the time that were higher than the rate chosen by the USDA. See, e.g., Complaint ¶ 24. These comparative data cannot carry the day, for the issue here is whether the rate chosen was reasonable; and there is simply nothing in the record to suggest that the rate chosen was anything but reasonable. Appellant does not even claim that the method by which the USDA calculated its rate was, in any way, flawed. In fact, appellant concedes that the USDA used the “methodology set forth” in the USDA’s own handbook. Brief for Appellant at 10.

In short, appellant has failed to establish that the agency’s action with respect to its DAP claim was unreasonable. Accordingly, we uphold the determination of the NAD with respect to this claim as well.

III. Conclusion

For the reasons stated above, we affirm the judgment of the District Court.

So ordered.