Com. of Mass. v. US

• Decision Date: 06 April 1992
• Docket Number: Civ. A. No. 86-2132-Y.
• Citation: 788 F. Supp. 1267
• Parties: COMMONWEALTH OF MASSACHUSETTS, and Massachusetts Department of Public Welfare, Plaintiffs, v. UNITED STATES of America, United States Secretary of Agriculture, Massachusetts State Board of Food Appeals, and United States Department of Food & Nutrition Service, Defendants.
• Court: U.S. District Court — District of Massachusetts

788 F. Supp. 1267

COMMONWEALTH OF MASSACHUSETTS, and Massachusetts Department of Public Welfare, Plaintiffs, v. UNITED STATES of America, United States Secretary of Agriculture, Massachusetts State Board of Food Appeals, and United States Department of Food & Nutrition Service, Defendants.

Civ. A. No. 86-2132-Y.

United States District Court, D. Massachusetts.

April 6, 1992.

Douglas H. Wilkins, Asst. Atty. Gen., Boston, Mass., for plaintiffs.

Jeffrey Martin, Asst. U.S. Atty., Steven Zelinger, U.S. Dept. of Justice, Washington, D.C., for defendants.

MEMORANDUM AND ORDER

YOUNG, District Judge.

The Commonwealth of Massachusetts appeals from a determination by the Food Stamps Appeals Board ("Appeals Board") that Food and Nutrition Services ("Food and Nutrition"), the administrator of the food stamps program,¹ properly levied sanctions totalling $1,323,864 for violations from April to September 1982 of the Food Stamp Act of 1964 ("the Act"), 7 U.S.C. §§ 2011-2030. The parties have filed cross-motions for summary judgment on the validity of the agency action. The Commonwealth additionally seeks judgment that it is entitled to a "good cause" penalty waiver.²

The Commonwealth agrees that the mathematical computations involved comply with the applicable regulations, but it argues that the regulations themselves are biased against states in their choice of a sampling-type computation method.³ The Appeals Board affirmed the validity of the sanctions, finding that the sampling methods were "consistent with currently recognized approaches outlined in statistical texts and practices used in other governmental programs." See Letter from Orval Kerchner to Michael Putnam ("Kerchner Letter") at 1978. The Appeals Board ruled that the denial of a "good cause" penalty waiver was also proper, because Massachusetts did not exhibit any "exemplary effort" to reduce its error rate. Id. at 1979.

I. The Food Stamp Act of 1964

Congress, in an attempt to supplement the nutritional needs of low income families and "to alleviate ... hunger and malnutrition," 7 U.S.C. § 2011, enacted the Food Stamp Act of 1964. The Act provides for food stamp distribution to eligible families in order to increase their food purchasing power. The Secretary of Agriculture was empowered to formulate regulations consistent with the Act, including uniform eligibility requirements. 7 U.S.C. § 2013(c). The Secretary delegated nation-wide supervision of the program to Food and Nutrition. The states that choose to participate in the plan are required to designate an agency responsible for day-to-day operations. See 7 C.F.R. § 272.2 (1990). In Massachusetts, the Department of Public Welfare ("the Department") administers the food stamp program. The Commonwealth and the federal government share the administrative costs equally. See 7 U.S.C. § 2025(a).

A. Error Rates

In order to verify state compliance with the program, the federal government imposes a target error rate each year. The target error rate provides a margin for error within which there is no liability for administrative errors. If the official error rate exceeds the target error rate, then liability results. Indeed, the federal government is empowered to reduce food stamp funding by five percent for every percentage point by which the official error rate exceeds the range of the target error rate. 7 C.F.R. § 275.23(d)(3)(i) (1991). Any over-issuance of food stamps, either an issuance to ineligible households or an issuance of excessive amounts to eligible households, results in liability.

Rather than compile actual percentages based on the Department's distribution to individual households, Food and Nutrition uses a "sampling method" in which a cross-section or sub-sample of the households is evaluated in order to determine the official or actual error rate. 7 C.F.R. § 275.10-11. First, the state determines its own official error rate using a statistical sample, then Food and Nutrition chooses a sub-sample from the state's larger sample and reviews the official error rate.

The target rate computation is set out in 7 C.F.R. § 275.23(e)(2). For 1983 and subsequent years, the regulations enumerate the target error rates. For the relevant period — April to September of 1982 — however, the regulations provided a choice between two formulae for calculating the target error rate. These are the higher of either the national standard payment error rate during the first-half of 1981 or the state's actual payment error rate for the same period, with a mathematical allowance factored in. The parties do not dispute that the computed target error rate for the base period was 14.88 percent.⁴

Massachusetts' official payment error rate for the period was 16.35, exceeding the target rate by 1.47 percent.⁵ Food and Nutrition assessed the penalty for the period at $1,323,864. The Commonwealth does not dispute that Food and Nutrition accurately computed the penalty based on the regulations promulgated by the Secretary. Instead, Massachusetts calls into question the validity of the regulations themselves as biased in favor of the federal government and against the states.

B. Standard of Review

This Court is called upon to review the Appeals Board's approval of the action taken by Food and Nutrition. Pursuant to 7 U.S.C. § 2023, the Commonwealth is entitled to a trial de novo before this Court to review the actions of the Appeals Board.⁶ The Court is not limited to reviewing the record presented to the Appeals Board; the Court must examine all evidence and make its own findings of fact and conclusions of law. Modica v. United States, 518 F.2d 374, 376 (5th Cir.1975); Ramirez v. United States, 514 F.Supp. 759, 763 (D.P.R.1981). See also Dunn v. Secretary of U.S. Dep't of Agric., 921 F.2d 365, 366 (1st Cir.1990), citing New England Legal Found. v. Massachusetts Port Auth., 883 F.2d 157, 167 (1st Cir.1989). This Court cannot, however, simply ignore the findings of the Appeals Board. Kulkin v. Bergland, 626 F.2d 181, 185 (1st Cir.1980) ("We think that the de novo judicial review provision, a procedure for determining facts, is not aimed at displacing the agency's role"); see also Redmond v. United States, 507 F.2d 1007, 1011 (5th Cir.1975).

The First Circuit has adopted a bifurcated standard of review in food stamp cases, applying a de novo standard to the determination of the violation and a limited administrative review standard to the sanction imposed. See e.g., Kulkin, 626 F.2d at 184; Broad St. Food Mkt., Inc. v. United States, 720 F.2d 217, 220 (1st Cir.1983). In Broad St., for example, the First Circuit explained that review of the imposition of sanctions takes place under the "arbitrary and capricious" standard germane to the Administrative Procedure Act, 5 U.S.C. §§ 701—706, while review of agency fact finding takes place under a de novo standard. In that case, Food and Nutrition had suspended the food stamps program at a market for one year. The market did not challenge the finding of violations, but, on de novo review, it submitted new evidence that the sanction of disqualification was onerous to the surrounding community. The federal government objected to the new evidence. New evidence, the First Circuit held, is generally admissible in a de novo review, but the District Court had overstepped its bounds by examining new evidence regarding the choice of sanction. The sanction review⁷ was limited to an examination of "the sanction imposed in light of the administrative record to judge whether the agency properly applied its regulations." Broad St., 720 F.2d at 220. This is the standard imposed by the Administrative Procedure Act and espoused by the Supreme Court in Butz v. Glover Livestock Comm. Co., 411 U.S. 182, 185-86, 93 S.Ct. 1455, 1457-58, 36 L.Ed.2d 142 (1973). See also Kulkin, 626 F.2d at 184-85 (adopting the Butz standard for review of assessed sanctions).

This bifurcated approach gives appropriate controlling weight to the discretion of the administrative agency in choosing a sanction. Broad St., 720 F.2d at 220 (quoting Kulkin, 626 F.2d at 184 choice of sanction is "peculiarly a matter for administrative competence" and may be overturned only if it results from "arbitrary or capricious" agency action or is "unwarranted in law and in fact"). Moreover, the Act's legislative history supports this bifurcated approach. See H.R.Rep. No. 464, 95th Cong., 1st Sess. 397-98, reprinted in 1977 U.S.C.C.A.N. 1705, 1978, 2326-27 ("The trial de novo as set forth in 7 U.S.C. § 2023 should be limited to a determination of the validity of the administrative action, but not of the severity of the sanction. Review of the factual determination is normal grist for the courts; review of the length of highly discretionary a sic sentence of disqualification is not").

If the sanction is not warranted in law or fact, or if the Department of Agriculture acted in an "arbitrary and capricious" manner in imposing sanctions, this Court may reduce the penalty or instruct the agency to review the penalty. 7 U.S.C. § 2025(a); see also Broad St., 720 F.2d at 220. Summary judgment is an appropriate remedy if no genuine issues of material fact exist for trial. Modica, 518 F.2d at 376.

C. The Sampling Method and Penalty Bias

The stratified sampling method employed by Food and Nutrition is a reasonable and efficient method for determining the official error rate. See Hansen and Tepping Joint Declaration ¶ 16.⁸ Though alternative methods exist, the two defense experts hail the Food and Nutrition choice⁹ as having a statistical advantage over other methods. Id. It is undisputed, though, that some bias against the states does arise from the use of the stratified sampling method. The experts disagree on the impact that the bias has upon state liability levels. Hansen and Tepping describe the bias as "trivial," although in Massachusetts the bias represents 0.15 percent of federal administrative costs or 9.8 percent of the...