California Human Development Corporation v. William E. Brock, Secretary, Department of Labor

762 F.2d 1044, 246 U.S. App. D.C. 65, 1985 U.S. App. LEXIS 30030

• Court: Court of Appeals for the D.C. Circuit
• Decided: May 28, 1985
• Docket: 84-5321
• Status: Published

Opinions

BAZELON, Senior Circuit Judge:

Appellants challenge the Department of Labor’s (DOL) method for allocating funds under section 402 of the Job Training Partnership Act (JTPA).¹ Under this statute, the DOL distributes grants among states to support programs servicing migrant and seasonal farmworkers. The district court, on summary judgment, upheld the DOL’s formula for allocating funds. Because the DOL’s method is reasonably consistent with its congressional mandate, the district court’s decision is affirmed.

I. Background

The appellants, grantees of farmworker training funds under the JTPA, contest the DOL’s formula for distributing the funds among the states. These grantees are nonprofit organizations residing in states receiving reduced allocations in fiscal year 1984.² They claim that use of the challenged formula results in allocations inconsistent with the proportional number of eligible migrant and seasonal farmworkers in each state.

This controversy centers on how the DOL should annually distribute $57 million in farmworker training funds.³ Here the total amount of the budget is not at issue. Instead, the parties dispute the method for distributing these extremely limited public funds among many needy programs. There are enough funds to service only one-tenth of the total eligible population.⁴ Thus, no matter how the agency chooses to allocate its funds, only a fixed and small proportion of the eligible poor can be benefitted.

Under the Job Training Partnership Act, the DOL administers federal grant programs for the training and reemployment of low-income workers. Section 402 of JTPA provides the statutory basis for a particular program that makes .federal grants to eligible public or private agencies for the purpose of providing job training, employment opportunities, and other services to migrant and seasonal farmworkers. In establishing this program, Congress found that “chronic seasonal unemployment and underemployment in the agricultural industry, aggravated by continual advancements in technology and mechanization resulting in displacement, constitute a substantial portion of the Nation’s rural employment problem and substantially affect the entire national economy.”⁵

Under the predecessor statute to JTPA— the Comprehensive Employment Training Act (CETA)⁶ — the Secretary of Labor used 1977 Social Security industrial data in estimating the location and extent of the beneficiary farmworker population.⁷ The DOL allocated federal funds in proportion to the estimated number of eligible beneficiaries in each state. The DOL calculated the total number of individuals in each state working in agriculturally-related industries with reported earnings of $3,000 or less⁸ and for whom Social Security taxes are paid.

The data base for determining allocations was changed when the JTPA was enacted in 1982.⁹ The JTPA mandated a switch in the data base, stating that: “All data relating to economically disadvantaged and low-income persons shall be based on 1980 Census or later data.” ¹⁰ Furthermore, the legislation set forth procedures that the DOL must follow if the agency wished to rely upon a particular formula in its allocations:

Whenever the Secretary utilizes a formula to allot or allocate funds made available for distribution at the Secretary’s discretion under this chapter, the Secretary shall, not later than 30 days prior to such allotment or allocation, publish such formula in the Federal Register for comments along with the rationale for the formula and the proposed amounts to be distributed to each State and area. After consideration of any comments received, the Secretary shall publish final allotments and allocations in the Federal Register.¹¹

The new legislation also established deadlines for dispersal of the federal funds to states.¹²

Given the imposed time restraints, the Department was forced “to make complex decisions in a relatively short period of time.” ¹³ The DOL sought to establish the allocation apparatus in time for the transition period October 1983 through June 1984 during which $43 million was allocated. In so doing, the DOL dutifully complied with the procedures outlined in the JTPA.¹⁴'' On August 12, 1983, the Department published for review and comment the proposed formula based on the 1980 Census data.¹⁵ The DOL explained that “the proposed planning estimates are based on the total number of people in each State who worked in qualifying agricultural occupations and reported a poverty level income ____”¹⁶

The DOL’s regulations concerning the future application of the formula were effective October 1, 1983. The regulations provided that up to six percent of section 402 funds may be used for discretionary federal projects,¹⁷ and that “[n]o less than 94 percent of the funds received for section 402 activities shall be allocated for farm-worker programs in individual States in an equitable manner using the best data available as to the farmworker population as determined by the Department.” ¹⁸ Recognizing that the congressionally-mandated shift in data bases could result in “disruption of funding levels from one year to the next,” the DOL included a “hold harmless” provision limiting to twenty-five percent any reduction in a State’s allotment from the prior year.¹⁹

The regulations also included a definition of “farmwork” as “work performed for wages in agricultural production or agricultural services as defined in the most recent edition of the Standard Industrial Classification (SIC) Code definitions ----”²⁰ However, the regulations expressly provided that the definition is applicable only “for eligibility purposes,” as distinguished from allocation purposes. Under the eligibility regulations, a low-income farmworker is eligible if at least fifty percent of his income or time during a specified twelve-month period was in farmwork; if he performed farmwork for wages; earned at least $400 performing farmwork, or performed farmwork for at least twenty-five days; and was employed on a seasonal basis, without a constant year-round salary. Dependents of these individuals are also eligible.²¹

In December 1983, following review of public comments, specific allotments for each state were fixed for the transition period October 1983 through June 1984. The DOL further explained that “[t]he most recent available [Social Security] data are for 1977,” and therefore it had to rely on the 1980 Census under the JTPA’s “latest available data” requirement.²² The DOL indicated that the Social Security data included persons in agriculturally-related industries. In contrast, the Census count — as relied upon by the DOL — used occupational classifications. The industrial classifications or codes, that were formerly used, “captured many non-farm-workers such as clerical, administrative, and technical employees of agricultural firms, and therefore vastly overstated the farmworker population.”²³

In January 1984, the DOL announced the state allotments for program year July 1984 through June 1985. The same formula was used to distribute $57 million during this second time interval. Thus, the DOL acted expeditiously to meet statutory deadlines. However, the DOL had assured concerned parties that there would be “on-going review,” so in subsequent years the best data available (as determined by the Secretary of Labor) would be used.²⁴ The DOL has employed an interagency task force to review the data base for the purpose of suggesting improvements and refinements.²⁵

In this suit, the appellants challenge the DOL’s allocation of JTPA section 402 funds, alleging that: (1) the DOL violated its own regulations by failing to use the best data available and by failing to allocate funds equitably; (2) the DOL violated the JTPA by not providing a rationale for the formula and by not considering comments; and (3) the DOL’s action was arbitrary and capricious.

After a period of discovery, the district court dismissed the suit on summary judgment. The district court concluded that the DOL arrived at an allocation formula based on factual data reasonably available to it, and that the Department’s actions could not be characterized as arbitrary and capricious.²⁶ Because the DOL’s shift in policy was mandated by Congress and because the Department did not violate its own regulations, the district court’s decision is affirmed.

II. Discussion

The DOL’s actions in allocating JTPA section 402 funds must conform to relevant statutes and to the agency’s own regulations.²⁷ The JTPA does not preclude judicial review.²⁸ Indeed, the statute expressly requires the Secretary of Labor to follow a carefully delineated process that includes providing a “rationale for the formula,” ²⁹ if it decides to use a formula.

In this case, the DOL substantially conformed to the JTPA’s requirements and to the Department’s own prescriptions. In evaluating the reasonableness of the agency’s compliance, this court recognizes the statistical expertise of the agency.³⁰ Furthermore, the DOL’s actions, while not elegant, were not improper given several ameliorating considerations. First, the “hold harmless” clause cushions the impact of the new formula. Second, although the DOL was forced to make difficult decisions in short time, provisions exist to modify the formula as more substantial information becomes available.³¹ And finally, here the DOL was allowed under the JTPA not to use any formula. Instead, the Department could have decided to allocate all the funds at its discretion. Mindful of these •considerations, our review is limited to whether the formula and its application resulted from a consideration of relevant factors and whether there was a clear error of judgment.³² The estimates made by an agency, if explained, are entitled to deference by courts.

A. The Agency’s Regulations

Even if the action of the agency is discretionary by statute, the agency’s own regulations must be complied with by the agency after they are issued.³³ Appellants contend that the DOL violated its own regulations (1) by not using the best data available as to the farmworker population as required by 20 C.F.R. § 633.105(b)(1) (1983); (2) by not distributing the funds in an equitable manner as required by the same regulation, id.; and (3) by not relying in the formula upon industrial classifications, instead of occupational classifications, which are referred to in the definition of “farmwork” in id. § 633.104.

1. “Best Data Available ” Requirement

The “best data available” requirement is difficult to analyze in this instance.³⁴ Unfortunately, no one knows the exact extent or location of eligible seasonal and migrant farmworkers. Neither the Social Security nor the Census data provide a direct answer — approximations must be deduced from sets of classifications or codes. While Census data do not include those who fail to respond to the Census questionaires, Social Security data do not include those whose Social Security taxes are not accounted for.

Even the different kinds of classifications or codes within these data bases result in either under- or over-inclusiveness. The appellants point out that because the DOL used Standard Occupational Classification (SOC) codes, the data are overinclusive. Occupational codes include self-employed individuals who are not wage-earners, contrary to eligibility requirements. On the other hand, industrial codes are also overinclusive by counting non-farmworkers such as clerical, administrative, and technical employees of agricultural firms.³⁵ The DOL, furthermore, explained that because the allocation data base specifies only poverty-level agricultural workers, the resulting approximation using occupational classifications would include poverty-level tenant farmers and share-croppers who supplement their marginal farming receipts with farm wages.³⁶ These poor have been traditionally served by the program. Given the uncertainty surrounding these statistics, as well as the promised on-going review of the data base by the DOL, we cannot find the Department’s actions unreasonable.

2. “Equitable Distribution” Requirement

Appellants raise similar issues with regard to the “equitable distribution” requirement.³⁷ They contend that it was inequitable for the DOL not to include in the data base specific eligibility requirements: wage earners, SIC codes, public assistance participants, lower living standard income level, or dependents. Aside from the above-mentioned uncertainty in the accuracy of available data and the provision for on-going review procedures, we note that the DOL invoked the “hold-harmless” provision that cushioned the impact of the congressionally-mandated shift in database.³⁸ Hence, the Department did consider the past expectations of parties and the unfairness of abruptly shifting policies. The DOL continued, in an equitable fashion, to use an allocation formula that was applied uniformly among states, proportionate to the farmworker tally in each state.

With respect to the DOL’s failure to include all eligibility factors, the DOL has never used a data-base that precisely identified the groups of persons eligible to receive services. As a practical matter, this is an impossible task. The Census Bureau does not maintain data on the number of seasonal and migrant farmworkers. The 1980 Census household data does not indicate whether income is earned on a seasonal basis rather than a year-round basis, and does not show if reported earnings were actually received from working in the reported occupation.³⁹

The DOL’s continued use of a poverty level factor of $3000 was reasonable. Given the time constraints, the DOL chose to use this factor as an appropriate approximation of eligible beneficiaries — those below 70 percent of the lower living standard income level or on public assistance. The district court noted that the DOL was following precedent that previously had not been challenged.⁴⁰

Similarly, regarding dependents of eligible farmworkers, the district court correctly concluded that the DOL’s decision to use the number of farmworkers as an approximation for the proportional number of dependents was reasonable. Dependents could qualify for only 15 percent of grant funds for non-training supportive services.⁴¹ The district court found that “there is no basis for assuming that dependents would be more concentrated in particular geographical areas.”⁴²

3. Industrial Classifications and Eligibility Requirements

Section 633.104 includes the following definition:

Farmwork shall mean, for eligibility purposes, work performed for wages in agricultural production or agricultural services as defined in the most recent edition of the Standard Industrial Classification (SIC) Code definitions____⁴³

This definition apparently had been carried over from the previous CETA allocation regulations under which the DOL had relied upon industrial classifications when the agency used the Social Security data base. Because this definition is expressly applicable only for eligibility purposes, the DOL is not required under its own rules to use industrial classifications for allocation purposes.

B. The Job Training Partnership Act

Appellants argue that the DOL did not provide a rationale for the formula and that the DOL failed to consider comments as required by the JTPA. Both contentions must be rejected. The DOL’s resort to 1980 Census data was a major policy shift mandated by Congress. The Department determined that “the 1980 Census data best meets the JTPA, Section 162(a) ‘latest available data’ requirement.”⁴⁴ The formula was chosen because it was most consistent with the will of Congress.⁴⁵

After reviewing the administrative record, we agree with the district court that the DOL responded in a reasonable manner to the important comments that were submitted. The agency explained that Census occupational data was superior to Social Security industrial data since it counted only persons actually performing agricultural work, and described the Census allocation formula that was used. The agency furthermore promised “on-going review and analysis of data sources” in future years.⁴⁶

C. The Reasonableness of the Agency’s Actions

Finally, the appellants contend that the DOL’s actions were arbitrary, capricious, and an abuse of discretion.⁴⁷ The district court, however, was correct in finding that there was a reasonable relationship between the eligibility criteria and the allocation data base.⁴⁸ Again we emphasize that this conclusion was reached in view of the “hold harmless” clause and the on-going review provision by the DOL. The DOL’s actions were rational, given the information that the DOL had at the time the agency promulgated the regulations. Complex decisions had to be made in a short time span. The change in allocation pattern was mainly due to the substitution of the 1980 Census data for the 1977 Social Security data. “[TJhat choice must be laid at the doorstep of the Congress.”⁴⁹ At least for the DOL’s fiscal year 1983 and 1984 allocations, this court cannot find the agency’s allocation formula to be arbitrary and capricious.

Summary judgment was appropriately issued in favor of the DOL by the district court. The agency factfinding procedures were found to be adequate, and reasonable explanations were provided. The DOL, especially under these circumstances, acted well within its discretion. The reasonableness of the agency’s justifications results from putting the focal point of review on the administrative record.⁵⁰

III. Conclusion

For the reasons cited above, the decision of the district court is

Affirmed.

1 . 29 U.S.C. § 1672 (1982).

2 . California’s percentage share was reduced from 21.69 to 8.80; Florida’s from 7.68 to 4.83; New York’s from 2.68 to 2.47; and Utah’s from 0.47 to 0.38. In contrast, South Dakota’s percentage share increased from 0.47 to 2.48; Kentucky’s from 1.16 to 3.26; and Iowa's from 1.86 to 4.12. See Joint Appendix (J.A.) at 135. But see infra, note 38 ("hold harmless” provision by agency limits impact of change in percentage share).

3 . Job Training Partnership Act; Program Year 1984 Transition Period; State Allotments for Migrant and Seasonal Farmworker Programs, 49 Fed.Reg. 3942, 3943 (1984).

4 . See Appellee’s Brief at 21.

5 . 29 U.S.C. § 1672(a)(1) (1982).

6 . 29 U.S.C. § 873 (1976).

7 . Prior to CETA, funds were allocated under the Economic Opportunity Act on a discretionary basis without use of a formula.

8 . This statistical cut-off was consistent with the poverty level for a single person defined by the Office of Management and Budget.

9 . The JTPA was enacted on October 13, 1982, to be effective by October 1, 1983.

10 . 29 U.S.C. § 1572(a) (1982).

11 . Id. § 1572(d).

12 . See id. § 1572(c).

13 . California Human Dev. Corp. v. Donovan, No. 84-5321, Memorandum Opinion at 6 (1984) [hereinafter cited as Memo.Op.].

14 . See supra note 11. In July 1983, the DOL published proposed regulations to implement the JTPA program.

15 . Employment and Training Administration; Migrant and Seasonal Farmworker Programs; Fiscal Year 1984 Proposed State Planning Estimates, 48 Fed.Reg. 36688 (1983).

16 . Id.

17 . 20 C.F.R. § 633.105(a) (1983).

18 . Id. § 633.105(b)(1).

19 . Id. § 633.105(b)(3).

20 . Id. § 633.104.

21 . Id. § 633.107.

22 . Job Training Partnership Act; Fiscal Year 1984 Transition Period; State Allotments under Title IV-A, Section 402 Migrant and Seasonal Farmworker Programs, 48 Fed.Reg. 54914, 54915 (1983). Additionally, Social Security data included only those individuals who paid Social Security taxes. Id.

23 . Memo. Op. at 3-4; see 48 Fed.Reg. at 54915 (1983).

24 . 48 Fed.Reg. at 54915. The DOL acknowledged in its brief that on-going administrative review of the problem was required by the statute and the regulations. Appellee’s Brief at 43; see 29 U.S.C. § 1572(a) (1982) (requiring use of "latest available data”); 20 C.F.R. § 633.-105(b)(1) (1983) (requiring use of “the best data available").

25 . Appellee's Brief at 42-43.

26 . Memo.Op. at 7-8.

27 . While the district court, in dicta, indicated that "[a]t worst from the plaintiff s'point of view, the Secretary’s choice of data base is entirely nonrevicwable ..." Memo. Op. at 7 (emphasis added), it did in fact review the reasonableness of the formula and found that the DOL's actions were not arbitrary and capricious. Id. at 7-8.

28 . 5 U.S.C. § 701(a) of the Administrative Procedure Act "withholds agency action from judicial scrutiny when: (1) statutes preclude judicial review; or (2) agency action is committed to agency discretion by law." Neither exception to the general presumption of judicial review applies in this instance. The § 701(a)(1) exception requires explicit statutory language precluding review, which the JTPA does not contain. The § 701(a)(2) exception is a very narrow one to be applied in rare instances when "statutes arc drawn in such broad terms that in a given case there is no law to apply.” Citizens to Preserve Overton Park v. Volpe, Inc., 401 U.S. 402, 410, 91 S.Ct. 814, 821, 28 L.Ed.2d 136 (1971); see Local 1219, Am. Fed'n of Gov't Employees v. Donovan, 683 F.2d 511 (D.C.Cir.1982).

Judicial review of allocation formulae cannot be described as "unaccustomed." Indeed, we have previously indicated its importance. See Batterton v. Marshall, 648 F.2d 694 (D.C.Cir.1980). This is not an instance where "there is no law to apply.” Appellants alleged that the Department failed to comply with the JTPA because it did not provide a rationale and failed to consider comments. Appellant’s Brief at 24-32. Furthermore, it is at least conceivable that an allocation formula may be so invidiously discriminatory that it becomes "inequitable” in violation of 20 C.F.R. § 633.105(b).

The concurrence relies upon Heckler v. Chaney, -U.S.-, 105 S.Ct. 1649, 84 L.Ed.2d 714 (1985), to conclude that the agency’s discretion is unreviewable. Chaney, however, did not involve an analogous situation. It concerned an agency’s refusal to take enforcement action under a statute. Hence, judicial review was inappropriate because of the lack of statutory guidelines and the longstanding tradition of absolute discretion in prosecutorial decisionmaking. — U.S. at-, 105 S.Ct. at 1655. In the instant controversy, the agency decided to act under the authority of regulations approved by Congress. Judicial review is necessary to determine if the agency is following congressionally mandated guidelines that require a rationale, consideration of comments, and equitability. We apply a reasonableness standard, which gives considerable deference to the agency’s expertise and time constraints, in defining these requirements. This case is analogous to the traditional requirement that' if an agency decides to promulgate rules, then it is bound by its own regulations even if the action of the agency was discretionary. Service v. Dulles, 354 U.S. 363, 77 S.Ct. 1152, 1 L.Ed.2d 1403 (1957).

29 . 29 U.S.C. § 1572(d) (1982).

30 . The regulations contain limiting language that implicitly recognizes the DOL’s expertise. See 29 U.S.C. § 1572(a) (1982) ("All allotments and allocations under this chapter shall be based on the latest available data and estimates satisfactory to the Secretary.") (emphasis added); 20 C.F.R. § 633.105(b)(1) (1983) (section 402 funds shall be allocated “in an equitable manner using the best data available as to the farmworker population as determined by the Department ”) (emphasis added); see abo Chemical Mfrs. Ass’n v. NRDC, — U.S.-,-, 105 S.Ct. 1102, 1110, 84 L.Ed.2d 90 (1985); State v. EPA, 696 F.2d 147, 159-60 (2d Cir.1982).

31 . See supra note 24; see, e.g., National Soft Drink Ass’n v. Block, 721 F.2d 1348, 1354-55 (D.C.Cir.1983).

32 . Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 823, 28 L.Ed.2d 136 (1971).

³³

34 . This requirement is followed by limiting language in the regulation that implicitly recognizes the agency’s expertise. See supra note 30.

35 . See supra note 23.

36 . See 48 Fed.Reg. at 54915; Memo.Op. at 9.

After oral argument, the appellants moved to supplement the record with newly-available data allegedly demonstrating that only a low percentage of this population is eligible. We deny this motion because the information was not submitted to the agency at the time it made its allocations for the fiscal years 1983 and 1984.

37 . See supra notes 30 & 31.

38 . For the fiscal year 1984, no state would be reduced by more than 25 percent of the amount received in fiscal year 1983. The amounts are reduced an additional 25 percent each year over three years. Thereafter each state is allocated the amount it would receive by direct application of the data without a hold harmless provision. 48 Fed.Reg. 36688 (1983).

39 . Appellee’s Brief at 22.

40 . Memo.Op. at 5; see supra note 8.

41 . 20 C.F.R. § 633.304 (1983).

42 . Memo.Op. at 5-6.

43 . 20 C.F.R. § 633.104 (1983) (emphasis added).

44 . 48 Fed.Reg. 54915 (1983).

45 . Memo.Op. at 7.

46 . 48 Fed.Reg. 54915 (1983). See National Soft Drink Ass’n v. Block, 721 F.2d 1348, 1354-55 (D.C.Cir.1983) (affirming agency's imprecise category approach because full provision made to update categories as more information is forthcoming).

47 . See Administrative Procedure Act, 5 U.S.C. § 706(2)(A) (1982).

48 . "It is not necessary in this case to explore in what circumstances that relationship would be so far distant as to be no longer reasonable. In view of the many imponderable factors, perfect symmetry [between eligibility criteria and the allocation data base] was not achievable____” Memo.Op. at 5 n. 7.

49 . Id. at 6.

50 . See National Org. for Women, Washington, D.C. Chapter v. Social Security Admin., 736 F.2d 727, 745 (D.C.Cir.1984) (Mikva and McGowan, JJ., concurring).