Opinion
Civil No. 03-624 (JRT/FLN).
August 26, 2004
Elizabeth Hendricks Schmiesing, Richard A. Duncan, Brian Boru O'Neill, and Colette Routel, FAEGRE BENSON, Minneapolis, MN, for plaintiffs.
Joan Humes, Assistant United States Attorney, OFFICE OF THE UNITED STATES ATTORNEY, Minneapolis, MN, for defendants Bosworth and Venemen.
David Russell Oberstar, FRYBERGER BUCHANAN SMITH FREDERICK, Duluth, MN, for intervenor-defendants Conservationists with Common Sense, Ely Outfitters Association, Gunflint Trail Outfitters Association, Seagull-Saganaga Homeowners Association
MEMORANDUM OPINION AND ORDER GRANTING PLAINTIFFS' MOTION FOR SUMMARY JUDGMENT AND DENYING DEFENDANTS' MOTIONS FOR SUMMARY JUDGMENT
Plaintiffs, organizations dedicated to the protection and enjoyment of national forests and wildlife areas, challenge the United State Forest Service's ("USFS") decision to increase the available permits for motorized boats in the Boundary Waters Canoe Area Wilderness ("BWCAW"). By statute, the number of permits cannot exceed the statutory cap. Plaintiffs challenge the USFS's allegedly unauthorized recalculation of the statutory cap; the recalculation of the cap in an allegedly arbitrary and capricious manner; and the increase in quotas in certain areas of the BWCAW. The government defendants argue that the decision to raise the quotas was not arbitrary and capricious, and was necessitated by the Eighth Circuit's decision in Friends of the Boundary Waters Wilderness v. Dombeck, 164 F.3d 1115 (8th Cir. 1999). A group of concerned citizens, home and resort owners, and outfitters intervened in support of the USFS's decision to increase the available motorboat permits. All parties have cross-moved for summary judgment. For the reasons discussed below, the Court grants plaintiffs' motion for summary judgment and denies the defendants' motions.
BACKGROUND
I. BOUNDARY WATERS CANOE AREA WILDERNESSThe Superior National Forest in northeastern Minnesota spans 150 miles along the United States-Canadian border. Established as a National Forest in 1909 by proclamation of President Theodore Roosevelt, the Superior National Forest encompasses a "unique wilderness" — the Boundary Waters Canoe Area Wilderness. The BWCAW consists of approximately one million acres of forested lake land along the Minnesota-Ontario border. See Dombeck, 164 F.3d at 119; Minnesota v. Block, 660 F.2d 1240, 1245 (8th Cir. 1981).
The Eighth Circuit described the formation of the BWCAW in an unrelated opinion:
A sponsor of this legislation described the area in introducing the BWCAW Act on the House floor:
The Boundary Waters Canoe Area is the largest wilderness area east of the Rocky Mountains and the second largest in our wilderness system. It is our Nation's only lakeland canoe wilderness a network of more than 1,000 lakes linked by hundreds of miles of streams and short portages which served as the highway of fur traders who followed water routes pioneered by Sioux and Chippewa Indians. Despite extensive logging, the BWCA still contains 540,000 acres of virgin forests, by far the largest such area in the eastern United States. * * * (M)any western wilderness areas lack such complete food chains. This natural ecosystem is a valuable educational and scientific resource; it has been the focal point of important research in wildlife behavior, forest ecology, nutrient cycles, lake systems, and vegetation history.
The BWCA is complemented on the Canadian side of our border by the Quetico Provincial Park of Ontario where commercial logging and nearly all motorized recreational activity are prohibited. Together, these areas encompass an area the size of Yellowstone National Park and constitute one of the finest wilderness areas on our continent. Not surprisingly, the BWCA is the most heavily used unit in the national wilderness system, drawing people from throughout the country who seek the solitude of a wilderness experience.Block, 660 F.2d at 1245 (citing 123 Cong. Rec. H621-22 (daily ed. Jan. 31, 1977), reprinted in Legislative History of the Boundary Waters Act of 1978, at 1-2.)).
Visitor demand on the BWCA continues to be impressive. The Superior National Forest's website indicates that the BWCA Wilderness attracts thousands of visitors each season.
In the Wilderness Act of 1964, Congress recognized the unique features and value of the BWCAW, and designated the BWCAW as one of the initial wilderness areas protected by the Act. See 16 U.S.C. § 1132. The 1964 Act prohibited use of motorized vehicles in any national wilderness area. 16 U.S.C. § 1133(c). That same Act, however, provided a specific exception for the BWCAW:
Other provisions of this chapter to the contrary notwithstanding, the management of the Boundary Waters Canoe Area * * * shall be in accordance with regulations established by the Secretary of Agriculture in accordance with the general purpose of maintaining, without unnecessary restrictions on other uses, including that of timber, the primitive character of the area, particularly in the vicinity of lakes, streams, and portages: Provided, That nothing in this chapter shall preclude the continuance within the area of any already established use of motorboats.16 U.S.C. § 1133(d)(5) (1976). This was not, however, Congress's last word on motorized boats in the BWCAW. "In response to the confusion and litigation generated by the above proviso, as well as in reaction to threatened deterioration of the wilderness from excessive use, Congress enacted the [BWCAW Act] of 1978." Block, 660 F.2d at 1246. The BWCAW Act eliminated all motorized use within the wilderness except on those lakes specifically enumerated in the Act. Pub.L. No. 95-495, 92 Stat. 1649 (1978).
In the 1978 Act, Congress directed the Secretary of Agriculture to develop and implement entry point quotas to govern and restrict the use of motorboats on those particular lakes where Congress had legislatively authorized their restricted use. 92 Stat. at 1651, § 4(f). "The only specific guidance given to the Secretary concerning these quotas was a statutory cap on motorboat use, which prescribes that motorboat use "shall not exceed the `average actual annual motorboat use' during the years 1976 through 1978." Dombeck, 164 F.3d at 1120 (quoting 92 Stat. at 1651 § 4(f)).
Congress specifically directed the Secretary to:
[D]evelop and implement, as soon as practical, entry point quotas for use of motorboats within the wilderness portions of the lakes listed in subsection c, the quota levels to be based on such criteria as the size and configuration of each lake, and the amount of use on that lake: Provided, That the quota established for any one year shall not exceed the average actual annual motorboat use of the calendar years 1976, 1977, and 1978 for each lake, and shall take into account the fluctuation in use during different times of the year: Provided further, That on each lake homeowners and their guests and resort owners and their guests on that particular lake shall have access to that particular lake and their entry shall not be counted in determining such use.92 Stat. at 1651, § 4(f) (emphasis added).
Although Congress clearly directed the Secretary to establish a statutory cap, as the USFS points out, Congress did not establish the method for establishing the cap. In 1981 the USFS established a methodology for calculating "average actual annual use" by assembling sources including computer analysis of available permit data; other records of motor use in 1976, 1977 and 1978; and public comments received by mail, at meetings, and during preparation of the plan. The USFS suggests that as part of the original calculation of the statutory cap, the USFS did not count use by homeowners, resort owners and their guests, when that use was on the "particular lake" on which the home or resort owner resided, and also did not count such use on the "chain of lakes" of which "that particular lake" was part. (Gov't Def. Mem. in Supp. of Summ. J. at 9-10.)
The USFS's interpretation of the phrase "particular lake" to include "chain of lakes" was challenged and rejected in Dombeck, discussed below.
Using such a method, the USFS determined the "average actual annual motorboat use" from 1976-78. (Administrative Record ("A.R.") Tab 53 at 1.) This number — 14,925 — then became the statutory cap. The BWCA Wilderness is now managed in "accordance with a 1986 Land and Resource Management Plan for the Superior National Forest, as amended by the BWCA Wilderness Management Plan and Implementation Schedule of 1993 (the Wilderness Plan). Dombeck, 164 F.3d at 1120. The Wilderness Plan announced motorboat quotas that the USFS "established . . . after considering the pertinent legislation, [USFS] policy, the needs of the environment, the historic uses of the area, and the recreational needs of the visitors . . . the Wilderness Plan restricts visitor and motorboat use within the BWCA through a quota system, entry point restrictions, special permits for commercial towboats and a special exemption from the motorboat quota system for homeowners, resort owners, and their guests." Id. (footnote omitted).
Motorboat quotas have been reduced since 1981, in part based on overcrowding in the wilderness, and in part based on phase-outs required by the BWCAW Act. Prior to the recalculation at issue here, the statutory cap had been reduced to 10,539. BWCAW Act § 4(c)(3)-(4).
II. FRIENDS OF THE BOUNDARY WATERS WILDERNESS v. DOMBECK
The dispute in this case is fairly characterized as an extension of one of the issues the Eighth Circuit addressed in Friends of the Boundary Waters Wilderness v. Dombeck, 164 F.3d 1115 (8th Cir. 1999). In Dombeck, a group of concerned citizens and outfitters (many of the same groups intervening here) challenged the motorboat quotas as determined by the USFS. The outfitters claimed that the motorboat quotas, visitor use restrictions, and the USFS's definition of "guest" unduly limited access to the BWCAW. A group of environmental organizations (many of the same organizations that are plaintiffs here) intervened and challenged two aspects of the USFS Wilderness Plan. The environmental organizations argued that the USFS's use of "special use permits" for towboats, and the Wilderness Plan's grouping of certain listed chains of lakes on which homeowners, resort owners, and guests were exempt from the quota system, allowed excessive motorized use in the area in violation of the BWCA Wilderness Act and the Administrative Procedures Act.
The "chains of lakes" that the USFS initially considered as one lake for the purpose of administering the BWCAW Plan were (1) the Moose Lake chain (Moose, Sucker, Newfound, and Birch Lakes); (2) the Farm Lake Chain (White Iron, Farm, Garden, and South Farm Lakes); (3) Saganaga Lake Chain (Seagull River, Gull Lake, and Saganaga Lake).
After a discussion of the history of the BWCAW, the Dombeck Court rejected the outfitters' challenges to the USFS's definition of "guest." Id. at 1123 (finding reasonable the USFS definition of "guest" to include only nonpaying, overnight guests of homeowners and resort owners). The court also rejected the environmental organizations' contention that the special use permit system for towboats was unreasonable. Id. at 1121-22 (concluding that a separate system for monitoring commercial towboats is reasonable, and reasoning that "[t]he combined number of motorized boat trips that the Plan allows (1,342 + 7,902) totals 9,244 trips, which does not exceed the 10,539 motorboat trips cap mandated by the BSCA Wilderness Act.").
The Eighth Circuit also rejected the outfitters' claim that the USFS had violated the National Environmental Policy Act ("NEPA") and held that the environmental impact statement underlying the plan was adequate under NEPA. The NEPA issue is not relevant to this lawsuit.
The Eighth Circuit, however, upheld the environmental organizations' challenge to the USFS's reading of the phrase, "that particular lake," in relation to the quota exemption for homeowners and resort owners. Id. at 1123-25. The court found unreasonable the USFS's interpretation of the phrase, "that particular lake," to include certain chains of lakes. Id. at 1123-24.
The Eighth Circuit characterized the successful argument as follows:
The Environmentalists challenge the agency's decision to consider each chain of lakes as one lake for purposes of exempting property owners and their guests from the motorboat use quotas, contending that this policy impermissibly expands the motorboat use quota exemption. The Environmentalists contend that this statutory exemption from motorboat use quotas gives property owners and their guests quota-free access only to the named lake that abuts their property.Id. at 1124.
The Eighth Circuit found this argument persuasive, and rejected the USFS's rationale for including certain "chains of lakes" in the exemption. Although the court recognized that the USFS had consistently interpreted the phrase in the same manner, the court found the phrase "that particular lake" to be an unambiguous statement by Congress. Id. at 1124. "It is not reasonable," the Dombeck Court stated, "for the Forest Service to broaden this restrictive phrase by redefining the term `lake' to include several connected but individually named lakes. . . ." Id. The court concluded, "[g]iven the plain language of the subsection at issue and its context, we are convinced that the homeowner/resort owner and guest exception to the motorboat use quota system should remain as narrow as Congress has written it and that the agency's interpretation is not permissible because it is contrary to the clearly expressed intent of Congress." Id. The environmentalist plaintiffs were awarded summary judgment on that issue. Id.
The case was remanded, and the District Court issued a short order pursuant to the remand Specifically, the District Court ordered summary judgment entered in favor of plaintiffs on Count III of their complaint. The District Court ordered the USFS to "take the necessary steps to ensure that the BWCA Wilderness Management Plan and Implementation Schedule and supporting documents are consistent with the Eighth Circuit opinion that property owners and their guests are exempt from motorboat quotas only on the lake immediately adjoining their property." (Order dated September 3, 1999 in Civil No. 5-95-10 (JMR).)
III. USFS RECALCULATION AND QUOTA INCREASE FOLLOWING DOMBECK
After the Dombeck decision, the USFS began to recalculate the statutory cap as the USFS believed was required by the decision. The USFS proceeded by determining the number of home/resort owners who they speculated had been counted in calculating "average actual annual use" in 1981. To determine the number of home/resort owners, the USFS used sources including 1982 aerial photographs; county assessor records; USFS reports; and interviews with home/resort owners. The USFS then multiplied that result by the number of days of use. Days of use per affected party also required a number of sources, according to the USFS. This product [(number of uncounted home/resort owners) X (days of use)] then was used to establish the purportedly corrected average actual annual use by home/resort owners in 1976, 1977, and 1978. The revised statutory cap added 9,445 motorboat permits to the previous total of 14,925.
Although Dombeck issued in 1999, the USFS was unable to implement changes to the cap until 2002, because of the urgent work necessitated by a massive blow-down of trees in 1999.
The USFS argues that it merely corrected the cap by taking the 1981 cap and adding the affected party use which would have been counted had the USFS originally used the Dombeck definition of "particular lake." The result of this recalculation is a statutory cap of 12,650 motorboat permits on the Moose, Farm, and Saganaga Lake Chains. (Admin. Record Tab 17 at 9.) The original statutory cap was 3,205 motorboat permits for the lake "chains." After the USFS recalculated the statutory cap, they also revisited the motorboat quotas, and issued the Chain of Lakes Environmental Assessment ("EA") on September 12, 2001. The EA proposed alternatives for the quota numbers, one of which, Alternative 2, proposed an increase from 2,376 to 6,892. This Alternative was adopted by the Forest Supervisor on February 11, 2002. Plaintiffs appealed the decision, and the Regional Forester denied the plaintiffs' appeal in its entirety.
IV. THE CHALLENGE
Plaintiffs challenge the recalculation on several grounds. First, they argue that the BWCAW Act does not authorize the USFS to recalculate the statutory cap — plaintiffs argue that the BWCAW Act provided for a single occasion to calculate the statutory cap. As a subpart of this argument, plaintiffs argue that recalculation is demand driven and not required by Dombeck.
Plaintiffs alternatively argue that the USFS recalculation of the statutory cap was arbitrary and capricious. In support of this argument, plaintiffs claim that the USFS relied on unsupported assumptions, "double-counted" exempt users on the Moose Lake Chain; conducted surveys that are fatally flawed; and improperly used general occupancy data from 1979 to estimate actual resort occupancy on the three at-issue lake chains.
Plaintiffs complain about how the USFS counted use in September of 1976. In that month no "outside" motorboat users were allowed into the BWCAW due to fire danger. Plaintiffs argue that the plain language of the statute requires an entry of "0" for September of 1976, because that is what the "actual use" was. The USFS instead "averaged" the use in September 1977 and 1978 as an estimate for September 1976. Finally, the plaintiffs suggest that the increase in motor quotas was arbitrary and capricious. Plaintiffs note that if the new statutory cap is invalidated, the new quotas are illegal because they exceed the 1981 statutory cap. Plaintiffs argue that raising the cap is contrary to Congress's expressed intent to keep the wilderness "wild." Plaintiffs argue that the USFS did not consider mandatory factors when setting the new cap/quotas, because USFS did not consider wilderness values, and did not consider the individual characteristics of the individual lakes.
ANALYSIS
The Court agrees with the USFS's suggestion that "the Court must determine here . . . whether the USFS decision to correct the statutory cap and entry point quotas to incorporate the holding in Dombeck was arbitrary, capricious or otherwise not in accordance with law." (Gov't Def. Mem. at 18.)
I. ADMINISTRATIVE PROCEDURES ACT — 5 U.S.C. §§ 553-559
The Court's review of agency decisions is limited by the Administrative Procedures Act ("APA"). Plaintiffs bring this action under both Sections 706(2)(A) and 706(1). Review is deferential under either section, although the relief that is allowed to successful plaintiffs differs.
Section 706(2)(A) requires courts to overturn an agency decision that is arbitrary, capricious, an abuse of discretion, or contrary to law. 5 U.S.C. § 706(2)(A). The reviewing court is not to substitute its judgment for that of the agency, as long as the agency's determination is supported by any rational basis. Section 706(1) provides that "the reviewing court shall decide all relevant questions of law, interpret constitutional and statutory provisions, and determine the meaning or applicability of the terms of an agency action. The reviewing court shall . . . compel agency action unlawfully withheld or unreasonably delayed." 5 U.S.C. § 706(1) (emphasis added). See Mt. Emmons Mining Co. v. Babbitt, 117 F.3d 1167, 1170 (10th Cir. 1997). Agencies must be compelled to act where the agency fails to carry out a mandatory, nondiscretionary duty. Forest Guardians v. Babbitt, 174 F.3d 1178, 1187-88 (10th Cir. 1999). See also Cobell v. Norton, 240 F.3d 1081, 1096 (D.C. Cir. 2001) (noting that, as a rule, Section 706 leaves courts discretion to decide whether agency action (or inaction) is or is not unreasonable).
II. DOMBECK'S AUTHORIZATION AND REQUIREMENTS.
This Court owes no deference to the agency's interpretation of Dombeck. The USFS argues that Dombeck "redefined one of the assumptions the USFS had used to calculate `average actual annual' use for purposes of establishing the statutory cap. . . . [T]he Eighth Circuit ruling rendered the 1981 calculation of the statutory cap (and resulting quotas) invalid, because they were based on a different calculation of `average actual annual use.'" (Gov't Def. Mem. at 11.) Plaintiffs, ostensibly the "winners" in Dombeck, challenge the USFS's reaction to the case. Plaintiffs argue that the Eighth Circuit in Dombeck took the statutory cap as given, and did not direct any recalculation. (Plfs.' Opp'n Mem. at 5.)
Plaintiffs emphasize that their successful summary judgment on the "particular lake" issue entitled the Dombeck plaintiffs to the relief requested in their complaint, which was
Preliminary and permanent injunctive relief ordering the Secretary [of Agriculture] to amend the BWCAW Plan and all necessary supporting and implementing documents to provide that only motorboat use on Moose lake and Saganaga lake by homeowners and resort owners with properties on those particular lakes and their guests shall be considered exempt from motor quotas.
( Id. at 7.) Plaintiffs complain that it is "preposterous for the Forest Service to claim that its actions were compelled by Dombeck, when the Eighth Circuit's decision limited motorized use in the BWCAW while the Forest Service's actions increased motorized use by 400%." ( Id. at 7.)
Plaintiffs bolster their reading of Dombeck by analyzing another section of the opinion. The Eighth Circuit upheld the USFS special use permit system for commercial towboats, and in doing so the Circuit assumed the statutory cap was 10,539 motorboat trips. Dombeck, 164 F.3d at 1122. In fact, the Court specifically noted that "10,539 motorboat trips" is the "cap mandated by the BWCA Wilderness Act." Id.
There is no question that Dombeck is silent on the recalculation of the statutory cap, and the Court is not persuaded that the Dombeck Court contemplated such a recalculation. There is no indication that the Dombeck Court fathomed that the USFS would use its decision as justification to take the rather extraordinary step of recalculating a statutory cap that had been set for nearly two decades. The Dombeck Court expressly noted the cap of 10,539 in upholding the special permit system for towboats. If the Eighth Circuit merely intended to uphold the method for the original calculation, the opinion likely would have noted that the cap was subject to revision using the same methodology. The Eighth Circuit did not do so, and the parties do not indicate that the issue was identified for the Court.
In addition, the USFS urges a reading of Dombeck that results in the opinion being internally inconsistent. Specifically, the USFS invites the Court to read Dombeck as sub silentio overruling the statutory cap, even though Dombeck expressly relies on that cap for the "towboat" section of the opinion. The inconsistency is compounded when the USFS also asks the Court to read Dombeck as sub silentio approval of the methodology used by the USFS in setting the cap. Such a contorted reading of Dombeck is unpersuasive, and the opinion simply cannot be read as authorization for such a radical adjustment that is not otherwise authorized by Congress.
Similarly, the District Court's Order on Remand, dated September 3, 1999, did not contemplate or authorize a recalculation. As noted above, the District Court ordered that "the United States Forest Service will take the necessary steps to ensure that the BWCA Wilderness Management Plan and Implementation Schedule and supporting documents are consistent with the Eighth Circuit decision that property owners and their guests are exempt from motorboat quotas only on the lake immediately adjoining their property." (Order dated September 3, 1999 in Civil No. 5-95-10 (JMR).)
The USFS urges a reading of Dombeck that cannot be reconciled with the Eighth Circuit's language in that opinion. Neither the Eighth Circuit's opinion in Dombeck, nor the District Court's Order on remand contemplated a recalculation of the statutory cap. Absent authority to recalculate the statutory cap, to do so amounts to an arbitrary and capricious action that is an abuse of discretion and is not otherwise in accordance with law. 5 U.S.C. § 706(2)(A). The Court must therefore set aside the agency decision, and order the agency to take corrective action to comply with the Eighth Circuit's direction in Dombeck unless authority for the recalculation can be presumed from the language of the delegation in the BWCAW Act.
The Court is not dissuaded by the defendants' argument that plaintiffs' interpretation would render judicial review of agency action a nullity. Plaintiffs' argument is more fairly characterized as the suggestion that the USFS is misconstruing Dombeck — not that federal agencies are constrained from responding to judicial review. Understood this way, plaintiffs simply make the unremarkable argument that agency action cannot be contrary to judicial opinions on that issue.
III. OTHER AUTHORIZATION TO RECALCULATE STATUTORY CAP.
Agencies have only the authority delegated to them by Congress. Drake v. Honeywell, 797 F.2d 603, 607 (8th Cir. 1986). Congress directed the USFS to calculate, "as soon as practical entry point quotas for use of motorboats within the wilderness portions of [listed] lakes . . . Provided, That the quota established for any one year shall not exceed the average annual motorboat use of the calendar years 1976, 1977, and 1978 for each lake. . . ." BWCAW Act § 4(f). In 1981, the USFS complied with this congressional directive, and in the BWCAW Final Implementation Plan, calculated a statutory cap of 14,925 day use motorboat permits on the enumerated lakes. (A.R. Tab 53 at 1.) The BWCAW Act delegated to the USFS the authority to set motorboat quotas within this statutory cap. This discretion to set quotas, however, was subject to a limitation — specifically, the USFS was given no authority to exceed the statutory cap, once that cap was calculated. The Act also did not include any delegation of authority to the USFS to decide when the statutory cap should be recalculated. Similarly, the Act specifies no criteria for recalculation, or for determining when or under what circumstances a recalculation should occur. Agency action that exceeds statutory authority is arbitrary and capricious.
It is clear that absent Dombeck, the USFS would not have revisited the statutory cap. Indeed, the USFS conceded at oral argument that Dombeck alone provided the authority for the recalculation. There is no doubt that an agency would need a strong statement by the Courts to modify the otherwise clear intent of Congress. The USFS's analysis of Dombeck, however, is unavailing. As the above analysis indicates, it is simply untenable that Dombeck instructs the USFS to recalculate the cap.
Indeed, the USFS argues that any court's power to order a "recount" is severely limited — that is, the court can tell the USFS to do it again, but cannot tell the USFS how to do the recount. Though this might be something of an overstatement, the Court agrees that it lacks authority to prescribe the methodology for any recount. It is likely, however, that once a court finds a given methodology unreasonable, the USFS would hesitate before using that same methodology again.
IV. METHODOLOGY ("Average Actual Annual" Use)
Even if the Dombeck decision had authorized a recalculation, the Court would likely be required to set aside the recalculation because plaintiffs have established that the recalculation is arbitrary, capricious, and an abuse of discretion, or otherwise is not in accordance with the law.
The Eighth Circuit recently reiterated the appropriate judicial review of agency methodology:
If the administrative record contains evidence that supports the positions of both the agency and the party seeking relief, the agency is entitled to rely on its experts' tests and observations, and decisions made in such reliance are not arbitrary and capricious. Even if the agency's data is flawed, if the agency has relied on a number of findings and only some are erroneous, we must reverse and remand only if there is a significant chance that but for the errors the agency might have reached a different result. The question for us is not whether there might have been a better way for the agency to resolve the conflicting issues with which it was faced, but whether the agency's choice is a reasonable one.Cent. S.D. Coop. Grazing Dist. v. Sec'y of the U.S. Dep't of Ag., 266 F.3d 889, 898-99 (8th Cir. 2001) (internal quotations and citations omitted).
The USFS argues that it relied on the "best available data" to calculate average actual annual use. Plaintiffs disagree, and suggest that the methodology was so flawed as to render the USFS's result arbitrary and capricious.
A. Surveys
Critical to the "recalculation" was survey data obtained in a survey designed and conducted by USFS employees. Although the "surveys" were only one type of data on which the USFS claims to have relied, the USFS does not argue that the surveys were unimportant to the recalculation. Plaintiffs raise a strenuous challenge to the methodology of these surveys, and have provided an expert opinion that the surveys are unreliable and had an error rate of more than two standard deviations.
Despite the USFS's criticism of citation to cases discussing the admissibility of survey evidence, the inquiry here is not unrelated to that the Court would undertake in determining admissibility under the Rules of Evidence. The Rules are concerned with reliability, and it is that same concern underlying the Court's consideration of the survey in this instance. Though the level of reliability required for consideration might be less for an APA-style review than that required to admit a survey into evidence, the factors courts rely on to determine a survey's admissibility are a useful starting point for the Court's consideration of this survey.
Plaintiffs point to cases in which courts have refused to validate agency rules where the agency relied on unsound scientific data. St. James Hosp. v. Heckler, 760 F.2d 1460 (7th Cir. 1985); We Who Care, Inc. v. Sullivan, 756 F. Supp. 42 (D. Me. 1991). In We Who Care, the district court remanded the matter to the agency, because the record did not contain the survey on which the agency relied in making its determination. Id. at 46-47. The district court therefore determined that the decision was arbitrary and capricious given the lack of supporting evidence. Id.; see also Brown v. Sec'y of Health and Human Serv., 46 F.3d 102, 107 n. 7 (1st Cir. 1995) (implicitly approving of the We Who Care decision by noting that when the AFDC recipients first challenged the regulation, the Secretary was unable to produce a copy of the food stamp study, and the district court "[a]ccordingly . . . found the regulation arbitrary and capricious given the lack of supporting data in the record."). The We Who Care case is fairly cited for the proposition that it is appropriate for the Court to examine survey data to ensure that reliance on that information is reasonable.
Aid to Families with Dependent Children.
In defense of its reliance on its survey evidence, the USFS cites cases discussing the Health and Human Services Secretary's use of survey data in the decision regarding an AFDC automobile exemption. See Champion v. Shalala, 33 F.3d 963 (8th Cir. 1994); Brown v. Sec'y of Health Human Serv., 46 F.3d 102 (1st Cir. 1995). At issue in these cases was a $1,500 automobile-resource exemption set by the Secretary for AFDC recipients — exemption excluded from AFDC families who had more than $1,500 equity in an automobile. Champion, 33 F.3d at 965. Plaintiffs who would otherwise have been eligible for AFDC, but who had an ownership interest in one automobile in excess of $1500, challenged the cap. Plaintiffs particularly challenged the Secretary's reliance on a 1979 Food Stamp survey to determine the appropriate dollar amount of the automobile-resource exemption. The plaintiffs argued that the Food Stamp population was not relevant to the AFDC population (because families who qualified for one might not qualify for the other), and also argued that the cap was counter to the statutory goal of aiding recipients in achieving self-sufficiency.
Despite the USFS's valiant efforts, these cases favor the plaintiffs' position. The Food Stamp survey at issue in these decisions is readily distinguishable from the informal and unscientific "survey" relied on by the USFS. In the AFDC cases, the Eighth Circuit summarized the survey and why the Secretary's reliance on it was reasonable. Id. at 966. In particular, the court noted that the "survey collected data on household assets and incomes from a nationwide, statistically valid sample of 11,300 households." Id. (emphasis added). Although Food Stamp recipients were (and remain) a different population than AFDC recipients, the Eighth Circuit held that reliance on the nationwide, statistically valid survey was reasonable, in part because the undisputed record evidence showed a significant overlap between the two populations. Id. In addition, the court noted that the use of the Food Stamp data was actually helpful to AFDC recipients, because Food Stamp recipients who are not on AFDC tended to be more affluent than AFDC recipients. Id. at 966-67.
Unlike the formal survey in the AFDC cases, the results of which were reported to Congress, the informal survey here was designed and conducted by agency staff and not designed for external use. Though it is undisputed that the USFS has "discretion to rely on the reasonable opinions of its own qualified experts," Marsh v. Or. Natural Resources Council, 490 U.S. 360, 378 (1989), the presumption of agency expertise can be rebutted if the decision is not reasonable. See Defenders of Wildlife v. Babbitt, 958 F. Supp. 670, 679 (D.D.C. 1997); see also Greenpeace, Am. Oceans Campaign v. Nat'l Marine Fisheries Serv., 237 F. Supp. 2d 1181, 1188 (W.D. Wash. 2002) (noting that courts' deference is not unlimited, "and the presumption of expertise may be rebutted if the agency's decisions are based on science but are shown to be not reasonable").
The study was titled: Assets of Low Income Households: New Findings on Food Stamp Participants and Nonparticipants, Report to the Congress, January 1981, Food and Nutrition Service, U.S. Dep't of Agriculture. Cited in Brown, 46 F.3d at 106 n. 7.
In addition to the statistical differences between the sound Food Stamp study and the instant case, the AFDC cases are distinguishable because in those cases, there was an absolute delegation of authority to the agency, and the Secretary's decision to leave the cap at $1,500, rather than raise the cap, was explicitly considered by Congress and not modified. Here, the USFS's discretion is not absolute — the USFS is not free to set the quotas or statutory cap at any level the agency may prescribe. See 42 U.S.C. § 602(a)(7)(B)(i) (Supp. V 1993) (providing that "so much of the family member's ownership interest in one automobile as does not exceed such amount as the Secretary may prescribe.") (emphasis added). There is no indication here that the increase in the statutory cap has been brought to the attention of Congress, let alone considered and not modified by that body. Another critical difference is that in the AFDC cases, the Secretary's action was consistent with Congressional purpose of lowering the overall cost of the AFDC program. In contrast, here, the agency's action is at odds with at least one stated purpose of the BWCAW Act.
It may be true, as the USFS argues, that nothing requires the USFS to "hire and train professional interviewers, hire an expert professional statistician, [or] develop a method for drawing a random sample of residents from over 20 years ago." (Gov't Def. Mem. at 18.) However, once the USFS decided to conduct a survey, and to rely on the resulting data, the USFS was obligated to ensure that the survey produced at least reasonably reliable data. St. James Hosp., 760 F.2d at 1467 n. 5 ("[I]t is an agency's duty to establish the statistical validity of the evidence before it prior to reaching conclusions based on that evidence.").
In contrast to the survey in the AFDC cases, the USFS survey cannot be described as statistically valid. That survey relied on interviews with only five Moose Lake homeowners to determine 1978 motorboat use on Newfound and Sucker Lakes. One does not need to be a statistician to apprehend the numerous flaws in this sample. The use of only five respondents contrasts sharply to the survey approved in the AFDC cases in which a statistically significant 11,300 responses were gathered. Further, the survey asked homeowners to recall their use of lakes over twenty years ago, and the sample included only those individuals who had retained their property for at least twenty years. In the AFDC cases, the survey was of current Food Stamp recipients and did not ask individuals to recall their habits from decades past. In addition, the homeowners were aware of the purpose of the study, and had an interest in increasing their estimates of past use. These combined errors render the USFS data so flawed that reliance on the survey is unreasonable.
Simply put, the survey is unworthy of consideration in recalculating the statutory cap. Because the survey data was instrumental to the recalculation, it is clear that "there is a significant chance that but for the errors the agency might have reached a different result." Cent. S.D. Coop. Grazing Dist., 266 F.3d at 898-99. This survey did not result in reasonably reliable data and reliance on the survey is not reasonable. See Menorah Med. Ctr. v. Heckler, 768 F.2d 292, 295-96 (8th Cir. 1985) (adopting St. James Hosp. v. Heckler, 760 F.2d 1460 (7th Cir. 1985) (concluding that the at-issue rule was arbitrary and capricious because the Secretary based the rule on a faulty study and failed to consider critical aspects of the problem)).
B. Other Flawed Data
The USFS argues that its consideration of numerous sources of information insulates its result from attack. Regardless of how many sources of information are consulted, if the source itself is flawed, as plaintiffs indicate many are here, the result cannot be sound. Because the Court found the "survey data" fatally undermines the recalculation, the Court will not discuss at length the plaintiffs' additional specific complaints. The Court does note, however, that some issues plaintiffs have identified are more troublesome than others. The Court is especially concerned with the USFS's reliance on assumptions that seem to lack support in the lengthy administrative record. For example, plaintiffs suggest that the USFS assumed that all homeowners on the Moose and Saganaga Lake Chains owned motorboats during the relevant period, but plaintiffs point to record evidence which seems to contradict this assumption. (A.R. Tabs C, K.) The Court also questions whether the USFS had a reasonable basis on which to assume that home/resort owners and guests always traveled outside the particular lake their property abuts, and whether it was reasonable to assume common use patterns for the different lake chains.
Plaintiffs argue, in essence, "garbage in, garbage out." The computer axiom is commonly understood as shorthand to express that if invalid data is entered into a computer system, the resulting output will also be invalid. Although originally applied to computer software, the axiom applies to many decision-based systems, such as that of the USFS here.
Though the Court notes its concern, the Court also is mindful of the deference allotted to agency discretion. Absent the unreasonable reliance on unsound survey data, these concerns might not be enough to deem the agency's action arbitrary, capricious, or otherwise contrary to law.
V. CONCLUSION
Agency action must be based in authority delegated to the agency by Congress. Absent such a delegation, agency action is invalid. In this instance, Congress did not delegate to the USFS the authority to recalculate the statutory cap. Similarly, the Dombeck decision did not contemplate such a recalculation. Even if Dombeck fairly could be read as providing such authority, the recalculation, attempted twenty years later, was based on flawed data, and in some cases, an absence of data of "actual" use. Therefore plaintiffs are entitled to summary judgment. The USFS shall take the necessary steps to ensure that the BWCA Wilderness Management Plan and Implementation Schedule and supporting documents are consistent with the Eighth Circuit opinion that property owners and their guests are exempt from motorboat quotas only on the lake immediately adjoining their property, without using the arbitrary and capricious recalculation of the statutory cap.
ORDER
Based on the foregoing, all the records, files, and proceedings herein, IT IS HEREBY ORDERED that:
1. Plaintiffs' Motion for Summary Judgment [Docket No. 34] is GRANTED.
2. Defendants Bosworth and Venemen's Motion for Summary Judgment [Docket No. 26] is DENIED.
3. Intervenor-Defendants' Motion for Summary Judgment [Docket No. 30] is DENIED.
LET JUDGMENT BE ENTERED ACCORDINGLY.