Florida Manufactured Housing Association, Inc., Homes Ofmerit, Inc., Jacobsen Manufacturing, Inc., Andnobility Homes, Inc., Petitioners, v. Henry G. Cisneros, Secretary of the United States Departmentof Housing and Urban Development, Respondent, 53 F.3d 1565 (11th Cir. 1995)

Mark S. Weiss, Roberts & Weiss, Woodbridge, VA, for amicus Ass'n for Regulatory Reform.

Robert E. Sasser, James D. Hamlett, Sasser & Littleton, Montgomery, AL, for amicus Alabama Manufactured Housing Institute, Inc.

William A. Pyle, Pyle, Dreher, Mills & Dye, P.A., Ridgeland, MS, for amicus Mississippi Manufactured Housing Ass'n.

Dick W. Wilson, Jr., Peterson, Dillard, Young, Asselin, Powell & Wilson, Atlanta, GA, for amicus Georgia Manufactured Housing Ass'n, Inc.

Petition for Review of an Order of the U.S. Department of Housing and Urban Development.

Before ANDERSON and CARNES, Circuit Judges, and RONEY, Senior Circuit Judge.

CARNES, Circuit Judge:

This case involves regulations promulgated under the National Manufactured Housing Construction and Safety Standards Act of 1974, 42 U.S.C. §§ 5401-5426 ("the Manufactured Housing Act" or "the Act"). After the tremendous damage caused by Hurricane Andrew in 1992, the Department of Housing and Urban Development ("HUD"), pursuant to the Act, revised its wind resistance standards for manufactured homes. The Florida Manufactured Housing Association, Inc. and the other petitioners in this case (referred to collectively as "the manufacturers") challenge several aspects of that rulemaking. They contend (1) that the agency did not adequately consult with its Advisory Council as required by the Act; (2) that the agency misinterpreted the meaning of "cost" as used in the statutory criteria; and (3) that the new wind standards are arbitrary and capricious for four different reasons. We hold that the manufacturers' arguments are without merit and deny their request that we set aside the regulations and that we remand to the agency for further proceedings.

As declared by Congress, the purposes of the Manufactured Housing Act are to "reduce the number of personal injuries and deaths and the amount of insurance costs and property damage resulting from manufactured home accidents and to improve the quality and durability of manufactured homes." 42 U.S.C.A. Sec. 5401 (West 1983). In order to achieve these objectives, the Act authorizes the Secretary of HUD to "establish by order appropriate Federal manufactured home construction and safety standards. Each such Federal manufactured home standard shall be reasonable and shall meet the highest standards of protection, taking into account existing State and local laws relating to manufactured home safety and construction." 42 U.S.C.A. Sec. 5403(a) (West 1983). The construction and safety standards issued under the Act supersede state and local standards for manufactured homes. 42 U.S.C.A. Sec. 5403(d) (West 1983); Scurlock v. City of Lynn Haven, Fla., 858 F.2d 1521, 1524-25 (11th Cir. 1988). The regulations are issued under the rulemaking procedures mandated by the Administrative Procedure Act, 5 U.S.C. § 553. See 42 U.S.C.A. Sec. 5403(b) (West 1983).

When promulgating manufactured housing standards, the Act directs the Secretary of HUD to:

(1) consider relevant available manufactured home construction and safety data, including the results of the research, development, testing, and evaluation activities conducted pursuant to this chapter, and those activities conducted by private organizations and other governmental agencies to determine how to best protect the public;

(2) consult with such State or interstate agencies (including legislative committees) as he deems appropriate;

(3) consider whether any such proposed standard is reasonable for the particular type of manufactured home or for the geographic region for which it is prescribed;

(4) consider the probable effect of such standard on the cost of the manufactured home to the public; and

(5) consider the extent to which any such standard will contribute to carrying out the purposes of this chapter.

42 U.S.C.A. Sec. 5403(f) (West 1983). Pursuant to this statutory authority, HUD has promulgated standards covering "all equipment and installations in the design, construction, transportation, fire safety, plumbing, heat-producing and electrical systems of manufactured homes which are designed to be used as dwelling units." 24 C.F.R. Sec. 3280.1 (1994).

Before HUD issued the regulations challenged in this appeal, the agency's wind resistance standards for manufactured homes divided the United States into two wind zones: a "standard" wind zone and a "hurricane-resistive" wind zone. Under those regulations, manufactured homes in the hurricane-resistive zone were required to withstand winds of approximately 80 miles per hour. The devastation caused by Hurricane Andrew in August of 1992 convinced HUD that the existing standards were inadequate. According to HUD, 97% of all manufactured homes in Dade County were destroyed, compared with 11% of single-family, non-manufactured homes. In Florida and Louisiana, 11,213 manufactured houses were destroyed and 3,016 suffered major damage. Approximately 36% of all housing units destroyed by Hurricane Andrew were manufactured homes. Moreover, the wind turned parts of some manufactured homes into flying missiles, causing additional damage to other structures. Manufactured Home Construction and Safety Standards on Wind Standards, 59 Fed.Reg. 2456, 2457 (1994) ("Final Rule"). Damage to manufactured homes from high winds is "primarily in the form of roof failure, loss of roof diaphragm material, connection failures, and tiedown/foundation failures." Id.

After Hurricane Andrew, HUD initiated field investigations in Florida as part of a general review of manufactured home standards. The investigations revealed various deficiencies in manufactured homes' resistance to wind storms, such as inadequate connections between roofs, walls, and floors, and between exterior roof and wall coverings and supporting sheathing or framing. Those deficiencies led to other problems, including water damage, damage from increased internal pressures, and missile damage to other structures.

In addition to HUD's investigations, the National Institute of Standards and Technology issued a report ("National Standards Report") comparing the wind protection provisions of selected codes and standards. In assessing the impact of Hurricane Andrew on manufactured housing, the National Standards Report found that:

Damage to manufactured homes ranged from loss of roofing to total destruction.... Commonly observed failures include loss of roof membranes and blow-off of roof sheathing, failure of uplift straps at truss-to-wall connections where staple crowns pulled through the strap material, loss of cladding on endwalls and near corners where large negative (suction) pressures develop, loss of add-ons with resulting missile damage and damage to the parent unit at points of attachment, complete separation of superstructure from floor and underframe, and loss of the complete unit due to failure of tiedown straps or withdrawal of soil anchors.

The National Standards Report recommended that HUD use the wind load requirements of the American Society of Civil Engineers' model standard, ASCE 7-88, as the basis for the new rules.

On April 14, 1993, HUD published a Notice of Proposed Rulemaking, proposing that the manufactured home standards be amended to conform with the ASCE 7-88 standard, 58 Fed.Reg. 19,536 (1993) ("Notice of Rulemaking"), which is what the National Standards Report had recommended. The stated goal of the proposed standards was to "increase the safety of manufactured homes in areas where wind-induced damage is a special hazard." Id. at 19,536. In order to have the new standards in place by the 1993 hurricane season, the Notice of Rulemaking indicated that HUD would use an abbreviated thirty-day public comment period for the proposed rules. However, after that deadline passed, HUD subsequently extended the comment period for another thirty days, because the agency decided that the new standards could not be implemented in time for the 1993 hurricane season. HUD received over one thousand comments from the public, although most were duplicative or identical form letters. Comments submitted by the manufacturers, as well as others, raised a number of objections to the proposed standards, including criticisms of the increased costs and predictions of deleterious effects on the industry and on low-income families.

The Manufactured Housing Act also requires that HUD, before establishing, amending, or revoking any manufactured housing standard, must consult "to the extent feasible" with the National Manufactured Home Advisory Council ("Advisory Council"). 42 U.S.C.A. Sec. 5404(b) (West 1983). At first, HUD determined that it was not feasible to consult with the Advisory Council because of the agency's expedited schedule; however, after it extended the comment period, HUD did consult with the Advisory Council, which it convened for a two-day session in July of 1993. At that meeting, the Advisory Council adopted a resolution calling for more studies, and more public input, and recommending that the Council "be reconvened to review the public comments and the analysis and studies by HUD" that the Council called for. HUD did not reconsult the Advisory Council before the Final Rule was published.

The Final Rule was published in the Federal Register on January 14, 1994. Under the Final Rule regulations, the United States is divided into three zones, classified according to their susceptibility to hurricanes and high winds. Wind Zone III includes parts of the coast in Alaska, southern Florida, Louisiana, and North Carolina; all of Hawaii; and various United States territories. Wind Zone II covers selected areas in Alabama, Florida, Georgia, Louisiana, Maine, Massachusetts, Mississippi, North Carolina, South Carolina, Texas, and Virginia. Wind Zone I is all parts of the country not in Wind Zones II or III. Final Rule, 59 Fed.Reg. at 2470-72. In general terms, manufactured homes in Wind Zone III must withstand a "design wind speed" of 110 miles per hour, and Wind Zone II manufactured homes must withstand a design wind speed of 100 miles per hour. HUD decided not to change the standards for Wind Zone I; thus, manufactured homes designated for parts of the country located in that zone are only required to comply with the previous wind standards governing those parts of the country. In addition, the regulations set forth specific technical requirements manufactured homes must meet in order to comply with the ASCE 7-88 criteria.

The Final Rule stated that, in formulating the new wind standards, HUD had balanced the competing goals of improving safety and retaining manufactured homes as a viable source of low-cost housing. HUD acknowledged that the costs to consumers in Wind Zones II and III would rise, but found that the price increases were justified because of reductions in future losses to consumers and the public, as well as reductions in the "inestimable costs of devastation to people's lives and emotional health and to the communities" in hurricane-prone areas. Final Rule, 59 Fed.Reg. at 2457-58. HUD also stated that some of the benefits and costs (such as deaths, injuries, and uninsured costs) were difficult to quantify and that "its statutory mandate to reduce the number of personal injuries and deaths and the amount of insurance costs and property damage resulting from manufactured home accidents, and to improve the quality and durability of manufactured homes, requires that the Department look beyond affordability issues. In promoting homeownership opportunities for lower-income persons, the Department strongly believes that such housing must also be safe." Id. at 2461-62.

HUD also explained in the Final Rule that the agency had considered the comments of the Advisory Council, and did not consider it necessary to reconvene the council. Id. at 2456, 2461. According to HUD, it modified several aspects of the proposed regulations in response to Advisory Council recommendations. HUD decided not to impose any new requirements in Wind Zone I, and it reduced the size of Wind Zone II by changing the high wind speed boundary between Wind Zone I and Wind Zone II from 80 miles per hour to 90 miles per hour, as well as making several other changes. Additional modifications HUD made in response to the Advisory Council's recommendations are discussed on pages 1575-76, below.

On March 14, 1994, the manufacturers, who were still dissatisfied with the scope of the new wind standards, filed a petition for review with this Court, as they are entitled to do under the Act. See 42 U.S.C.A. Sec. 5405(a) (1) (West 1983). At the same time, the manufacturers submitted to HUD an Application for Stay of Effective Date of Rule Amendments Pending Judicial Review ("Application for Stay"). The Secretary of HUD denied the Application for Stay on April 1, 1994, and the manufacturers subsequently filed in this Court a motion to stay the new regulations pending judicial review. That motion was denied on April 28, 1994.

The manufacturers' challenges implicate both HUD's interpretation of the Act and the procedural propriety of the agency's rulemaking. The different issues involve different standards of review:

The standard of review for an agency's interpretation of a statute is governed by the two-prong test outlined in Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 104 S. Ct. 2778, 81 L. Ed. 2d 694 (1984). First, the Court must determine whether Congress has directly and unambiguously spoken to the issue at hand. If so, that is the end of our inquiry and we must give effect to the expressed intent of the legislature. Id. at 842-43, 104 S. Ct. at 2781.

However, if Congress has not directly addressed the issue, we then proceed to the second prong of Chevron. At this stage, "the court does not simply impose its own construction on the statute, as would be necessary in the absence of an administrative interpretation." Id. at 843, 104 S. Ct. at 2782 (footnote omitted). As the Supreme Court and this Court have recognized, " 'the resolution of ambiguity in a statutory text is often more a question of policy than of law.' " Georgia, Dep't of Medical Assistance v. Shalala, 8 F.3d 1565, 1568 (11th Cir. 1993) (quoting Pauley v. Bethenergy Mines, Inc., 501 U.S. 680, 696, 111 S. Ct. 2524, 2534, 115 L. Ed. 2d 604 (1991) (citations omitted)). For that reason, " [t]he court need not conclude that the agency construction was the only one it permissibly could have adopted to uphold the construction, or even the reading the court would have reached if the question initially had arisen in a judicial proceeding." Chevron, 467 U.S. at 843 n. 11, 104 S. Ct. at 2782 n. 11; see Jaramillo v. INS, 1 F.3d 1149, 1152-53 (11th Cir. 1993); Lipscomb v. United States, 906 F.2d 545, 548 (11th Cir. 1990). Instead, the Court must defer to the agency's construction if it is reasonable. The agency's construction is reasonable if it is not "arbitrary, capricious, or manifestly contrary to the statute." Chevron, 467 U.S. at 844-45, 104 S. Ct. at 2782-83; see Alabama Power Co. v. Federal Energy Regulatory Comm'n, 22 F.3d 270, 272 (11th Cir. 1994).

Under the Manufactured Housing Act, HUD issues its regulations through the informal rulemaking procedures of Sec. 553 of the Administrative Procedure Act ("APA"), and a court reviews HUD's orders in accordance with Secs. 701-706 of the APA. See 42 U.S.C.A. Secs. 5403(b), 5405(a) (3) (West 1983). The APA provides that an agency action promulgated under the informal rulemaking procedures may be set aside if it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C.A. Sec. 706(2) (A) (West 1977). This standard of review is "highly deferential," Hussion v. Madigan, 950 F.2d 1546, 1553 (11th Cir. 1992) (citation and internal quotation marks omitted), and "presumes the validity of the agency action," Charter Fed. Sav. and Loan Ass'n, West Point, Ga. v. Office of Thrift Supervision, 912 F.2d 1569, 1580 (11th Cir. 1990) (citations and internal quotation marks omitted). The Supreme Court has set forth the factors relevant to this review:

Normally, an agency rule would be arbitrary and capricious if the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.

Motor Vehicle Mfrs. Ass'n of the United States, Inc. v. State Farm Mutual Auto Ins. Co., 463 U.S. 29, 43, 103 S. Ct. 2856, 2867, 77 L. Ed. 2d 443 (1983).

"Along the standard of review continuum, the arbitrary and capricious standard gives an appellate court the least latitude in finding grounds for reversal." North Buckhead Civic Ass'n v. Skinner, 903 F.2d 1533, 1538-39 (11th Cir. 1990) (footnote omitted). We are limited to "a determination of whether the agency has considered the relevant factors and articulated a rational connection between the facts found and the choices made." Charter Fed. Sav. and Loan Ass'n, 912 F.2d at 1580 (citation and internal quotation marks omitted). When the agency is confronted with opposing views among specialists, it must be given the discretion to rely on the reasonable opinions of its own experts, even if a court finds other views more persuasive. Marsh v. Oregon Natural Resources Council, 490 U.S. 360, 378, 109 S. Ct. 1851, 1861, 104 L. Ed. 2d 377 (1989); Hussion, 950 F.2d at 1554. Thus, " ' [a]dministrative decisions should be set aside in this context ... only for substantial procedural or substantive reasons as mandated by statute ... not simply because the court is unhappy with the result reached.' " North Buckhead Civic Ass'n, 903 F.2d at 1539 (quoting Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, Inc., 435 U.S. 519, 558, 98 S. Ct. 1197, 1219, 55 L. Ed. 2d 460 (1978)).

In addition to arguing that the new standards are arbitrary and capricious, the manufacturers contend alternatively that the closer scrutiny "substantial evidence" test should be applied in this case. They point out that the statutory provision for judicial review in the Manufactured Housing Act is derived from the National Traffic and Motor Vehicle Safety Act of 1966, 15 U.S.C. § 1381 et seq. In reviewing the Motor Vehicle Safety Act, the Supreme Court in State Farm seemed to suggest that the substantial evidence test, as well as the arbitrary and capricious standard, might be applicable to the review of agency findings. State Farm, 463 U.S. at 43-44, 103 S. Ct. at 2867.

We are unpersuaded by the manufacturers' argument. The Supreme Court in State Farm referred to the substantial evidence standard only after citing legislative history from the Motor Vehicle Safety Act that explicitly stated Congress's intent that that standard of review be used. Id. The Court has repeatedly held that unless an agency's organic statute contains a specific provision to the contrary, the substantial evidence standard is used only to review formal "on the record" agency actions, not those resulting from the informal rulemaking procedures of Sec. 553 of the APA which are incorporated into the Manufactured Housing Act. See American Paper Inst., Inc. v. American Elec. Power Serv. Corp., 461 U.S. 402, 412 n. 7, 103 S. Ct. 1921, 1927 n. 7, 76 L. Ed. 2d 22 (1983); FCC v. National Citizens Comm. For Broadcasting, 436 U.S. 775, 802-03, 98 S. Ct. 2096, 2116, 56 L. Ed. 2d 697 (1978); see also 5 U.S.C. § 706(2) (E); Phillips Petroleum Co. v. Federal Energy Regulatory Comm'n, 786 F.2d 370, 373-74 (10th Cir.), cert. denied, 479 U.S. 823, 107 S. Ct. 92, 93 L. Ed. 2d 44 (1986); Western Union Tel. Co. v. FCC, 665 F.2d 1126, 1148 n. 45 (D.C. Cir. 1981); cf. Aqua Slide 'N' Dive Corp. v. Consumer Prod. Safety Comm'n, 569 F.2d 831, 837 (5th Cir. 1978) (applying substantial evidence review to informal rulemaking because it was required by the agency's statute). Therefore, we will use the arbitrary and capricious standard set forth in Sec. 706 of the APA in our review of HUD's rulemaking in this case.

Under the Act, a petition for judicial review must be filed "prior to the sixtieth day after such order is issued." 42 U.S.C.A. Sec. 5405(a) (1) (West 1983) (emphasis added). The Final Rule was published in the Federal Register on January 14, 1994; however, the last page of the Final Rule indicated that the document itself was "dated" January 4, 1994. The petition for judicial review was filed with this Court on March 14, 1994. Thus, the petition was timely filed if the date the regulations were "issued" is the date of publication in the Federal Register, but not if it is the date placed on the Final Rule itself.

HUD argues that the rule was issued when the rulemaking decision was dated, not when it was published. The agency cites the Federal Register Act to support its contention that there is a distinction between the date of publication and issuance. See 44 U.S.C.A. Sec. 1503 (West 1991) (discussing special provision for filing with the Office of the Federal Register " [w]hen the original is issued, prescribed, or promulgated outside the District of Columbia"). HUD also points to other statutory schemes to demonstrate that when Congress wants to provide that the date of publication initiates a filing period, it has expressly done so; since Congress did not do so here, HUD argues that the date of publication is not the date the Final Rule was "issued."

HUD's argument, however, is contradicted both by its own prior interpretations of the statutory term "issued" and by the plain meaning of that term. The Act authorizes a range of dates for the time an order amending a standard can take effect. The effective date "shall not be sooner than one hundred and eighty days or later than one year from the date the order is issued," unless the Secretary publishes his finding that a departure is in the public interest. 42 U.S.C.A. Sec. 5403(e) (West Supp.1995) (emphasis added). In its Notice of Rulemaking in this very case, HUD equated the issuance date, which starts the one-year period running, with the publication date, stating: "Under ... [42 U.S.C. § 5403], standards are to become effective not sooner than 180 days following publication of a final rule in the Federal Register." 58 Fed.Reg. at 19,536 (emphasis added). Moreover, the Final Rule provides that some standards are to become effective on July 13, 1994 and others on January 17, 1995. 59 Fed.Reg. at 2456. These dates correspond to 180 days and one year, respectively, after the date of publication, not the date placed on the Final Rule.¹  As the manufacturers note, the latter effective date would be prohibited by the Act (without publication of a finding of good cause) if HUD's argument on appeal were adopted, because it would be "later than one year" after the date on the Final Rule.

Even assuming that we otherwise would give deference to an agency's interpretation of a statute concerning the jurisdiction of federal courts,²  we will not do so in this case. We do not defer to an agency's post hoc "convenient litigating position" that is wholly unsupported by prior regulations, interpretations, rulings, or administrative practices. Bowen v. Georgetown Univ. Hosp., 488 U.S. 204, 212-13, 109 S. Ct. 468, 473-74, 102 L. Ed. 2d 493 (1988); accord Alabama Power Co., 22 F.3d at 273; USX Corp. v. Director, Office of Workers' Compensation Programs, 978 F.2d 656, 658 (11th Cir. 1992); McKee v. Sullivan, 903 F.2d 1436, 1438-39 n. 3 (11th Cir. 1990). In addition, although an agency's interpretation may receive some deference even if it has changed over time, the consistency of its interpretation is an important factor in determining the amount of deference owed. See Thomas Jefferson Univ. v. Shalala, --- U.S. ----, ---- - ----, 114 S. Ct. 2381, 2388-89, 129 L. Ed. 2d 405 (1994); Good Samaritan Hosp. v. Shalala, --- U.S. ----, ----, 113 S. Ct. 2151, 2161, 124 L. Ed. 2d 368 (1993); INS v. Cardoza-Fonseca, 480 U.S. 421, 446 n. 30, 107 S. Ct. 1207, 1221 n. 30, 94 L. Ed. 2d 434 (1987); Georgia, Dep't of Medical Assistance, 8 F.3d at 1568. Therefore, even if we would otherwise give deference to HUD's interpretation of the term "issued" in the statutory provision that determines this Court's jurisdiction, the agency's change of position within the same rulemaking process, seemingly for the sole purpose of triggering a jurisdictional challenge, convinces us to give the agency's latest interpretation no deference.

Moreover, HUD's latest interpretation contravenes the plain meaning of the term "issued." The verb "issue" clearly refers to an act of public announcement and not to the act of arriving at a private decision within the agency. See Random House Unabridged Dictionary 1015 (2d ed. 1993) (defining the transitive verb form of this term in the first two entries as "to put out"; "deliver for use, sale"; "put into circulation"; and "to mint, print, or publish for sale or distribution"). Giving the term its plain meaning also comports with common sense and avoids a result that Congress could not have intended. Under the interpretation of "issued" now urged by HUD, an agency would have the power to manipulate the jurisdiction of the federal courts. In this case, there was only a ten-day delay between the date of the internal decision (January 4), and the date the Final Rule was published (January 14). However, if HUD's interpretation were adopted, the agency conceivably could release its final rule to the public thirty, forty, fifty, or more days after the stated date of decision and thereby impede or prevent any judicial review. As a matter of fairness, the sixty-day filing period should not begin to run until the public has notice of the final rule's content. Cf. Northwest Tissue Ctr. v. Shalala, 1 F.3d 522, 530 n. 8 (7th Cir. 1993) ("Before any litigant reasonably can be expected to present a petition for review of an agency rule, he first must be put on fair notice that the rule in question is applicable to him."); RCA Global Communications, Inc. v. FCC, 758 F.2d 722, 730 (D.C. Cir. 1985) ("Although statutory time limitations on judicial review of agency action are jurisdictional, self-evidently the calendar does not run until the agency has decided a question in a manner that reasonably puts aggrieved parties on notice of the rule's content." (citation omitted)). For all these reasons, we hold that the petition for review was timely filed so that this Court does have jurisdiction under the Act to decide it.

The Manufactured Housing Act requires that HUD consult with the Advisory Council "to the extent feasible" when amending manufactured home standards. 42 U.S.C.A. Sec. 5404(b) (West 1983).³  In anticipation of having the rules in effect before the next hurricane season, HUD initially took the position that it was not "feasible" for it to meet with the Advisory Council; however, after the expedited schedule was relaxed, the agency convened the Advisory Council for a two-day session in July 1993. The Advisory Council adopted a resolution, recommending among other things that further studies be conducted, and that the Advisory Council be reconvened in the future to review public comments about HUD's analysis of the further studies the Council had recommended. HUD published the Final Rule in January 1994 without reconvening the Advisory Council. HUD explained that it had "considered the conclusions and recommendations of the Advisory Council" and did "not believe there is any significant advantage in, or that the public interest would be served by, reconvening the Advisory Council." Final Rule, 59 Fed.Reg. at 2456.

The manufacturers claim that HUD erred in interpreting the provision for consultation "to the extent feasible" to permit it to refuse to convene the Advisory Council for a second time, particularly since six months had elapsed between the consultation in July 1993 and the January 1994 publication of the Final Rule. The manufacturers argue that the failure to reconvene the Council denied it an opportunity to comment on the modified final standards and on the data on which those standards were based. Under these circumstances, they contend, HUD's failure to request further consultation was an error of law. HUD responds that it complied with the Act by convening the Advisory Council once and by carefully considering its comments. Although it disagreed with many of the Advisory Council's recommendations, the agency did make some modifications in its proposed standards based in part on the Council's recommendations. In addition to withdrawing the proposed regulations for Wind Zone I and expanding the size of that zone, the following five requirements that had been included in the proposed rules were left out of the final standards:

(1) Maximum dimension of 12"' for roof overhangs;

(2) Lower load duration factor than provided in the 1991 National Design Specification for Wood Construction (NDS);

(3) Requirement for a 1.5 safety factor to calculate a resistance of anchoring and foundation systems to higher design forces in Wind Zones II and III;

(4) Manufacturer's design and details for a permanent foundations system (certified by a registered professional engineer or architect) applicable to each manufactured home design; and

(5) Shortened period for implementation of the standards after publication.

Final Rule, 59 Fed.Reg. at 2457. In view of the changes it made, HUD asserts that the manufacturers "have confused a requirement of consultation with one of 100% adoption."Congress did not indicate exactly what it meant when it required HUD to consult with the Advisory Council "to the extent feasible." Thus, we cannot decide this statutory interpretation question under the first prong of Chevron. Instead, we move on to the second prong, which requires us to determine whether HUD's single consultation with the Advisory Council under the circumstances of this rulemaking process comported with a reasonable interpretation of the requirements of 42 U.S.C. § 5404(b). We hold that it did.

As we have discussed, based upon its consultation with the Advisory Council, HUD did make some modifications to the proposed standards, with the result that they became less stringent. The Advisory Council wanted more modifications, more studies, and more consultation, but HUD is the decisionmaker and the process must end sometime. If we were to require reconsultation whenever the Advisory Council demanded it, the process might never end. Further delay in this instance would have resulted in the continuation of standards that even the manufacturers concede were inadequate and needed revision. The manufacturers' brief states that they "have never disputed the need for some strengthening of manufactured home standards to conform with those generally accepted for site-built housing ...." Although HUD abandoned its goal of having the rules in place by the 1993 hurricane season, there was still a vital interest in not postponing the effective date of the amended rules too far into the future. Public safety was involved, and it is no exaggeration to say that too much delay could have resulted in the loss of life. We hold that it was reasonable for the agency to interpret the feasibility language in Sec. 5404(b) as not requiring it to reconvene the Advisory Council under these circumstances.

The manufacturers rely on National Constructors Ass'n v. Marshal, 581 F.2d 960 (D.C. Cir. 1978), but that case is distinguishable. In it, an order by the Secretary of Labor was remanded, because the department had issued its rule without adequately consulting with an advisory committee as required by statute and regulation. Id. at 971-72. However, in that case the operative language directed that the agency "shall consult," without limiting that duty "to the extent feasible," id. at 964 n. 4, 967, which is a significant qualifying phrase in the statutory language before us. Moreover, in Marshal, the advisory committee was consulted on one proposed standard, but was not consulted at all about a subsequently developed standard that was "fundamentally differ [ent]" from the first. Id. at 970 n. 27. Because the second standard was not a "logical outgrowth" of the proposal originally presented to the committee, the Marshal court held that the mandatory consultation requirement had not been met. Id. at 970-71 & n. 27. In this case, we do not believe that the final standards were so substantively different that they cannot be considered a "logical outgrowth" of the proposed standards upon which the Advisory Committee was consulted.⁴  This conclusion is bolstered by the fact that the final rules were less, not more, rigorous than the proposed rules. The statutory requirements and rulemaking history in this case are simply not analogous to those involved in the Marshal decision.

The statute requires HUD to "consult" with the Advisory Council, and the agency did so. Even if HUD could have reconvened the Advisory Council in the six months between the Council's only meeting and publication of the Final Rule, under the facts of this case, it was reasonable not to do so. We therefore conclude that the agency's determination of non-feasibility under the facts of this case was not "arbitrary, capricious, or manifestly contrary to the statute." Chevron, 467 U.S. at 844, 104 S. Ct. at 2782.B. THE MEANING OF "COST"

As discussed on p. 1569 above, the Manufactured Housing Act, 42 U.S.C. § 5403(f), sets out five specific requirements for HUD to follow when establishing manufactured housing standards. The fourth such criterion directs HUD to "consider the probable effect of such standard [s] on the cost of the manufactured home to the public." 42 U.S.C.A. Sec. 5403(f) (West 1983). The manufacturers argue that this factor refers solely to the consumer purchase price of manufactured homes and that HUD has misinterpreted the meaning of "cost" by giving it a much broader definition. The manufacturers point to a statement made by HUD in its denial of the manufacturers' Application for Stay in which the agency stated that the meaning of cost is "not limited to [a manufactured home's] purchase price but includes the potential costs to the public if the home is involved in a disaster."

According to the manufacturers, this interpretation is contrary to the plain meaning of "cost" and thus the standards were not promulgated in accordance with the law. Instead of simply considering consumer costs, they contend, HUD merged the consumer cost factor into a cost-benefit analysis and concluded that the benefits to society outweighed any price increases. The manufacturers argue that when the price increases were subsumed into a more general weighing of societal costs and benefits, the true impact of price increases was minimized because they were offset by general benefits to the government and public at large, as distinguished from the "public" referred to in Sec. 5403(f) (4), which the manufacturers contend means only those members of the general public who live in manufactured homes. Thus, they argue, HUD's interpretation denied consumer price increases their proper consideration as an independent factor.

We need not determine the meaning of "cost" or "public" in Sec. 5403(f) (4), because we find that HUD has complied with the Act even under the limited meaning the manufacturers would give those two terms. The manufacturers acknowledge that HUD has considered the effect of the regulations on consumer purchase prices; in fact, they contest HUD's purchase price impact figures in a separate argument. See infra at Part IV.C.1. The crux of their claim is that HUD has erred by merging cost into a general cost-benefit analysis, instead of considering it as a separate, independent criterion. Although this claim may have some semantic appeal, there is no statutory support for it. The Act requires that the agency "consider" cost, which the manufacturers concede was done, but the Act does not indicate precisely how HUD is to consider this factor, or how much weight the agency should give cost in weighing it against other factors. When it prescribed the factors HUD should consider, Congress did not establish a strict algebraic formula in which the agency simply plugs in the numbers, as the manufacturers seem to suggest. As this Court recently stated: "we decline the ... invitation to require an agency to accord greater weight to aspects of a policy question than the agency's enabling statute itself assigns to those considerations." Hussion, 950 F.2d at 1554. As long as the agency gives fair consideration to the relevant factors mandated by law, the importance and weight to be ascribed to those factors is the type of judgment that courts are not in a position to make. Instead, that judgment is for the agency:

While agencies are not directly accountable to the people, the Chief Executive is, and it is entirely appropriate for this political branch of the Government to make such policy choices--resolving the competing interests which Congress itself either inadvertently did not resolve, or intentionally left to be resolved by the agency charged with the administration of the statute in light of everyday realities.

When a challenge to an agency construction of a statutory provision, fairly conceptualized, really centers on the wisdom of the agency's policy, rather than whether it is a reasonable choice within a gap left open by Congress, the challenge must fail. In such a case, federal judges--who have no constituency--have a duty to respect legitimate policy choices made by those who do.

Chevron, 467 U.S. at 865-66, 104 S. Ct. at 2793.

Moreover, the fifth requirement under Sec. 5403(f) mandates that the agency consider "the purposes of this chapter," which are defined as reducing injuries, deaths, insurance costs, and property damages, see 42 U.S.C.A. Sec. 5401 (West 1983). As HUD points out, even if the definition of "cost" under Sec. 5403(f) (4) does not include a consideration of general societal costs other than purchase price increases, Sec. 5403(f) (5) certainly does. Not only is it permissible for HUD to balance the cost to the general public of wind damage to manufactured housing against price increases, Sec. 5403 requires the agency to do so. Whether the agency does so under its consideration of Sec. 5403(f) (4) or Sec. 5403(f) (5) is immaterial. Even if we accept the manufacturers' definition of "cost" and "public" in Sec. 5403(f) (4), the agency complied with the Act.

C. ARBITRARY AND CAPRICIOUS REVIEW OF THE NEW STANDARDS

The manufacturers contend that the new wind standards are arbitrary and capricious in four different respects. We consider each of these arguments separately.

1. The Cost and Benefits of the New Standards

In the Final Rule, HUD acknowledged that the new regulations would increase costs to consumers, but justified those increases on the grounds that the stricter standards would reduce losses from wind damage to manufactured housing occupants and the general public, and because they would help to avoid the "inestimable costs of devastation to people's lives and emotional health and to the communities" caused by severe hurricanes. Final Rule, 59 Fed.Reg. at 2457-58. HUD explained that its statutory mandate required it to look at more than economic considerations in promulgating regulations. It also concluded that the new regulations were in the "optimal societal interest" because the "savings in storm damage repair, loss of personal property, and potential personal injury or loss of life, in addition to other expected benefits, exceeds the cost differential" for the new wind standards. Id. at 2461-62.

HUD quantified these costs and benefits in its Regulatory Impact Analysis ("RIA"). Using data from the damage caused by Hurricane Andrew, the RIA delineated three types of benefits that would result from the new standards: (1) reductions in property damage borne by residents and insurers; (2) reductions in federal disaster relief expenditures; and (3) reductions in deaths and injuries. According to the RIA, the new standards would reduce 75% of the property damage sustained by manufactured housing during "severe wind events" in Wind Zone II, and 83% of the damage in Wind Zone III. Balanced against these benefits were increased costs of production caused by compliance with the new standards, although HUD determined that only a portion of the increased costs would be passed on to consumers in the form of higher prices.⁵  HUD concluded in the RIA that the new standards would produce an annual benefit of $83.8 million and an annual cost of $51.7 million, resulting in an annual net benefit of $32.1 million.

After the Final Rule and the RIA were issued, the manufacturers submitted their Application for Stay to HUD, requesting that the agency stay the effective date of the new regulations pending judicial review. The manufacturers contended that two reports submitted with the Application for Stay (the "Meeks Report" and the "De Alessi Report") demonstrated that the data and methodology HUD used to calculate the costs and benefits were fundamentally flawed. HUD rejected the manufacturers' arguments and denied the Application for Stay.

Before this Court, the manufacturers again challenge HUD's determination in the Final Rule and the RIA that the new wind standards would produce a net economic benefit, and claim that HUD has not adequately addressed the arguments made in the Application for Stay and the Meeks and De Alessi Reports. The manufacturers do not contend that the Act requires the agency to undertake a cost-benefit analysis before promulgating new regulations; in fact, as discussed in the preceding section, they argue that HUD as a matter of law should not subsume consumer costs into a general analysis of societal costs and benefits. However, to the extent that HUD relied on projections indicating that the benefits of the new standards will outweigh the costs as the primary basis for issuing the new standards, the manufacturers assert that these figures relied upon are flawed, making adoption of the new regulations arbitrary and capricious.

Before considering the manufacturers' challenge, we address the dispute between the parties as to what materials are properly included in the rulemaking record. HUD argues that the Meeks and De Alessi Reports and the arguments first made by the manufacturers in their Application for Stay should not be considered by this Court on appeal because the studies and arguments were submitted after HUD issued its Final Rule. HUD also contends that the RIA is not an appropriate object of attack because it was undertaken pursuant to an Executive Order⁶  solely as an internal managerial tool for the federal government. In response, the manufacturers argue that the Application for Stay and the accompanying reports should be included in the rulemaking record because the economic data and analysis used by HUD to support the new wind standards were not disclosed to the public until the Final Rule and the RIA were issued. They also state that the RIA should be open to challenge because it was cited by HUD in the Final Rule to demonstrate that the agency had considered consumer costs as mandated by the Act.

We need not resolve this dispute about the rulemaking record, because HUD's reliance on its cost and benefit figures as support for the new wind standards is not arbitrary and capricious, even assuming that the manufacturers' Application for Stay and the two economic reports accompanying it are included as part of the rulemaking record. Because we assume for the manufacturers that the studies and arguments contained in their Application for Stay are part of the record, we must also assume that HUD's denial of the Application for Stay, which responded to these studies and arguments, is part of the record, too. With these twin assumptions, we proceed to explain why the agency's cost and benefit analysis was not arbitrary and capricious.

In the Application for Stay, the manufacturers first challenged the RIA's projections of the cost increases associated with the new standards. Using estimates from an industry study instead of the figures calculated by HUD, the manufacturers argued that the cost of producing manufactured houses complying with the new wind regulations would be double that of HUD's predictions. Moreover, they asserted that the RIA underestimated the amount of the cost increases that would be passed on to consumers as higher prices. According to the De Alessi Report, the latter error resulted from HUD's use of the national housing market, instead of the individual manufactured housing submarkets of Wind Zones II and III, to measure the elasticity of supply of the manufactured housing industry. The elasticity of supply measures how the market supply will respond to changes in price and is an important factor in forecasting how much consumer prices will rise when production costs are increased.

Arguing that the RIA overstated the extent of losses inflicted on manufactured housing by Hurricane Andrew, the manufacturers also contended that HUD's estimates of the benefits accruing from the new standards were significantly exaggerated. Specifically, the Meeks Report criticized HUD's estimate of insured losses from Hurricane Andrew, because more recent data indicated that the actual amount of insurance payments caused by that storm was almost half the figure used by HUD. In addition, the manufacturers argued that HUD's projections of uninsured losses and federal disaster relief expenditures caused by Hurricane Andrew were incorrectly calculated. Because the only available data on uninsured losses and federal relief payments induced by Hurricane Andrew were for all types of housing, HUD estimated the amounts directly attributable to manufactured housing by multiplying the total figure by the ratio of the number of destroyed manufactured houses to the number of all destroyed houses. The manufacturers claimed that that ratio was methodologically unsound, because the median cost of a manufactured house is only 37% of the median cost of conventional housing. Accordingly, they asserted, the actual amount of uninsured losses and federal relief payments attributable to manufactured housing should have been much smaller than that projected by HUD.

Applying their adjusted cost and benefit calculations, the manufacturers contended in the Application for Stay that the new standards would result in a net loss to society of $61.3 million per year. On appeal, they argue that because there will be a net societal loss rather than gain as a result of the new standards--and because HUD's regulations were justified by its determination that the standards would be economically beneficial--the new wind standards are arbitrary and capricious and should be set aside.

We disagree. We believe that HUD has given a logical explanation for the cost and benefit figures it relied on in promulgating the new wind standards, and that it provided in the Final Rule and the denial of the Application for Stay a reasoned response to the manufacturers' comments and criticisms. Specifically, we conclude that HUD is entitled to rely on the cost estimates calculated by its own engineering staff rather than the figures submitted by the industry's trade association, because our review of the record does not indicate that the agency's projections are either flawed or unreasonable. See North Buckhead, 903 F.2d at 1539 ("When specialists express contrary views, an agency must have discretion to rely on the reasonable opinions of its own qualified experts even if, as an original matter, a court might find contrary views more persuasive." (citation and quotation marks omitted)). The record also demonstrates that HUD's estimate of the amount of cost increases that will be passed on to consumers is based on rational economic analysis.

In addition, we find that HUD has justified its reliance on the public and private benefits cited in the RIA. For example, in response to the manufacturers' attack on the ratio used to calculate the uninsured losses of and federal relief payments to manufactured housing owners caused by Hurricane Andrew, HUD's explanation included the fact that even though manufactured houses have a lower value on average than other housing, manufactured houses generally suffer a higher level of damage. Moreover, although the manufacturers are correct that more recent data suggest that insurance payments for property damage did not reach the levels originally projected by HUD, there is still a net societal benefit even if the updated insurance figures are substituted into the analysis.

Simply put, the manufacturers have not convinced us that HUD's explanation "runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise." State Farm, 463 U.S. at 43, 103 S. Ct. at 2867. The role of this Court is not to decide whether HUD or the manufacturers used the better technical data and methodologies; instead, our task is to determine whether HUD's explanation of its administrative action demonstrates that it has considered the appropriate factors required by law and that it is free from clear errors of judgment. Id. at 43, 103 S. Ct. at 2866-67.

Our conclusion that HUD's cost and benefit projections are not arbitrary and capricious is bolstered by two additional, but related, considerations. First, the manufacturers' projection of a net annual loss does not take into account their own concession that the wind standards needed to be amended, at least to some extent. Thus, the manufacturers' claim of a $61.3 million net loss loses some of its force, because both parties agree that the preamendment status quo is not a viable option.

Second, we agree with HUD that it is not confined by the Act to promulgating new wind standards that will produce a net economic gain. HUD did assert in the Final Rule and the RIA that the projected economic benefits of the amendments would outweigh the economic costs. In addition, however, HUD explained that the need to increase safety and prevent future devastation to communities, such as that caused by Hurricane Andrew, justified increasing consumer prices for manufactured housing. This explanation is consistent with the agency's statutory mandate. Congress has required HUD to consider, in addition to consumer cost, the purposes of the Act--which include reducing injuries, deaths, insurance costs, and property damage. 42 U.S.C.A. Secs. 5401, 5403(f) (West 1983). Our review of the rulemaking record as a whole convinces us that HUD, in determining that the necessity of improving the safety of manufactured housing made the attendant cost increases acceptable, used its "best judgment in balancing the substantive issues." North Buckhead, 903 F.2d at 1539. We will not substitute our judgment for that of the agency. Id.

Because wind damage to structures is affected by natural topography and man-made construction, the ASCE 7-88 standard, which the agency adopted as the basis for the new regulations, establishes four "terrain exposure categories" to reflect the surrounding terrain in which a structure is to be sited. These exposure categories can be briefly described as follows: Exposure A: large city centers; Exposure B: urban, suburban, and wooded areas; Exposure C: open terrain with scattered obstructions; and Exposure D: flat areas exposed to wind approaching over large bodies of water.

The manufacturers argue that the new wind regulations are arbitrary and capricious because they impose on all manufactured homes, regardless of the actual exposure category in which the homes may be located, standards that are based on Exposure C areas. For example, they say that even though 80% of the area in Dade and Broward Counties in Florida qualifies as Exposure B, manufactured homes located there must meet the stricter requirements for the Exposure C category. Thus, the manufacturers argue, the wind regulations are arbitrary and capricious because they ignore the statutory criterion that the standards be reasonable "for the geographic region for which [they are] prescribed." 42 U.S.C.A. Sec. 5403(f) (3) (West 1983).

HUD has provided a sensible explanation for adopting a single exposure level for the new wind standards. It chose Exposure C in the reasonable expectation that most of the manufactured housing will be located in Exposure C conditions. Although HUD could have permitted homes located in Exposure B areas to have somewhat more lenient standards and could have required homes located in Exposure D areas to have somewhat more stringent standards, we cannot conclude that it was unreasonable to choose an intermediate exposure category that provides reasonable protection for all areas in Wind Zones II and III.⁷  HUD explained that it chose a single exposure category so that dealers could stock inventories, and would not be required to order each home based on the individual customer's location. Because HUD neither refused to consider the appropriate factors nor committed a clear error of judgment, the agency's decision to adopt Exposure C wind standards for all Wind Zone II and III areas was not arbitrary and capricious.

3. Achieving HUD's "Stated Purpose"

The manufacturers next contend that the new wind standards are arbitrary and capricious because they cannot achieve the "stated purpose" of the regulations: preventing manufactured housing damage from another Hurricane Andrew. According to the manufacturers, since Andrew's 145 miles-per-hour winds would cause damage even to manufactured homes built according to the revised standards, the cost increases associated with the new standards are not justified.

Of course, the manufacturers previously argued that the new standards are arbitrary and capricious because they go too far, and now they contend that the regulations are arbitrary and capricious because they do not go far enough. Moreover, it is not at all clear that the "stated purpose" of the wind standards is to prevent damage from another Hurricane Andrew. Although Hurricane Andrew was the catalyst behind strengthening the standards, which even the manufacturers admit are too lax, the purpose of the changes according to the Final Rule is simply to increase safety "in areas where wind-induced damage is a particular hazard and risk." Final Rule, 59 Fed.Reg. at 2456. That they do.

Even if the standards were promulgated in order to prevent a reoccurrence of damage of the extent caused by Hurricane Andrew, we still would not find that these standards are arbitrary and capricious. We agree with HUD that the regulations are not invalid merely because they fail to solve every weather-related problem and cannot completely prevent damage from another storm of the ferocity of Hurricane Andrew. Like most regulations of this nature, the utility of the revised wind standards is measured not in terms of absolutes, but in increments of improved safety. According to HUD, estimates of Hurricane Andrew's maximum wind speeds indicated that approximately three percent of the storm area experienced wind speeds of 140 miles per hour or more. Even if manufactured homes constructed under the new regulations would not have been able to survive Hurricane Andrew's peak winds, the new wind standards would have helped to prevent significant damage to the manufactured homes located outside of those areas hit by the storm's strongest winds. That may be the most that can be hoped for. It is enough to convince us to reject the manufacturers' all or nothing contention.

In a related argument, the manufacturers contend that because hurricane risks cannot be eliminated completely, consumers must be given the informed option to sacrifice some hurricane protection in exchange for lower housing costs. According to the manufacturers, by including cost and geographic reasonableness criteria in the Manufactured Housing Act, Congress indicated that consumer choice should be accommodated. Citing Chrysler Corp. v. Department of Transportation, 472 F.2d 659 (6th Cir. 1972), the manufacturers analogize the new wind standards to the regulation of convertibles and sports cars under the automobile safety standards of the Motor Vehicle Safety Act. In Chrysler, the Sixth Circuit observed that soft top convertible cars were "inherently incapable" of meeting certain safety standards, such as rollover requirements, imposed on the automobile industry by that legislation. Finding that the legislation was not intended to eliminate sports cars and convertibles from the market, the court remanded the case to the agency for reconsideration of the safety standard. Id. at 679-80. The Chrysler decision, the manufacturers assert, means that HUD should be prevented from the kind of "governmental paternalism" that denies consumers the ability to choose to live more cheaply and less safely. Because the new standards foreclose the cheap and dangerous option, they are arbitrary and capricious, the argument goes.

We find this argument unpersuasive for three reasons. First, the Manufactured Housing Act is not analogous to the Motor Vehicle Safety Act, because the safety standards at issue under the two acts have different purposes. The regulations under the Motor Vehicle Safety Act that were challenged in Chrysler concerned passive restraint devices designed to protect occupants of an automobile. Id. at 664. In contrast, the wind standards promulgated under the Manufactured Housing Act are designed to protect not only the occupants of manufactured homes, but also other members of the public who could be affected by flying debris during high winds. See Final Rule, 59 Fed.Reg. at 2457-58. HUD quoted in the Final Rule a Federal Emergency Management Agency study that found the disintegration of siding and roofs of manufactured homes " 'contributed significantly to the generation of airborne debris' " during Hurricane Andrew. Id. at 2462. Potential victims of flying debris from manufactured housing, unlike the purchasers of convertibles, do not have the opportunity to choose between cost and safety. What the manufacturers propose would be the equivalent of allowing automobile purchasers to buy at a discount automobiles with unsafe brakes, a consumer choice option that would sacrifice the safety of innocent people who would be given no choice in the matter.

A second difference between the Chrysler decision and this case is that convertibles and sports cars were found to be "inherently incapable" of complying with some of the requirements in the Motor Vehicle Act that hard top vehicles could meet. Chrysler, 472 F.2d at 679. There is no suggestion in this case that the technology does not exist for the industry to comply with the Manufactured Housing Act. In the Chrysler decision, a convertible could not comply with the automobile rollover standard and still remain a convertible. In this case, a manufactured home that conforms to the new wind standards is still a manufactured home, albeit a safer and more expensive one.

Finally, and most fundamentally, the consumer's "right to choose" is not a criterion for decisionmaking under the Manufactured Housing Act. Allowing consumers to knowingly assume the risk of unsafe housing may or may not be a good idea, but it is not one Congress included in the statutory scheme. If the manufacturers want the statutory criteria for promulgating manufactured home standards changed, they should direct their arguments to Congress.

For the reasons stated in this opinion, the manufacturers' petition for review is DENIED.