Suquamish Tribe v. Central Puget Sound Growth Management Hearings Board

Van Deren, J.

¶1 The Suquamish Tribe, Kitsap Citizens for Rural Preservation, and Jerry Harless1 appeal the Central Puget Sound Growth Management Hearings Board (Board) decisions validating Kitsap County’s (County) comprehensive plan under Washington’s Growth Management Act (GMA), chapter 36.70A RCW. The Citizens argue that the Board erred because (1) it applied bright-line rules to approve minimum urban density and maximum rural density; (2) substantial evidence did not support its findings that the County’s reduced urban density plan, the County’s adoption of the minimum density in the land capacity analysis, and the County’s Rural Wooded Incentive Program all complied with the GMA; and (3) it did not decide all issues related to urban density, the land capacity analysis, the Rural Wooded Incentive Program, and the County’s rural element goal harmonizing document. We reverse those parts of the Board’s decisions that the Citizens challenge on appeal and remand for the Board to reconsider appropriate urban and rural densities without use of bright-line rules and for it to address all of the issues that the Citizens raised relating to urban density, the land capacity analysis, the Rural Wooded Incentive Program, and the County’s rural element goal harmonizing document.

FACTS

I. Kitsap County’s Growth Management Act Comprehensive Plan History

¶2 In 1990, Kitsap County began creating its first comprehensive plan under the GMA. In 1992, the County adopted its first set of countywide planning policies to guide *749its development regulations and comprehensive plan creation. The County’s original countywide planning policies set a development goal that allocated 83 percent of growth to urban areas and 17 percent of growth to rural areas.

¶3 In 1998, the County adopted the comprehensive plan and, on review, the Board found it valid. The plan established various urban density levels and a minimum density for the urban low and urban cluster designations of five dwelling units per acre and one dwelling unit per five acres for areas with the rural residential designation.

¶4 The County’s 2002 buildable lands report showed that, from 1995 to 2000, urban growth had not occurred at minimum density levels (five dwelling units per acre) established in the 1998 comprehensive plan. The buildable lands report also showed that growth did not occur in accordance with the 1992 countywide planning policies’ goal to allocate growth primarily in the urban growth areas (UGAs). In fact, 43 percent of all growth occurred in the UGAs and 57 percent occurred in rural areas. To remedy the inconsistency between actual growth and the projected growth in the UGAs and the rural areas, the County adopted “reasonable measures” in 2004 to remedy the discrepancy.2 Administrative Record (AR) at 2150. It also updated the countywide planning policies, reducing the urban growth target to 76 percent and delaying the 1992 goal of 83 percent of growth in UGAs until the County could achieve the 76 percent target.

¶5 In 2006, the County completed a 10-year update of its comprehensive plan. The 2006 updated plan included the land capacity analysis that the County used to set the size of its UGA based on (1) population projections, (2) the urban *750growth target of 76 percent,3 (3) available land in the county, (4) a market factor related to the likelihood of redevelopment of different urban parcels, and (5) the minimum density within each zoning area.4 The latest plan’s various zoning densities were: urban restricted residential (one to five dwelling units per acre), urban low residential and urban cluster residential (four to nine dwelling units per acre), urban medium residential (10-18 dwelling units per acre), and urban high residential (19-30 dwelling units per acre).5

¶6 The updated 2006 comprehensive plan also included a Rural Wooded Incentive Program with the goal to preserve open space and forestry through partial residential development at densities higher than those otherwise allowed by zoning. The Rural Wooded Incentive Program allows development of rural wooded land6 at residential densities equivalent to one dwelling unit per five acres in exchange for setting aside 75 percent of the land for open space. It also allows clustered developments in parcels ranging from 20 to 500 acres. The County may approve projects until the total acreage developed under the Rural Wooded Incentive Program reaches 5,000 acres, at which point the county commissioners must decide whether to authorize another 5,000 acre phase. The program restricts clusters to a maximum of *75125 houses, sets no minimum lot size,7 creates minimum setbacks of 20 feet in the front yard and 5 feet on the sides and rear, and requires a 150 foot buffer between clusters. It also contains hearing examiner approval criteria for proposed projects and a postdevelopment monitoring and evaluation component that allows the county commissioners to review each 5,000 acre phase.

II. Citizens’ Petition for Review of County Plan to the Board

¶7 The Citizens raised three broad issues8 in their petition for review of the County’s 2006 comprehensive plan to the Board that are relevant to this appeal:

• Whether the County’s decision to reduce urban density and to expand the UGA complied with the GMA’s (1) goals for urban growth, sprawl, transportation, housing, environment, and public facilities and service; (2) requirements for UGA designation; and (3) requirement for internal consistency.
• Whether the County’s land capacity analysis complied with the GMA’s (1) urban growth and antisprawl goals, (2) requirements for UGA designation, and (3) requirement for internal consistency.
• Whether the County’s Rural Wooded Incentive Program complied with the GMA’s (1) environment, open space, and recreational goals; (2) rural element requirement; and (3) definition of “rural development.”[9]

A. The Board’s Initial Decision in Response to Petitions

1. Minimum Density of Four Dwelling Units per Acre

¶8 In its review, the Board found that the County designated a minimum density of four dwelling units per acre for *75290 percent of its urban areas. The Board approved this density:

At the outset, the Board acknowledges that 4 [dwelling units per acre] is an “appropriate” urban density; it is not low-density sprawl. In fact, the County is correct in noting that since 1995, 4 [dwelling units per acre] has been an approved and accepted minimum urban density for Kitsap County.

AR at 1952 (citing Bremerton v. Kitsap County, No. 95-3-0039, 1995 WL 903165, 1995 GMHB LEXIS 384 (Cent. Puget Sound Growth Mgmt. Hr’gs Bd. Oct. 6, 1995) (Bremerton I). The Board acknowledged the Citizens’ “very impressive evidence” of the service efficiencies for higher density and agreed “that there is certainly persuasive evidence providing a solid basis and rationale for increased densities and compact urban growth.” AR at 1952-53. But it ruled that the reduction in density did not fall “outside the GMA’s requirements.” AR at 1953.

¶9 In response to the Citizens’ argument that the only local consideration came from a “Citizen Advisory Group,” the Board emphasized that “4 [dwelling units per acre] is an appropriate urban density and not the historic pattern previously found in Kitsap County.”10 AR at 1953. It concluded that the Citizens did not carry their burden of proof and it rejected the argument “that the County’s action of establishing a minimum urban density of 4 [dwelling units per acre] is clearly erroneous or contrary to the challenged provisions of the GMA.” AR at 1953. After reaching this conclusion, the Board failed to address the further assertion regarding inconsistency with the GMA goals and internal plan inconsistency.

2. Land Capacity Analysis Challenges

¶10 The Board also rejected the Citizens’ challenge to the methodology that the County employed in the land *753capacity analysis to calculate the size of UGAs because it had already adopted the County’s four dwelling units per acre figure as the proper minimum density for UGAs:

Just as the Board agreed with the County in regard to urban density, the Board here also agrees with the County on its methodology The [land capacity analysis] largely rests upon a residential density assumption of 4 [dwelling units per acre], which, as the Board has stated supra, is an “appropriate” urban density The consequence of adopting this lower assumption is, in fact, to demonstrate a need for more urban land. The methodology of the County is not flawed, nor is the use of a minimum of 4 [dwelling units per acre] rather than a trend or mid-range density flawed or in violation of any GMA directive. However, the Board does agree with [Citizens] that adopting this approach may dampen the recent success the County has had in encouraging higher densities in the UGAs, since the County concedes that between 2000 and 2005, the County achieved an average of 5.6 [dwelling units per acre] for urban low density plats.

AR at 1956. In reaching this conclusion based on the four dwelling units per acre density, the Board again failed to address whether the land capacity analysis’s use of this minimum density created internal inconsistencies with other portions of the comprehensive plan or whether the County clearly erred in its land capacity analysis calculations with a “reduction for critical areas” in combination with the Urban Restricted Residential zone.

3. Rural Wooded Incentive Program

¶11 The Board then addressed the Citizens’ challenges to the Rural Wooded Incentive Program. It reviewed one of its earlier decisions that had noted the large number of nonaggregated nonconforming lots (legacy lots), large lots, and ongoing forestry on nonresource lands in Kitsap County. The Board assumed that the earlier circumstances remained unchanged and that they served as the basis of the County’s Rural Wooded Incentive Program: “[T]he ‘local *754circumstances’ articulated in Bremerton II[11] have apparently not changed, but have not been specifically articulated in briefing. The Board therefore assumes that these ‘local circumstances’ have not changed and remain the problem that the County is attempting to solve with the [Rural Wooded Incentive Program].” AR at 1970.

¶12 Addressing the Citizens’ challenge to “clustering” of homes under the Rural Wooded Incentive Program, the Board ruled that clustering that would allow density of one dwelling unit per five acres was not clearly erroneous:

[U]nder the most generous option, a 100-acre parcel is allowed up to a maximum of 20 residences, a net residential density of 1 [dwelling unit per] 5 acres - a rural, not urban, density, that is consistent with preserving the rural character. The Board acknowledges that the clustered design of the development appears more dense when viewed in isolation, but it is nonetheless a rural density when viewed in the context of the entire parcel.

AR at 1975.

¶13 The Citizens also challenged the rural development design standards and County oversight of the Rural Wooded Incentive Program developments as inadequate to ensure that the rural areas were preserved. In addressing these challenges, the Board relied on the County’s specific design standards for development in the rural areas and ruled that the absence of a minimum lot size in the rural areas developed under the Rural Wooded Incentive Program did not affect the validity of the program because the County included a “directive to prevent clustering from becoming a predominant land use pattern in the rural area.” AR at 1975. The Board also relied on the County’s *755monitoring program12 because the County would review the process and development patterns on a biennial13 basis.

¶14 The Board also rejected the Citizens’ claim that the County’s 2006 comprehensive plan with regard to the Rural Wooded Incentive Program was clearly erroneous with regard to the provision of urban services:

Although the Board can understand that clustered residential development gives the appearance of a suburban environment and a need for urban services, the property subject to the [Rural Wooded Incentive Program] remains in the rural area and clusters are limited to 25 units with specific location and buffering requirements. The [Citizens] have failed to adequately demonstrate that clustered development within a rural area will result in a demand for urban services and the Board must presume that the County, and its examiner, will adhere to the approval criteria for the [Rural Wooded Incentive Program].

AR at 1977-78. Thus, the Board also ruled that “the initial scope of the [Rural Wooded Incentive Program] [Phase I - 5000 acres] is modest and reasonable and not clearly erroneous.” AR at 1973 (emphasis omitted) (second alteration in original).

¶15 The Board confronted the Citizens’ arguments that the Rural Wooded Incentive Program would lure away growth from designated UGAs:

On its face, permitting clustered development within the rural area seems contrary to a key tenet of the GMA - encouraging urban-style growth within urban area[s]. However, the GMA promotes the use of innovative land use management techniques such as clustering and the Act specifically defines rural development to include clustered residential *756development, at levels that are consistent with the preservation of rural character and the requirements of the rural element. Although clustering is permitted in the rural areas, the GMA is cognizant that the magnitude of such clustering can potentially affect rural character. This is why it is important that rural clustering be monitored to ensure that its magnitude and extent is not overreaching. The County has included such monitoring provisions in the [Rural Wooded Incentive Program].
Approval criteria for [Rural Wooded Incentive Program] developments specifically require that rural character be preserved. Comprehensive Plan policies and design standards are also in place to ensure that one of the stated purposes of the [Rural Wooded Incentive Program] - “produce a development pattern in rural areas that is consistent with rural character” - is met. These same design standards also address cluster size, location, and aesthetic buffering. The biennial monitor program also works to ensure maintenance of the qualities of rural character.
Given the GMA’s allowance for clustering within rural areas, the County’s [Rural Wooded Incentive Program], including the supporting Plan Policies and monitoring provisions to protect the rural character, cannot be seen to be in violation of the Act.

AR at 1978 (emphasis and citations omitted).

¶16 But here, the Board agreed with the Citizens that the written record did not demonstrate that the rural element in the Rural Wooded Incentive Program harmonized with the GMA planning goals. The Board remanded for the County to create a “document that accomplishes the harmonizing of goals” and to provide sufficient evidence of the required balancing of the GMA goals. AR at 1982.

a. Does the Rural Wooded Incentive Program Harmonize with GMA Goals?

¶17 Following this first remand, the County submitted a document that it asserted explained how the Rural Wooded Incentive Program harmonized with the GMA’s planning goals. The Citizens argued that the County’s submitted goal harmonizing document did not consider local circum*757stances, that the goal-harmonizing document ignored the buildable lands report, and that the Board improperly assumed that local circumstances from one of its earlier decisions applied to this later plan. Again, the Board agreed with the Citizens and found that the existing goal harmonizing document inadequately explained the local circumstances prompting the Rural Wooded Incentive Program’s creation and again remanded to the County to demonstrate how the Rural Wooded Incentive Program reflected local circumstances.14

b. Does the Rural Wooded Incentive Program Reflect Local Circumstances?

¶18 The County submitted an updated goal-harmonizing document that added information about local circumstances:

The County’s GMA record reflects that while forestry operations occur in Kitsap County, such operations do not constitute “forest resource lands” as defined by the GMA. There exists no or limited infrastructure necessary to support such commercial operations. Nevertheless, there remains some small-scale forest operations scattered in rural areas throughout the County. The majority of these operations are under the control of a handful of landowners. Many of the existing rural parcels were segregated prior to GMA and may not meet current zoning. Thus, even with minimum lot sizes of 20 acres, such pre-GMA lots could be developed under Washington’s vested rights doctrine.
Kitsap County has been studying and proposing a rural clustering program since GMA’s inception. The rural clustering program has been refined over the years to provide an incentive [for] preserving large tracts of rural open space in exchange for an increase in density The program provides that such open spaces shall be associated as much as practicable to assure connection of the open space, trails and wildlife corridors. The [Rural Wooded Incentive Program] provides an incentive for *758owners of large tracts of land to realize a better economic return while preserving open space, and possibly encouraging aggregation of nonconforming rural lots.

AR at 4373-U.6. The Board noted that the buildable lands report showed that development from 1995 to 1999 was 43 percent urban and 57 percent rural and that the rate of urban development increased to 57 percent from 2000 to 2005, while the rate of rural development decreased to 43 percent in that same time period; thus, the “fear that development of [Rural Wooded Incentive Program] residential clusters will spawn further development of legacy lots is speculative,[15] but may not be unwarranted.” AR at 4755.

¶19 Nevertheless, the Board found that the additional information merited compliance on this remand issue. It concluded that “the County’s treatment of legacy lots is beyond the scope of the Board’s review for determining compliance on the limited remand issue for the [Rural Wooded Incentive Program]” and that the Citizens’ arguments about increasing urban growth did not “go to the remand issue of modifying the Goal Harmonizing Document to reflect local circumstances.” AR at 4755.

B. Citizens’ Request for Reconsideration — Urban Density and Land Capacity Analysis

¶20 The Citizens unsuccessfully requested reconsideration of the Board’s decision approving the urban density of four dwelling units per acre and the land capacity analysis. They argued that the Board did not address whether the 2006 comprehensive plan (1) created internal inconsistency when it set the minimum urban density of four dwelling units per acre, (2) created inconsistency with the GMA goals when it used four dwelling units per acre as a minimum urban density, (3) created internal inconsistency *759between a land capacity analysis using a reduced minimum density and the higher average density used to allocate land for development, or (4) underestimated the number of acres available in the County because the land capacity analysis simultaneously excluded lands from the calculation and included designated lands under the 2006 comprehensive plan with the lower density Urban Restricted Residential zone. The Board concluded that “[Citizens] simply reargue, or attempt to offer new argument pertaining to these Megal [i]ssues. The Board remains unpersuaded on these issues and finds and concludes that it has not misinterpreted the law.” AR at 2069.

III. Citizens’ Appeal to Thurston County Superior Court

¶21 The Citizens appealed the Board’s decisions to superior court. The superior court affirmed the Board’s August 15, 2007, final decision and order; the Board’s April 4, 2008, order finding partial compliance and noncompliance and invalidity; and the Board’s June 5, 2008, order finding compliance.

¶22 The Citizens appeal.

ANALYSIS

I. Deference, Burden of Proof, and Standards of Review

¶23 The GMA affords broad discretion to local governments in planning for growth, bounded only by the GMA’s goals and requirements. King County v. Cent. Puget Sound Growth Mgmt. Hearings Bd., 142 Wn.2d 543, 561, 14 P.3d 133 (2000); see also Thurston County v. Cooper Point Ass’n, 148 Wn.2d 1, 13-15, 57 P.3d 1156 (2002). Boards16 must afford a county’s actions great deference so long as the action complies with the GMA and is not clearly erroneous. RCW 36.70A.320(2); Lewis County v. W. Wash. Growth Mgmt. Hearings Bd., 157 Wn.2d 488, 497, 139 P.3d 1096 *760(2006). If a board affords a county’s action proper deference under the “clearly erroneous” standard, we, in turn, defer to the board’s decision. Quadrant Corp. v. Cent. Puget Sound Growth Mgmt. Hearings Bd., 154 Wn.2d 224, 238, 110 P.3d 1132 (2005). A board must presume that a county’s action is valid, leaving the challenger to meet the burden of establishing invalidity. RCW 36.70A.320(2); City of Redmond v. Cent. Puget Sound Growth Mgmt. Hearings Bd., 116 Wn. App. 48, 55, 65 P.3d 337 (2003).

¶24 “Judicial review of board actions is governed by the Administrative Procedure Act [(APA)], chapter 34.05 RCW.” Thurston County v. W. Wash. Growth Mgmt. Hearings Bd., 164 Wn.2d 329, 341, 190 P.3d 38 (2008). And under the APA, the “party appealing a board’s decision has the burden of demonstrating [its] invalidity.” Thurston County, 164 Wn.2d at 341; see also RCW 34.05.570(1)(a). “In reviewing an administrative action, we sit in the same position as the trial court and apply the APA standards directly to the administrative record.” Superior Asphalt & Concrete Co. v. Dep’t of Labor & Indus., 112 Wn. App. 291, 296, 49 P.3d 135 (2002).

¶25 In relevant part, a reviewing court must grant relief from an agency’s adjudicative decision if

(d) The agency has erroneously interpreted or applied the law;
(e) The order is not supported by evidence that is substantial when viewed in light of the whole record before the court, which includes the agency record for judicial review, supplemented by any additional evidence received by the court under this chapter;
(f) The agency has not decided all issues requiring resolution by the agency.

RCW 34.05.570(3). We review a board’s “legal conclusions de novo, giving substantial weight to its interpretation of the statutes it administers” and its “findings of facts for substantial evidence.” Manke Lumber Co. v. Cent. Puget Sound Growth Mgmt. Hearings Bd., 113 Wn. App. 615, 622, *76153 P.3d 1011 (2002). And as a threshold matter, we must determine whether an agency decided all of the issues necessary for judicial review. See Low Income Hous. Inst. v. City of Lakewood, 119 Wn. App. 110, 118-19, 77 P.3d 653 (2003).

II. Bright-Line Rules

¶26 The Citizens first argue that the Board erred when it (1) used a bright-line rule to approve minimum urban density, (2) approved a land capacity analysis that relied on the minimum urban density based on the bright-line rule to calculate land capacity, and (3) used a bright-line rule to approve appropriate rural density. The County argues that the Board did not apply a bright-line rule but, instead, the Board (1) found that four dwelling units per acre was an appropriate urban density for Kitsap County, (2) “summarized and reviewed the arguments of both parties” regarding the land capacity analysis and found that the County had “made a showing that the methodology was valid,” arid (3) did not mention a bright-line rule or apply it to the land capacity analysis. Br. of Resp’t at 34. According to the County, “[f ]ive-acre lots are clearly rural parcels under any planning principle.”17 Br. of Resp’t at 47. We agree with the Citizens, reverse the Board, and remand to the Board for further proceedings.

A. Standard of Review

¶27 Where a board erroneously interprets or applies the law, we review the issue de novo. Thurston County, 164 Wn.2d at 341; see RCW 34.05.570(3)(d). Whether a bright-line rule exists for appropriate density is a question of law. Thurston County, 164 Wn.2d at 358 n.19. And a board errs when it uses a bright-line rule to support its decision, regardless of whether its reliance is explicit or implicit. Thurston County, 164 Wn.2d at 358-59 & n.20.

*762¶28 A board cannot create bright-line rules “because ‘the growth management hearings boards do not have authority to make “public policy” even within the limited scope of their jurisdictions, let alone to make statewide public policy.’ ” Thurston County, 164 Wn.2d at 353 (quoting Viking Props., Inc. v. Holm, 155 Wn.2d 112, 129, 118 P.3d 322 (2005)). Instead, “the GMA creates a general ‘framework’ to guide local jurisdictions instead of ‘bright-line’ rules.” Viking Props., 155 Wn.2d at 129. Thus a county must base its action on “local circumstances” and the board, in its review, must engage in a focused factual inquiry and make its decision “on the specific circumstances of each case.” Thurston County, 164 Wn.2d at 353, 358 n.19. Where a board uses a bright-line rule, we must remand “to the Board to determine, without regard to the bright-line rule . . . , whether it was clearly erroneous for the County to include the challenged densities.” Gold Star Resorts, Inc. v. Futurewise, 167 Wn.2d 723, 735, 222 P.3d 791 (2009).

B. Urban Density

¶29 As the County does here, the County’s planning process repeatedly relied on and argued that the Board’s adoption of an earlier bright-line rule also demonstrated that four dwelling units per acre was an appropriate urban density for Kitsap County:

• “In Bremerton [I], the [Board] found that, as a general rule, [four] dwelling units per acre or more constitutes urban densities for Kitsap County. Four dwelling units per acre addresses GMA requirements specific to Kitsap County and these community desires.” AR at 2894.
• “Staff is reviewing lowering this minimum for the urban low zone to [four] (still considered urban according to . . . Bremerton [II]) and its effect o[n] the provision of public services and infrastructure.” AR at 1030.
• “The minimum density of [four dwelling units per acre] is a density that is considered urban for Kitsap County by the [Board] in Bremerton [I].”AR at 2901.
• “This reduces the capacity of the single-family designated areas; however, the new minimum of [four dwelling units *763per acre] still meets urban densities as defined in [Board] cases applicable to Kitsap County.” AR, at 3028.
• “[R] eduction in the minimum allowed density to [four dwelling units per acre] in the Urban Low and Urban Cluster designations would reduce some dwelling capacity, but still complies with GMA as it is an urban density.” AR at 2820.
• And to support its contention that four dwelling units per acre is “GMA compliant,” the County’s report quoted the language from Bremerton I: “ ‘[Generally, a]ny residential pattern [of four net dwelling units per acre], or higher, is ... compact urban development and satisfies the low end of the range required by the [GMA].’ ” AR at 3287 n.l (alteration in original); Bremerton I, 1995 WL 903165, at *35, 1995 GMHB LEXIS 384, at *102.

¶30 The Board employed similar bright-line rule language in its orders:

• “[Four dwelling units per acre] is an ‘appropriate’ urban density; it is not low-density sprawl.” AR at 1952.
• “[T]he County is correct in noting that since 1995, [four dwelling units per acre] has been an approved and accepted minimum urban density for Kitsap County. See Bremerton [I], [1995 WL 903165, 1995 GMHB LEXIS 384].” AR at 1952.
• “[Four dwelling units per acre] is an appropriate urban density and not the historic pattern previously found in Kitsap County.” AR at 1953.
• “The [land capacity analysis] largely rests upon a residential density assumption of [four dwelling units per acre], which, as the Board has stated supra, is an ‘appropriate’ urban density.” AR at 1956.

¶31 Although the County and the Board majority tried to characterize its 1995 decision as settling the issue of the appropriate Kitsap urban density based on local circumstances, the 1995 decision defies that characterization:

As noted above, the Board holds that up to 2.5-acre lots are urban. However, rather than adopt a maximum urban lot size, *764the Board instead adopts as a general rule a “bright line” at four net dwelling units per acre. Any residential pattern at that density, or higher, is clearly compact urban development and satisfies the low end of the range required by the Act. Any larger urban lots will be subject to increased scrutiny by the Board to determine if the number, locations, configurations and rationale for such lot sizes compl[y] with the goals and requirements of the Act, and the jurisdiction’s ability to meet its obligations to accept any allocated share of county-wide population. Any new residential land use pattern within a UGA that is less dense is not a compact urban development pattern, constitutes urban sprawl, and is prohibited. There are exceptions to this general rule. For example, 1-or 2.5-acre lots may be appropriate in an urban setting in order to avoid excessive development pressures on or near environmentally sensitive areas. However, this circumstance can be expected to be infrequent within the UGA and must not constitute a pattern over large areas.

Bremerton I, 1995 WL 903165, at *35, 1995 GMHB LEXIS 384, at *101-02 (boldface omitted).

¶32 Bremerton I fails to provide any rationale beyond its bright-line rule of four dwelling units per acre — in fact, early decisions of this Board focused on creating these broad and, subsequently rejected, wide-ranging rules. See 1995 WL 903165, at *31-32, 1995 GMHB LEXIS 384, at *90-94; see also Thurston County, 164 Wn.2d at 358 n.21. Our Supreme Court characterized the Board’s decision in Bremerton I as “adopting a bright-line urban density of a minimum of four dwelling units per acre.”18 Thurston County, 164 Wn.2d at 358 n.21.

*765¶33 The County and Board’s attempts to recharacterize Bremerton I as an opinion that considered local circumstances to justify four dwelling units per acre in Kitsap County fail. Moreover, even if the Board’s decision provided an appropriate urban density for Kitsap County in Bremerton I, the decision in 1995 cannot provide a local bright-line rule a decade later. Over time, circumstances can change in a county or in a portion of a county; and, as our Supreme Court has stated, “The update process ‘provides the vehicle ... for recognizing changes in land usage or population.’ ” Thurston County, 164 Wn.2d at 344 (quoting Gold Star Resorts, Inc. v. Futurewise, 140 Wn. App. 378, 390, 166 P.3d 748 (2007), aff’d in part, rev’d in part on other grounds, 167 Wn.2d 723 (2009). Furthermore, a bright-line rule for a specific county would “subject certain densities to increased scrutiny” and other densities to less scrutiny, a rule which would exceed a board’s authority. Thurston County, 164 Wn.2d at 359. Thus, we hold that the Board erred when it used a bright-line rule to approve the minimum urban density of four dwelling units per acre in Kitsap County.

C. Rural Density

¶34 Here, the Board approved “a net residential density of [one dwelling unit per five acres] - a rural, not urban, density, that is consistent with preserving the rural character.” AR at 1975. The Citizens challenge this determination as further application of a forbidden bright-line rule. We agree.

¶35 The GMA also establishes development guidelines encompassing “lands that are not designated for urban growth, agriculture, forest, or mineral resources.” RCW 36.70A.070(5). Under these development guidelines, a county’s comprehensive plan must “provide for a variety of rural densities, uses, essential public facilities, and rural governmental services needed to serve the permitted den*766sities and uses.” RCW 36.70A.070(5)(b). “To achieve a variety of rural densities and uses, counties may provide for clustering, density transfer, design guidelines, conservation easements, and other innovative techniques that will accommodate appropriate rural densities and uses that are not characterized by urban growth and that are consistent with rural character.” RCW 36.70A.070(5)(b).

¶36 Our Supreme Court has noted that “[t]he legislature did not specifically define what constitutes a rural density. Instead, it provided local governments with general guidelines for designating rural densities.” Thurston County, 164 Wn.2d at 359. And “[w]hether a particular density is rural in nature is a question of fact based on the specific circumstances of each case.” Thurston County, 164 Wn.2d at 359. Thus, a board should not reject or accept a density “based on a bright-line rule for maximum rural densities, but must . . . consider local circumstances and whether these densities are not characterized by urban growth and preserve rural character.” Thurston County, 164 Wn.2d at 360.

¶37 In Thurston County the Western Washington Growth Management Hearings Board (WWGMHB) “framed the issue as to whether the County’s comprehensive plan failed to comply with the GMA by allowing ‘development at densities of greater than one [dwelling] unit per five acres when this board has determined that such densities fail to comply with the GMA.’ ” 164 Wn.2d at 358 n.20 (quoting Thurston County AR at 2546). The WWGMHB concluded that Thurston County’s comprehensive plan did not provide for a variety of rural densities because it had invalidated the rural clustering program, leaving only one density, which did not satisfy the GMA’s requirement of a variety of densities in the rural element. Thurston County, 164 Wn.2d at 356-57. Our Supreme Court noted, “Although the [WWGMHB] did not explicitly adopt a five acre bright-line rule, such a rule was implicit in its decision because of the way the issue regarding rural densities was framed.” Thurston County, 164 Wn.2d at 358 n.20.

*767¶38 The court held that a board, “as a quasi-judicial agency, lacks the power to make bright-line rules regarding maximum rural densities” and “may not use a bright-line rule to delineate between urban and rural densities, nor may it subject certain densities to increased scrutiny.” Thurston County, 164 Wn.2d at 358-59. The court then remanded for the WWGMHB to review the comprehensive plan and consider, based on local circumstances, whether the rural densities were “characterized by urban growth” or failed to preserve rural character. Thurston County, 164 Wn.2d at 360.

¶39 Our Supreme Court has also emphasized that “when differentiating between urban and rural densities, [a b]oard cannot employ bright line rules.” Gold Star Resorts, 167 Wn.2d at 734. In Gold Star Resorts, Futurewise argued that WWGMHB did not use a bright-line rule, but the court pointed out the implicit application of a bright-line rule:

The [WWGMHBJ’s final decision and order shows that its decision on the rural density issue was based on its determination that “[rjesidential densities of greater than one dwelling unit per five acres are not considered rural densities,” and “[d]ensities that are not urban but are greater than one dwelling unit per five acres generally promote sprawl in violation of goal 2 of the GMA,” with the only exception being in [Limited Areas of More Intensive Rural Development]. Contrary to Futurewise’s claims, the [WWGMHB] did apply a bright line rule and Thurston County controls on this issue.

167 Wn.2d at 734-35 (second and third alterations in original) (citations omitted) (quoting Gold Star Resorts CP at 1566, 1567).

¶40 Here, the Board considered rural densities when it reviewed the Rural Wooded Incentive Program:

[T]here is no inherent error in the County’s clustering program provided for in the [Rural Wooded Incentive Program]. The Board notes that under the most generous option, a 100-acre parcel is allowed up to a maximum of 20 residences, a net residential density of [one dwelling unit per five acres] - a rural, not urban, density, that is consistent with preserving the *768rural character. The Board acknowledges that the clustered design of the development appears more dense when viewed in isolation, but it is nonetheless a rural density when viewed in the context of the entire parcel.

AR at 1975. This discussion relied on one dwelling unit per five acres as a per se rural density, thus using a bright-line rule.

¶41 It is not apparent from the Board’s decision that it considered whether the Rural Wooded Incentive Program’s provisions could create clusters of a significant size, allowing developers to site clusters relatively near to one another. The County’s 2006 comprehensive plan relied on, and the Board adopted, a bright-line rule for rural development that looked only at the entire parcel, not its relationship to other parcels.19 It is clear that the Board implicitly relied on this bright-line rule to support its conclusion that the “County’s clustering alternatives are not clearly erroneous.” AR at 1975 (boldface omitted).

¶42 Under these circumstances and on this record, we hold that the Board used a bright-line for rural density to approve the Rural Wooded Incentive Program without considering specific and current local circumstances and whether the prescribed densities would preserve rural character.20 Thus, the Board “improperly relied in this case on a ‘one residence per five acre’ rule.” Gold Star Resorts, 167 Wn.2d at 734.

*769¶43 Like a modern day Cassandra,21 our Supreme Court has warned the boards against bright-line rules, and this warning has fallen on deaf ears. Viking Props., 155 Wn.2d at 128-30; see Joseph W. Tovar, Puget Sound Reg’l Council, Appropriate Urban Densities in the Central Puget Sound Region: Local Plans, Regional Visions, and the Growth Management Act 15, 17-18 (Nov. 2, 2005), http://www.psrc.org/ assets/2030/appIF12-urban-densities.pdf. Thus, we reiterate that a board may not take a shortcut to approve or disapprove proposed urban or rural densities based on a bright-line rule — it must reevaluate local circumstances in each instance to account specifically for the inevitable changes occurring over time. Because the Board employed a bright-line rule to approve the minimum urban density of four dwelling units per acre and the rural density of one dwelling unit per five acres for the Rural Wooded Incentive Program, we remand for the Board to consider the current, specific local circumstances before resolving the issue of appropriate densities to be used in the County’s revisions to its comprehensive plan.

III. Lack of Substantial Evidence to Support Preservation of Rural Character

¶44 The Citizens next argue that substantial evidence does not support the Board’s conclusion that the Rural Wooded Incentive Program will preserve rural character22 because (1) the design standards do not protect rural character, (2) these proposed developments will require urban services, and (3) postdevelopment monitoring will *770not protect rural character.23 The County relies on former Kitsap County Code (KCC) 17.301.080(H)(4) (2008) to support its argument that the Rural Wooded Incentive Program requires that developments preserve rural character.24 We agree that substantial evidence does not support the Board’s conclusion that the design standards protect rural character, but the Citizens do not carry their burden of showing that the potential developments will require urban services on the record developed thus far.25

A. Standard of Review

¶45 We review findings of fact for substantial evidence in light of the whole record. RCW 34.05.570(3)(e). “A board’s order must be supported by substantial evidence,” and the evidence must be of a sufficient quantity “ ‘to persuade a fair-minded person of the truth or correctness of the order.’ ” Thurston County, 164 Wn.2d at 341 (internal quotation marks omitted) (quoting City of Redmond v. Cent. Puget Sound Growth Mgmt. Hearings Bd., 136 Wn.2d 38, 46, 959 P.2d 1091 (1998)). “Errors alleged under subsection (e) are mixed questions of law and fact, where the reviewing court determines the law independently, then applies it to the facts as found by the Board.” City of Arlington v. Cent. Puget Sound Growth Mgmt. Hearings Bd., 164 Wn.2d 768, 779-80, 193 P.3d 1077 (2008), affirming and adopting 138 Wn. App. 1, 154 P.3d 936 (2007).

B. Design Standards

¶46 The Citizens argue that there is not substantial evidence in the record “to support the Board’s finding of *771‘ample specific standards’ ” that govern the Rural Wooded Incentive Program. Br. of Appellant at 48 (quoting AR at 1975). The County counters that the decision criteria in the Rural Wooded Incentive Program require a hearing examiner to find that rural character is preserved before approving the project. Br. of Resp’t at 50-51.

¶47 The GMA defines “rural character”:

“Rural character” refers to the patterns of land use and development established by a county in the rural element of its comprehensive plan:
(a) In which open space, the natural landscape, and vegetation predominate over the built environment;
(b) That foster traditional rural lifestyles, rural-based economies, and opportunities to both live and work in rural areas;
(c) That provide visual landscapes that are traditionally found in rural areas and communities;
(d) That are compatible with the use of the land by wildlife and for fish and wildlife habitat;
(e) That reduce the inappropriate conversion of undeveloped land into sprawling, low-density development;
(f) That generally do not require the extension of urban governmental services; and
(g) That are consistent with the protection of natural surface water flows and ground water and surface water recharge and discharge areas.

Former RCW 36.70A.030(15) (2005). And the GMA’s “rural element” calls for specific measures:

The rural element shall include measures that apply to rural development and protect the rural character of the area, as established by the county, by:
(i) Containing or otherwise controlling rural development;
(ii) Assuring visual compatibility of rural development with the surrounding rural area;
(iii) Reducing the inappropriate conversion of undeveloped land into sprawling, low-density development in the rural area;
(iv) Protecting critical areas, as provided in RCW 36.70A-.060, and surface water and groundwater resources; and
*772(v) Protecting against conflicts with the use of agricultural, forest, and mineral resource lands designated under RCW 36.70A.170.

RCW 36.70A.070(5)(c).

¶48 In Kitsap County, rural character is satisfied for a hearing examiner’s purposes where a proposed plan incorporates:

a. Clustering of development, as applicable;
b. Preservation of critical areas, resource areas, groundwater recharge, and natural features;
c. Provision for a coordinated, comprehensive, interconnected, and integrated system of permanent open space areas; and
d. Placement of structures, circulation systems and utilities that minimize impervious surfaces and alteration of the land and also responds to physical characteristics of the property.

Former KCC 17.301.080(H)(4). But former KCC 17.301-.080(E), relied on by the Board, sets development standards for lots, cluster size, buffering between clusters, and open space.

¶49 Many, if not most, of the development standards related to rural character use precatory language such as “encourage,” “optimize,” and “should”:

5. Each cluster shall be limited to twenty-five units or fewer. Clusters within a development should be sited to achieve the following objectives and criteria. The director may allow exceptions based upon site-specific conditions or when conflicts occur between the criteria.
a. Optimize protection of critical areas, including wetlands and stream corridors, by keeping clusters and other development away from critical areas to the extent possible;
b. Optimize preservation and interconnectivity of open space as a permanent preservation of open space;
c. Avoid development on ridgelines, in the center of open field, or located on other prominent topographical features or scenic elements, visible to adjacent and vicinity properties when other locations are available; and
*773d. Minimize topographic alteration.
6. Clusters developed under this program shall provide a one-hundred-foot vegetated buffer from existing public roadways and adjoining properties and a one-hundred-fifty-foot buffer between clusters in order to preserve rural character and the aesthetic values of Rural Wooded lands.
a. Where two Rural Wooded Incentive Program developments abut each other, they are encouraged, where practical, to coordinate permanent open space to provide interconnectivity;
b. Buffers are encouraged to incorporate natural features to maximize retention of views and rural character;
c. Where native vegetation is available to create a sight-obscuring buffer, that vegetation should be preserved to the extent consistent with public safety. Hazard trees may be removed with approval of the director;
d. Preservation of trees greater than ten inches diameter breast height (dbh) is encouraged;
e. Except where an exception is needed to preserve or create scenic views from county or state roads, internal roads and building locations within a Rural Wooded development should be designed to maximize the extent to which the external buffer obscures the planned development from existing county or state roads; and
f. Where native vegetation is not available to create a sight-obscuring buffer between the planned development and existing county or state roads, fast-growing native vegetation that will grow to obscure the planned development should be planted within the buffer area.
11. In designating the areas for permanent open space tracts, the following is encouraged:
a. Preserve areas along saltwater shoreline;
b. Include open water bodies, creeks, rivers and other natural water features;
c. Protect scenic views and significant natural features;
d. Conserve areas of significant terrestrial wildlife, salmonid habitat, and groundwater supply; and
e. Coordinate with Washington State Department of Fish & Wildlife, Washington State Department of Natural Resources,
*774non-profit agencies, and local tribes to identify priority conservation areas.

Former KCC 17.301.080(E).

¶50 Although the listed items may constitute aspirational design standards, the failure of the ordinance to require the limitations as elements for each development under the Rural Wooded Incentive Program does not ensure that the resulting developments will accord with the protections due rural areas under the GMA or the County’s good intentions. Thus, we hold that substantial evidence does not support the Board’s finding of “ample specific standards here governing the [Rural Wooded Incentive Program] .”AR at 1975.

C. Urban Services

¶51 The Citizens also contend that clusters or multiple clusters of development at high densities will require urban services, but this contention is unsupported in their appellate brief by citation to evidence in the record and was not cited in briefs to the Board, and, after a search of the record, we cannot find evidence supporting the Citizens’ argument. Although these contentions make intuitive sense, the Citizens failed to carry their initial burden before the Board and continue to fail to carry their burden of proof on this issue.26 Because we remand for further consideration of the Rural Wooded Incentive Program, this issue may again become relevant if the parties supply additional information to the Board at future hearings.

IV. Failure of the Board to Decide All Issues

¶52 The Citizens further argue that the Board failed to decide all of the issues presented for review that related to urban density, the land capacity analysis, and the Rural *775Wooded Incentive Program.27 The County contends that the Board’s decision sets forth each issue, reviews the arguments, and sets forth its holdings on each issue, which satisfies RCW 34.05.570(3)(f) and that the Board did not need to set out “a specific ‘holding’ on each argument proffered.” Br. of Resp’t at 59. Although we have already remanded for further action related to these undecided issues because the Board erred and substantial evidence did not support all of its findings, we address these issues for purposes of judicial efficiency as they may remain relevant, depending on the County’s actions on remand.

A. Standard of Review

¶53 Under Washington’s APA, we grant relief “from an agency order in an adjudicative proceeding” where “[t]he agency has not decided all issues requiring resolution by the agency.” RCW 34.05.570(3)(f). The APA does not explain how to apply this standard, but we have interpreted RCW 34.05.570(3)(f) to require, as a threshold matter, that an agency decide an issue before a reviewing court can reach the merits of a party’s legal position28 on appeal. See Low Income Hous. Inst., 119 Wn. App. at 118-19.

¶54 The origins of RCW 34.05.570(3)(f) were veiled in mystery. Section (3)(f) did not exist in the original APA provisions of House Bill 1515 passed by the Washington House of Representatives. See House Journal, 50th Leg., Reg. Sess., at 1358-59 (Wash. 1988). The Senate added the provision and the House passed the amended bill after a free conference.29 House Journal, 50th Leg., Reg. Sess., at 1335, 1472 (Wash. 1988); Senate Journal, 50th Leg., Reg. *776Sess., at 991-92 (Wash. 1988). These proceedings and the final legislative report offer no explanation of the purpose underljdng this provision. See Final Legislative Report, 50th Leg., at 60 (Wash. 1988).

¶55 The relevant language in the original House bill came from a 1961 model act. See Model State Admin. Procedure Act § 15(g), 15A U.L.A. 12 (2000); House Journal, 50th Leg., Reg. Sess., at 1358-59 (Wash. 1988). But the language that the Senate inserted originates in a later model act. See Model State Admin. Procedure Act § 5-116(c) (amended 1981), 15 U.L.A. 145 (2000); Senate Journal, 50th Leg., Reg. Sess., at 991-92 (Wash. 1988). The 1981 model act states, “The court shall grant relief only if it determines that . . . [t]he agency has not decided all issues requiring resolution,” while the 1961 Model Act contains no similar provision. Model State Admin. Procedure Act § 5-116(c)(3), 15 U.L.A. 145; see Model State Admin. Procedure Act § 15(g), 15A U.L.A. 12. The accompanying commentary to the model act clarifies the provision:

Paragraph (c)(3) . . . deals with the possibility that the reviewing court may dispose of the case on the basis of issues that were not considered by the agency. An example would arise if the court had to decide on the facial constitutionality of the agency’s enabling statute, in a state where agencies are precluded from passing on such questions.[30] This provision is not intended to authorize the reviewing court to initially decide issues that are within the agency’s primary jurisdiction; such issues should first be decided by the agency, subject to the limited judicial review provided by this Act.

Model State Admin. Procedure Act § 5-116 cmt., 15 U.L.A. 146 (citation omitted). Thus the provision addresses situations where the agency cannot decide an issue, which the appellate court must decide, and situations where the agency failed to decide an issue within its purview, which the appellate court must remand for additional proceedings.

*777¶56 In Low Income Housing Institute, we held that a board did not decide all of the issues, as required by RCW 34.05.570(3)(f). 119 Wn. App. at 118-19. There, a board stated that the Low Income Housing Institute did not carry its burden to show an inconsistency between the affordable housing element in Lakewood’s comprehensive plan and Pierce County’s countywide planning policies. Low Income Hous. Inst., 119 Wn. App. at 117. We noted that, because of Lakewood’s geographic location, its comprehensive plan needed to consider Pierce County’s countywide planning policies. Low Income Hous. Inst., 119 Wn. App. at 118. We held that the Low Income Housing Institute presented evidence from a final environmental impact statement that the supply of affordable housing was likely to decrease in the city, which was not consistent with Pierce County’s countywide planning policies. Low Income Hous. Inst., 119 Wn. App. at 118. We concluded that, where the reviewing board “made no findings regarding the City’s current need or how the Plan will affect the future availability of affordable housing,” it did not articulate a basis for its decision, and we could not “review its analysis. It has failed to decide all issues requiring resolution as required by [former] RCW 36.70A.290U) [(1997)] and the APA (specifically RCW 34.05.570(3)(f)).” Low Income Hous. Inst., 119 Wn. App. at 118-19. Because we could not review the “analysis on issue number 11 under the summary order presented, we remand [ed] for more thorough findings and articulation of the basis for the ruling.” Low Income Hous. Inst., 119 Wn. App. at 119.

¶57 Only one other case has applied RCW 34.05.570(3)(f) at length: US West Communications, Inc. v. Washington Utilities & Transportation Commission, 134 Wn.2d 74, 949 P.2d 1337 (1997). In this rate case, US West argued that the Washington Utilities and Transportation Commission (Commission) failed to decide all of the issues when it did not consider evidence from a separately decided depreciation case. US West, 134 Wn.2d at 104. Our Supreme Court noted that, under RCW 80.04.200, the rate case did not require a *778rehearing about depreciation where the Commission issued its depreciation case decision only months before hearing the US West rate case. US West, 134 Wn.2d at 104-05. Only after the court held that the Commission “was not obligated to hear the [depreciation] issues again within the two-year period,” dispensing with the threshold matter of whether the Commission addressed all of the issues, did the court move on to the discussion on the merits as to whether (1) the Commission abused its discretion by not rehearing the depreciation evidence, (2) the Commission’s findings of fact and its consideration of the record were complete, and (3) the Commission unlawfully considered certain information in its decision. US West, 134 Wn.2d at 105-07.

¶58 Kitsap County characterizes the Low Income Housing Institute case as dealing with a situation where a board joined an issue with other issues and ruled that the Low Income Housing Institute did not prevail on two issues, and, because the Low Income Housing Institute did not prevail on those two issues, it did not prevail on the other similarly joined issues. But we also held that arguments about issue 11 remained undecided, in part, because the reviewing board erroneously ignored a second element of Pierce County’s countywide planning policies. Low Income Hous. Inst., 119 Wn. App. at 119-21.

¶59 When an agency fails to address an issue or supplies no reason for a decision based on an erroneous legal conclusion that leads an agency either to not decide or to inadequately decide an issue, a legal ground for remand under RCW 34.05.570(3)(f) and further proceedings before the agency arises.31 Where an issue is not decided but remains relevant to the challenged action, the appropriate remedy is to remand for the agency to exercise its judgment and make a decision. See Low Income Hous. Inst., 119 Wn. App. at 118, 121. We do not substitute our judgment and decide factual issues on appeal, but we may “substitute our *779judgment for that of the administrative body on legal issues.” Ames v. Dep’t of Health, Med. Quality Assurance Comm’n, 166 Wn.2d 255, 260-61, 208 P.3d 549 (2009), cert. denied, 130 S. Ct. 1528 (2010); Franklin County Sheriff’s Office v. Sellers, 97 Wn.2d 317, 324-25, 646 P.2d 113 (1982). Yet, because we accord substantial weight to an agency’s view of the law it administers, we remand for the agency to exercise its judgment within its primary jurisdiction and only then will we “substitute” our judgment. See Sellers, 97 Wn.2d at 325-26; William R. Andersen, The 1988 Washington Administrative Procedure Act — An Introduction, 64 Wash. L. Rev. 781, 820 (1989).

B. Urban Density

¶60 The Citizens also argue that the Board did not decide all of the issues when it relied on a bright-line rule to establish acceptable urban density because it did not independently determine whether reducing urban density (1) was consistent with GMA goals and (2) would create internal plan inconsistencies. The Citizens further argue that the County created an inconsistency between the GMA and comprehensive plan goals because the 2006 comprehensive plan promoted increased urban density while also reducing the minimum allowable urban density by 20 percent, an issue that the Board did not address.

¶61 Relying on its bright-line rule of four dwelling units per acre, the Board sidestepped the Citizens’ assertions that “(1) the County has failed to reasonably justify its change in minimum density and the subsequent UGA expansions”; “(2) reduction in residential densities contradicts both County policies and the GMA’s goals and requirements”; and (3) “the County’s rationale is not supported by the GMA’s requirements for urban density, transportation, affordable housing, urban facilities and services, and environmental protection.” AR at 1951-52. The Board noted these arguments but remained firm that its bright-line rule of four dwelling units per acre “is an ‘appropriate’ urban density.” AR at 1951-52. At no time did the Board articulate *780the local circumstances that justified this density, instead it viewed the bright-line rule as a “parameter! ]” within which the County had “discretion to plan for and regulate land use.” AR at 1951-52.

¶62 We remand to the Board for it to consider whether (1) local circumstances show that four dwelling units per acre is an appropriate urban density in Kitsap County at this time, (2) reducing minimum density is internally inconsistent with the comprehensive plan goals, and (3) reducing minimum density is consistent with the GMA’s goals.

C. Land Capacity Analysis

¶63 In addition, the Citizens argue that the Board failed to consider whether (1) the land capacity analysis created internal inconsistencies between the County’s comprehensive plan and the GMA’s goals and (2) the County counted land lost to wetlands and critical areas twice by reducing assumed future density of UGAs and reducing urban land density in the “Urban Restricted Residential designation.” Br. of Appellant at 35. The Citizens further contend that there is an internal inconsistency within the comprehensive plan and an inconsistency with the GMA goals because the land capacity analysis used a four dwelling units per acre minimum while the buildable lands report documented an “actual development trend which is on an upward trajectory from 3.8 [dwelling units per acre] in 1995 to 5.6 [dwelling units per acre] in 2005.” Reply Br. of Appellant at 20.

¶64 The Citizens argued before the Board that the County’s land capacity analysis represented clear error because (1) the urban restricted residential zone, combined with the reduction in the UGAs for critical areas, allowed the County to “double-dipt ]”; (2) the County created inconsistencies with the buildable lands report when it used a minimum density instead of actual existing and trending up average density; and (3) the density reduction created inconsistencies with “density factors and policies stated in the Comprehensive Plan.” AR at 1955.

*781¶65 The Board again based its approval of the minimum density on its unacceptable bright-line rule: “The [land capacity analysis] largely rests upon a residential density assumption of [four dwelling units per acre], which, as the Board has stated supra, is an ‘appropriate’ urban density.” AR at 1956. Furthermore, even if the density of four dwelling units per acre was an appropriate urban density, the Board (1) did not consider whether a reduction in the density of UGAs violated the GMA goals or whether current local circumstances supported using four dwelling units per acre in the land capacity analysis; (2) ignored the Citizens’ argument about the Countys “double-dipp[ing]”; and (3) even agreed “that adopting this approach may dampen the recent success the County has had in encouraging higher densities in the UGAs, since the County concedes that between 2000 and 2005, the County achieved an average of 5.6 units/net acre for urban low density plats.” AR at 1956.

¶66 We remand for the Board to decide, based on current local circumstances, and without reliance on the four dwelling units per acre bright-line rule, whether the County “double-dipped.” If local circumstances support a minimum urban density of four dwelling units per acre, the Board must also decide whether the County creates inconsistencies with the GMA’s goals, the buildable lands report, and the plan when it uses such a minimum density in the land capacity analysis.

D. Rural Wooded Incentive Program

¶67 The Citizens additionally argue that the Board did not decide all of the issues related to the Rural Wooded Incentive Program because it made no findings or conclusions on whether (1) the postdevelopment monitoring provision, done biennially, will protect rural character; (2) the design standards, which can be interpreted in multiple ways and are aspirational, provide adequate protection for rural character; and (3) the vesting law undercuts the monitoring provision. They also argue that the Board did not determine whether the possible Rural Wooded Incen*782tive Program developments would constitute urban growth or would destroy rural character. The Citizens press the urgency for the Board to review and decide issues relating to potential developments under the Rural Wooded Incentive Program because development rights may vest for significant numbers of homes if the comprehensive plan is approved as the County proposed and now is the only time to challenge these developments.

¶68 We have already decided that substantial evidence did not support the Board’s decision that the Rural Wooded Incentive Program would adequately ensure protection of rural character based on former KCC 17.301.080(E)’s design standards.

¶69 In addressing the Citizens’ challenge to the monitoring program, the Board relied on the standards throughout former KCC 17.301.080 but did not discuss how these development standards would ensure preservation of rural character when the language did not command compliance, nor did the Board discuss the vagueness of the language used; instead, the Board stated that the “biennial monitor [ing] program also works to ensure maintenance of the qualities of rural character.” AR at 1978. To determine whether the Board decided the issues the Citizens raised with regard to preserving the rural character, we now turn to the monitoring program.

¶70 The Citizens argued to the Board that “the monitoring and approval process does not thwart negative effects of incremental development, since it is ‘after the fact’ monitoring.” AR at 1968-69. The Citizens contended that the monitoring program did not account for vested rights and noted that these rights could allow problematic developments to continue because the County’s monitoring program might not detect adverse impacts until well after development rights vested.

¶71 Vesting “refers to the ‘notion that a land use application, under the proper conditions, will be considered only under the land use statutes and ordinances in effect at the time of the application’s submission,’ ” and it “determine [s] *783and fix[es] the rules that will apply to a land development.” E. County Reclamation Co. v. Bjornsen, 125 Wn. App. 432, 437, 105 P.3d 94 (2005) (quoting Friends of The Law v. King County, 123 Wn.2d 518, 522, 869 P.2d 1056 (1994)).

¶72 The Board knew that the Rural Wooded Incentive Program allowed vesting because the Board included the entire provision in an appendix. See former KCC 17.301-.080(1). The Board did not address this argument — a valid argument given that vested rights in nonconforming, preGMA rural parcels, the “legacy lots,” already prevented Kitsap County from controlling rural growth. But the Board ruled that “the initial scope of the [Rural Wooded Incentive P]rogram [Phase I — 5000 acres] is modest and reasonable and not clearly erroneous.” AR at 1973 (boldface omitted) (second alteration in original).

¶73 The Board’s assertion that the first phase of the Rural Wooded Incentive Program is “modest and reasonable” does not address the Citizens’ concern that development rights could vest and the County could release subsequent phases before the monitoring program detects problems with implementation of the Program. Thus, we hold that the Board did not decide all of the issues when it did not address the effects of vesting development rights on the County’s monitoring program’s effectiveness.

¶74 We also hold that the Board did not fully decide whether the monitoring program preserved rural character because the Board assumed that the monitoring program would maintain rural character. The Board’s passing discussion of this issue does not fully address whether the Rural Wooded Incentive Program’s control by aspirational standards and postdevelopment monitoring would actually preserve rural character.

*784E. Goal Harmonizing Document

¶75 Finally, the Citizens argue that substantial evidence32 does not support the Board’s finding of compliance for the goal-harmonizing document that the County produced, in which the County intended to explain how the rural element portion of the comprehensive plan “[hjarmoniz[es]” with the GMA’s planning goals. Br. of Appellant at 58. The Citizens also claim that, because the Board did not consider existing and potential “urban sprawl” development adjacent to the Rural Wooded Program lands, substantial evidence does not support the Board’s decision to accept the document as a satisfactory demonstration that the program harmonized with the GMA goals. Because the Citizens also question whether the Board decided all of the issues related to the goal harmonizing document, we review these issues under RCW 34.05.570(3)(f).

¶76 The County contends that the Board addressed this issue, which the County believes is irrelevant, and that the Board noted that nonconforming lots are generally under five acres in size, which would make participation in the Rural Wooded Incentive Program a boon for the County. The Citizens respond that the Board committed an “error of fact” when it “failed to consider problematic local circumstances [for parcels] surrounding the rural lands in which [Rural Wooded Incentive Program] developments ... occur” and failed to consider the impact of combining these local circumstances with the Rural Wooded Incentive Program. Reply Br. of Appellant at 38. The Citizens specifically noted that the final goal-harmonizing document did not consider the combined impact of the Rural Wooded Incentive Pro*785gram and the existing pattern of urban growth within the rural area, made possible by the existing legacy lots.

¶77 The Board characterized the Citizens’ argument as asserting that “while the Board is aware of the local circumstances present in rural Kitsap County, the County neglected to address them in the context of this document. [Citizens] then recite prior holdings of this Board pertaining to noncompliant rural densities.” AR at 4339 (citation omitted). Though the Board remanded to the County for a “simple explanation [of local circumstances] to be added,” the Board rejected the Citizens’ “continued questioning of rural densities . . . addressed in the [final decision and order’s] discussion of clustering, the bottom line being that the clustering provisions yielded rural densities.” AR at 4339. The Board then quoted language from the final decision and order that concluded one dwelling unit per five acres to be “ ‘a rural, not urban, density [ ] that is consistent with preserving the rural character. The Board acknowledges that the clustered design of the development appears more dense when viewed in isolation, but it is nonetheless a rural density when viewed in the context of the entire parcel.’ ” AR at 4340 n.7.

¶78 After the County added further explanation of local circumstances following the second remand to the County for expansion of the goal-harmonizing document, the Citizens renewed their argument that the updated goal-harmonizing document failed to address the local circumstance of the existing legacy lots and the impact of adding more small lots through the Rural Wooded Incentive Program.

¶79 The Board responded to the Citizens’ discussion of the legacy lots, stating that “the County’s treatment of legacy lots is beyond the scope of the Board’s review for determining compliance on the limited remand issue for the [Rural Wooded Incentive Program].” AR at 4755. The Board then found that the County’s additions to the goal-harmonizing document on the second remand merited a finding of compliance.

*786¶80 Because the Board improperly relied on a whole parcel calculation method and its bright-line rule for rural density, we hold that the Board did not decide whether the County clearly erred when the description of local circumstances in the goal-harmonizing document did not include a discussion of the impact on the rural element of higher density development under the Rural Wooded Incentive Program when combined with development on the existing legacy lots.

¶81 We reverse, in part, those orders by the Board that Citizens challenge on appeal33 and remand for further action. We hold that the Board erred when it used bright-line rules to find that the reduced urban density, the land capacity analysis, and the Rural Wooded Incentive Program’s rural density component complied with the GMA. We also hold that substantial evidence does not support the Board’s decision that the existing design standards in former KCC 17.301.080 protect rural character. And we hold that the Board did not decide all of the relevant issues that the Citizens raised; thus, we remand for the Board to decide all remaining relevant issues, including those regarding the reduction in urban density, the use of minimum density in the land capacity analysis, preservation of rural character, and the local circumstances supporting the goal-harmonizing document.

Penoyar, C.J., and Armstrong, J., concur.

Review denied at 170 Wn.2d 1019 (2011).

We refer to the parties collectively as the “Citizens.’

In 2006, we affirmed the Board’s order that required the County to take additional steps because the 2002 buildable lands report was inconsistent with the County’s planning policies and comprehensive plan, which were then in effect. See Kitsap County v. Cent. Puget Sound Growth Mgmt. Hearings Bd., 138 Wn. App. 863, 876-77, 881, 158 P.3d 638 (2007).

The 2006 comprehensive plan contained preliminary information, later confirmed in the completed 2007 buildable lands report, showing that the urban/rural growth rates had changed, so that 57 percent of growth occurred in UGAs and 43 percent occurred in rural areas. The 2007 buildable lands report also showed the average density for newly permitted developments was 5.6 dwelling units per acre in the UGA’s cuban low residential zone from 2000 to 2005.

The revised 2006 comprehensive plan assumed development at the lowest density available in each urban zoning category to calculate the amount of land needed for the total UGA.

The effect of the 2006 comprehensive plan was to expand the UGA boundaries by 12.7 square miles, thereby reducing Kitsap County's rural areas subject to the plan to accommodate the projected growth using these minimum densities.

Kitsap County contains 49,333 acres of “rural wooded” land. AR at 2815.

A household septic system’s drainfield size would not constrain the minimum lot size because the program contemplates group drainfields.

The Suquamish Tribe raised most of these issues in its petition.

9 Petitioners Kitsap Citizens for Rural Preservation and Harless also questioned the Rural Wooded Incentive Program’s compliance with the GMA’s (1) urban growth, antisprawl, and natural resource goals and (2) requirement for measures to govern rural development.

Board Member Margaret Pageler dissented from the urban density issue based on her reading of Viking Properties, Inc. v. Holm, 155 Wn.2d 112, 118 P.3d 322 (2005) and its prohibition on use of bright-line rules.

11 City of Bremerton v. Kitsap County, No. 04-3-0009c, 2004 WL 3249863, at *17-18, 2004 GMHB LEXIS 61, at **44-48 (Cent. Puget Sound Growth Mgmt. Hr’gs Bd. Aug. 9, 2004) (Bremerton II).

To support its finding, the Board emphasized “again that in the hypothetical situation, the most generous clustering provisions yields a net density of 1 [dwelling unit per ]5 acres on a 100 acre parcel - this is not an urban density or urban growth.” AR at 1976 n.16.

We and the parties use biennial review to mean review once every two years — as opposed to biannual review — which means review twice a year. Webster’s Third New International Dictionary 213, 211 (2002).

The Board also noted, “[Citizens’] continued questioning of rural densities was addressed in the [Final Decision and Order] in the discussion of clustering, the bottom line being that the clustering provisions yielded rural densities.” AR at 4339.

15 The Board did not refer to information in the record showing that development in the rural areas occurred at average densities far greater than the most generous zoned densities of one dwelling unit per five acres: Rural development ranged from 0.39 dwelling units per acre (the equivalent of one dwelling unit per 2.54 acres) to 2.74 dwelling units per acre.

At times, we refer to the growth management hearings boards generically as “board” and ‘boards.” We do not intend to cause confusion with an array of other administrative bodies.

Amicus Olympic Property Group also argues that the Board did not apply a bright-line rule because it did not explicitly label its action as such.

Some have described Bremerton I as the “seminal” case about urban bright-line densities. Joseph W. Tovar, Puget Sound Reg’l Council, Appropriate Urban DENSITIES IN THE CENTRAL PüGET SOUND REGION: LOCAL PLANS, REGIONAL VISIONS, AND THE Growth Management Act 12 (Nov. 2, 2005), http://www.psrc.org/assets/2030/ appIP12-urban-densities.pdf. In Viking Properties, our Supreme Court singled out the Board’s decision in Bremerton I, referred to it as a “bright-line” — setting a minimum density for urban development — and disclaimed the authority of the boards to make such public policy. Viking Props., 155 Wn.2d at 128-29. But even after Viking, one board decision recharacterized a bright line as a “ ‘safe harbor,’ which is to say that it is Em irrefutably valid appropriate urban density.” Tovar at 18 (quoting Fuhriman v. City of Bothell, No. 05-3-0025c, 2005 WL 3067434, at *17, *7652005 GMHB LEXIS 111, at *47 (Cent. Puget Sound Growth Mgmt. Hr’gs Bd. Aug. 29, 2005)).

Amicus Olympic Property Group contends that possible Rural Wooded Incentive Program developments, like its “String of Pearls Incentive,” should be permissible because they reserve the remaining 75 percent of land in permanent open space. Br. of Amicus at 17. Olympic Property Group admits, however, that “these tracts in some instances may resemble large, suburban lots,” but that these lots will have open space. Br. of Amicus at 17.

Because we hold that the Board used a bright-line rule and did not consider local circumstances, we leave for the Board to consider on remand (1) whether actual density can exist within a cluster that can be too high because the Rural Wooded Incentive Program also counts land outside the cluster to make the development appear less dense and (2) whether the clusters or groups of clusters allowed by the program actually allow urban growth outside the UGA.

Cassandra warned that the Grecian horse spelled Troy’s doom, but Apollo’s curse prevented her hearers from believing the prophecy. Vergil, The Aeneid 75-76, 83 (Robert Fagles trans., Viking Penguin 2006); Aeschylus, Agamemnon, in The Oresteia 150-52 (Robert Fagles trans., Penguin 1979). We leave to another day the casting of Apollo and the gift-bearing Greeks in Washington’s GMA epic.

As the Board relied on bright-line rules for urban density and the record does not demonstrate that it considered local circumstances to support its decision, we do not reach the Citizens’ arguments that it erred or that substantial evidence does not support its conclusions related to the County’s decision to (1) lower the minimum urban density and (2) use the minimum urban density in the land capacity analysis.

Because the Board did not address the Citizens’ monitoring argument as it relates to vesting and the design standards, we cannot review its decision on this issue as we discuss further in section IV.D.

Amicus Olympic Property Group makes a similar argument.

Although a subset of the design standards, the need for urban services is a independent issue under the GMA’s rural element. RCW 36.70A.070(5)(b); see also former RCW 36.70A.030 (15)-(19) (2005).

The Rural Wooded Incentive Program ordinance also states that a hearings examiner may grant approval to a development only if the “proposed development will not require the extension or provision of sanitary sewer service or other urban services to the development.” Former KCC 17.301.080(H)(2).

We also review whether the Board decided all of the issues related to the goal harmonizing document because Citizen’s also characterized the decision as failing to decide all of the issues.

Where an agency decides an issue and the legal and factual basis of that decision remains intact on appeal, we do not grant relief under section RCW 34.05.570(3)(f). See, e.g., Yakima Police Patrolmen’s Ass’n v. City of Yakima, 153 Wn. App. 541, 564-65, 567, 222 P.3d 1217 (2009).

A “free conference” allows a joint legislative conference committee to confer and debate the merits of different legislative approaches without “being confined *776to reasons adopted by either of the Houses.” Thomas B. Reed, Reed’s Rules: A Manual op General Parliamentary Law § 242, at 181 (1900).

30 Washington’s APA does not contain a similar provision. See Model State Admin. Procedure Act § 5-112(1), 15 U.L.A. 137.

Where an agency decides an issue in the alternative, remand may not be necessary under RCW 34.05.570(3)(f).

Although the Citizens argue that the Board’s finding of compliance lacked substantial evidence, we review this issue to ascertain whether the Board decided all of the issues in light of our holding that the County erred when it looked at the whole parcel and relied on the bright-line rule that one dwelling unit per five acres is a rural density without considering the local circumstances — i.e., whether the Rural Wooded Incentive Program’s clustering provisions could create urban growth when viewed without the open space, which reduces the apparent density of the development.

We address only those matters raised on appeal; we do not address the Board’s unchallenged decisions.