KISTLER, J.

The question in this case is whether a landowner holding a Measure 37 waiver had a common law vested right to construct a residential subdivision that he had begun but not completed by the effective date of Measure 49. Yamhill County found that the costs that the landowner had incurred were sufficient to establish a vested right to complete construction of the subdivision, and the circuit court upheld the county's decision on a writ of review. The Court of Appeals reversed the circuit court's judgment and remanded the case for further proceedings. Friends of Yamhill County v. Board of Commissioners, 237 Or.App. 149, 238 P.3d 1016 (2010). The Court of Appeals started from the proposition that, in the context of Measure 49, a common law vested right turns primarily on the ratio between the costs that a landowner has incurred and the projected cost of the development. It reversed because the county had given too little weight to that factor. We allowed the landowner's petition for review to clarify the standard for determining when, in the context of Measure 49, a common law right to complete a development will vest. We now affirm the Court of Appeals decision, although for different reasons than those stated in the Court of Appeals opinion.

Before turning to the facts of this case, we first discuss briefly the statutory context in which this issue arises. Comprehensive zoning laws first emerged in the early part of the twentieth century. See Village of Euclid v. Ambler Realty Co., 272 U.S. 365, 386-87, 47 S.Ct. 114, 71 L.Ed. 303 (1926) (discussing the origins of zoning laws); Eugene McQuillin, 8 The Law of Municipal Corporations § 25:3, 12-13 (3d ed. 2010) (same). Before then, local governments regulated the location of certain nuisance uses but did not control the use of land within their jurisdictions on a comprehensive basis. McQuillin § 25:3 at 12-13. As a result of increasing urbanization, local governments sought to organize more effectively the variety of different and, at times, incompatible uses of land within their communities. See Patricia E. Salkin, 2 American Law of Zoning § 12:2, 12-7 (5th ed. 2008); Village of Euclid, 272 U.S. at 386-87, 47 S.Ct. 114 ("Until recent years, urban life was comparatively simple; but with the great increase and concentration of population, problems have developed, and constantly are developing, which require, and will continue to require, additional restrictions in respect of the use and occupation of private lands in urban communities.").

To that end, state legislatures began enacting enabling legislation that authorized local governments to pass comprehensive zoning ordinances. See, e.g., Or. Laws 1919, ch. 311 (permitting municipalities to enact comprehensive zoning ordinances).1 In exercising that authority, local governments retained relative autonomy to craft the planning and zoning policies that governed their communities. See Hector Macpherson & Norma Paulus, Senate Bill 100: The Oregon Land Conservation and Development Act, 10 Will L J 414, 414 (1974). By the latter half of the twentieth century, however, that decentralized zoning authority had yielded, at least in Oregon, to statewide coordination of the standards governing land use. See id. at 415. In 1973, the Oregon legislature enacted Senate Bill 100, which established statewide land use planning goals and delegated the authority to modify and enforce those goals to a state agency. See generally ORS ch 197 (establishing the Land Conservation and Development Commission and setting out statewide land use planning goals).2

Oregon's statewide land use planning system did not come without controversy, however. In some parts of the state, implementing the statewide goals resulted in greater restrictions on the use of property than local zoning law previously had imposed. See Edward J. Sullivan & Jennifer M. Bragar, The Augean Stables: Measure 49 and the Herculean Task of Correcting an Improvident Initiative Measure in Oregon, 46 Will L. Rev. 577, 577-78 (2010). Some persons came to view the statewide goals, which were calculated to protect farm- and forest-resource lands, as unfairly limiting the rights of landowners who acquired their property before those restrictions were put in place. See id. In particular, they were concerned with the limitations that the land use laws placed on a landowner's ability to build homes on his or her land. Id.

Those concerns culminated in Ballot Measure 37. Or. Laws 2005, ch. 1, codified at former ORS 197.352 (2005). That measure provided landowners with "just compensation" for land use regulations, enacted after they had acquired their property, that restricted the use of the property and, as a result, diminished its value. See State ex rel. English v. Multnomah County, 348 Or. 417, 420-22, 238 P.3d 980 (2010) (describing Measure 37). When faced with a claim for "just compensation" under Measure 37, a government could choose: (1) to pay the landowner compensation for the diminished value of the property and enforce the regulation or (2) to waive the regulation and permit the owner "to use the property for a use permitted at the time the owner acquired the property." Former ORS 197.352(8) (2005).

Measure 37 also was not without controversy. Some believed that the measure went farther than many voters had intended in that it not only permitted landowners to build a small number of additional homes on their property, unrelated to the resource use of the land, but it also authorized the large-scale development of formerly protected lands. See Official Voters' Pamphlet, Special Election, Nov. 6, 2007, 20 (Legislative Argument in Support of Measure 49). In response to those concerns, the 2007 Legislative Assembly considered several draft bills intended to reform Measure 37. After several public hearings, those draft bills were consolidated into a single bill, House Bill 3540 (2007). See Tape Recording, Joint Special Committee on Land Use Fairness, HB 3540, Apr. 26, 2007, Tape 50, Side A (statement of Sen. Greg Macpherson). The legislature did not enact HB 3540 directly; instead, it referred the proposed legislation to the voters on June 15, 2007, as Ballot Measure 49. See Or. Laws 2007, ch. 750, § 2 (referring HB 3540 to the voters). In a special election held on November 6, 2007, the voters approved Measure 49 and, on December 6, 2007, the measure became effective.

Among other things, Measure 49 retroactively extinguished previously issued Measure 37 waivers of land use regulations. See Corey v. DLCD, 344 Or. 457, 466-67, 184 P.3d 1109 (2008) ("Measure 49 by its terms deprives Measure 37 waivers—and all orders disposing of Measure 37 claims—of any continuing viability"; emphasis in original). As a result, landowners who had begun developing their property under authorization granted by Measure 37 waivers could no longer automatically continue to do so. Instead, those landowners had to choose one of three alternative "pathways" moving forward: an "express pathway," a "conditional pathway," and "a third pathway for claimants that have vested rights to carry out claims that have already been approved." Tape Recording, Joint Special Committee on Land Use Fairness, SB 1019, Apr. 19, 2007, Tape 43, Side A (statement of Richard Whitman, Oregon Department of Justice, summarizing the proposed "framework" for amending Measure 37); see Or. Laws 2007, ch. 424, § 5 (setting out those three alternatives).

The express pathway entitles a landowner to obtain development approval for up to three additional homes on his or her property. See Or. Laws 2007, ch. 424, § 5(1) (identifying express pathway). Under the conditional pathway, a landowner can obtain approval for four to 10 homes if, among other conditions, the land use regulations prohibiting the construction of those homes resulted in a specified reduction of the fair market value of the property. See id. §§ 7 and 9 (setting out conditional pathway and describing conditions). Finally, the vested rights pathway permits a landowner who had obtained a Measure 37 waiver to "complete and continue the use described in the waiver," provided that the landowner could also demonstrate a "common law vested right" to complete that use. Id. § 5(3).

This case involves the third pathway that Measure 49 identifies. Gordon Cook owns approximately 40 acres of agricultural land in Yamhill County. Following the passage of Measure 37, Cook filed with the state and county written demands for compensation pursuant to that measure. Cook claimed that county land use regulations enacted after he acquired his property in 1970 both prevented him from developing a 10-lot subdivision on his land and also diminished the fair market value of that property by $1.6 million. The state explained that, in his written demands, Cook sought compensation in that amount or, alternatively, "the right to divide the 38.8-acre subject property into nine 2.69-acre parcels and one 12.4-acre parcel and [to] develop a dwelling on each 2.69-acre parcel." Instead of paying Cook monetary compensation, the state and county issued waivers of land use regulations in 2006, agreeing not to enforce the land use regulations that prevented Cook from developing a residential subdivision.3 See former ORS 197.352 (2005) (permitting that relief).

Having received those waivers, Cook could proceed to the first stage of his proposed development—that is, dividing his property "into nine 2.69-acre parcels and one 12.4-acre parcel[.]" To do so, Cook had to initiate a two-stage process for obtaining the county's approval of his proposed subdivision. See generally ORS 92.010 to 92.179 (setting forth procedures for the subdivision and partition of land). First, Cook had to apply for and obtain preliminary subdivision approval. See id. If he obtained preliminary approval of his proposed subdivision plan, then he could seek to satisfy any conditions to which final approval would be subject and, on completing those conditions and preparing and submitting a final plat of his proposed subdivision, he could seek final subdivision approval from the county. See id. For the county's part, that process permitted it to ensure that Cook's proposed development was "consistent with the [county's] comprehensive plan and applicable land use regulations." ORS 197.522 (requiring counties to approve subdivision plans if they are consistent with local land use regulations). Accordingly, based on the Measure 37 waivers issued by the state and the county, Cook submitted to the county an "application for preliminary subdivision approval."

Because Cook's application was predicated on the state and county's waiver of certain otherwise applicable land use regulations, in reviewing Cook's application, the county "k[ept] in mind that many of the standards [then in effect] would not [have been] applied when [Cook] first acquired the property." Instead, Cook's application was reviewed for compliance with the requirements of the zoning ordinance in effect in 1970.4 That ordinance, which the record refers to as "Ordinance 29," classified Cook's property as property subject to the county's former "A-1" agricultural zone. The A-l zone authorized as a permitted use a "[d]welling or dwellings for the owner, operator, or employe[e]s required to carry out" a farm- or forestry-resource use. It also authorized "[s]ingle-family dwelling" as a conditional use, provided that the dwelling was "developed in accordance with the requirements" of the county's R-S residential suburban zone.5

In reviewing Cook's application for preliminary subdivision approval, the county acknowledged that "Ordinance 29 did not establish minimum lot sizes within" the A-l zone. That is, unlike the restrictions set forth for other zoning districts under Ordinance 29, the ordinance did not identify the minimum size lot on which a dwelling could be sited as either a permitted use or as a conditional use in the A-l zone. With the apparent expectation, in light of Measure 37, of adjudicating land use applications seeking to develop property previously zoned as A-l Agriculture property, "[o]n April 2, 2007 the Board of Commissioners met for the purpose of interpreting Ordinance 29." At that meeting, the county adopted Board Order 07-289, which states, in part:

On May 11, 2007, after completing its review of Cook's application, the county granted him preliminary subdivision approval and conditional use approval to begin developing his 10-lot subdivision. The county's order approving Cook's application for preliminary subdivision approval also set forth the requirements that Cook would have to satisfy to obtain final subdivision approval from the county. No interested party appealed the county's approval of Cook's preliminary subdivision plan and conditional use.

Having obtained preliminary subdivision approval, Cook began developing his property. His efforts included clearing, excavating, and grading the property, as well as taking steps to satisfy the conditions to which final subdivision approval would be subject. Those conditions required Cook to retain a private firm to evaluate and plan a system for "on-site subsurface sewage disposal" for the nine smaller lots, and also required Cook to arrange for water and electrical power to the subdivision. Having begun development following his preliminary subdivision approval on May 11, 2007, Cook continued developing his property through the remainder of 2007. On December 5, 2007, the day before Measure 49 became effective, Cook obtained the county's approval of his final subdivision plat and recorded it. No interested party appealed the county's approval of Cook's final plat.

Were it not for Measure 49, Cook would have been entitled to proceed with the second stage of his proposed subdivision development —that is, "developing] a dwelling on each 2.69-acre" lot. As noted, however, Measure 49 retroactively invalidated Measure 37 waivers of land use regulations, and landowners such as Cook, who had been proceeding with a development under the authorization of a Measure 37 waiver, could no longer automatically continue to do so. Instead, they had to choose among three pathways: the express pathway, the conditional pathway, and the vested rights pathway. Cook chose the vested rights pathway. Accordingly, he sought a determination from the county that he had "a common law vested right" to complete the development contemplated in his Measure 37 waivers.6

To that end, Cook submitted an application for a county vesting determination pursuant to an ordinance that the county had enacted to implement the vested rights pathway of Measure 49.7 In his application, Cook provided the county with documentary evidence of his expenditures in seeking preliminary and final subdivision approval as well as expenses that he had incurred in developing his property. Specifically, Cook documented his legal costs; surveying and engineering costs; costs incurred in excavating, grading, and conducting other subsurface or surface-level development of his property; costs incurred in arranging for the provision of electrical power to the subdivision, in securing permits for the development, and in testing and planning septic systems for each homesite; and money paid to record Cook's final plat and to pay property taxes "imposed upon disqualification from special assessment."8

Based on that evidence, Cook provided three calculations of his total expenditures. Cook represented that, by June 28, 2007, a short time after the legislature had referred Measure 49 to the voters, he had expended a total of $120,494.06. On November 6, 2007, the day on which the voters approved Measure 49, Cook's calculations showed that his expenses totaled $143,764.85. Finally, when Measure 49 became effective one month later on December 6, 2007, Cook calculated that he had spent a total of $155,160.53 in developing his property.

Cook also included with his application estimates of what, in his view, the total cost of completing his proposed development would be. Originally, Cook had planned to develop only "finished lots." That is, Cook had planned to develop his property to the point where each new lot on the property was suitable for the construction of a dwelling, but Cook had not planned on constructing any of those dwellings himself. Instead, Cook had planned on selling the finished lots to developers or individuals who could then, as Cook saw it, arrange for the construction of dwellings on the lots on their own. As a result, Cook estimated that the total cost of completing his development—that is, the total cost of developing nine buildable lots— would be $204,660.53.

As Cook acknowledged in his application, however, "[s]ome opponents of Measure 37 have argued that the `project' must include the [cost of constructing] homes as well as the [cost of developing] finished lots." Although Cook disagreed with that view of Measure 37, he also offered an estimate of his project's total cost, including the cost of placing dwellings on the lots. Specifically, Cook added the cost of purchasing small, manufactured homes to be placed on the lots and the cost of providing necessary services to those dwellings. Under that modified characterization of his proposed development, Cook estimated that completing his proposed subdivision would cost a total of $871,158.54.

The county's ordinance permitting Cook to apply for a vested rights determination provides that "[p]ersons other than the [a]pplicant" may "submit written evidence, arguments or comments * * * for consideration by [the county]." Pursuant to that ordinance, Friends of Yamhill County filed a memorandum and exhibits in opposition to Cook's vested rights application. Among other exhibits, Friends of Yamhill County included printouts of online real estate listings, which advertised several of the lots on Cook's property for sale for between $325,000 and $329,000, and another listing that advertised a "[p]roposed new home by [a] renowned builder," to be built on one of Cook's lots, which could be purchased for $1.29 million. That advertisement depicted a sizeable permanent dwelling, somewhat dissimilar from the manufactured homes described in Cook's estimate of his total cost.

After reviewing the exhibits submitted by Cook and Friends of Yamhill County, the county determined that Cook had obtained a vested right to complete his proposed subdivision. The county rested its conclusion on two alternative bases. The county concluded initially that, because Cook had obtained both preliminary and final plat approval from the county, neither of which had been appealed, Cook had obtained "governmentally approved homesites that could be deeded to others," and, as a result, he had acquired a vested right to "develop and/or sell every one of the lots" before Measure 49 went into effect.

Alternatively, the county concluded that Cook had a vested right to complete his subdivision, based on a six-factor test set out in Clackamas Co. v. Holmes, 265 Or. 193, 508 P.2d 190 (1973). The parties primarily disputed three of the six Holmes factors. One of those factors required the county to examine the ratio that Cook's expenses bore to the cost of completing his development. A second factor required the county to determine whether Cook had begun developing his property after he received notice of Measure 49's pendency. The third factor required the county to assess Cook's good faith generally in proceeding with his development.

Considering the first Holmes factor, the county found that Cook's expenses were "substantial," but it did not determine the total cost of completing the project. In the county's view, doing so would be "a speculative analysis that may be necessary in other cases in which [the ratio] factor deserves to be given more weight, but is not necessary to properly address [that factor] in this case." The county also rejected Friends of Yamhill County's contention regarding the second factor. The county found that, because Cook had begun developing his property before the legislature referred Measure 49 to the voters, that factor did not provide any reason to say that Cook had acted in bad faith. Finally, regarding "good faith" generally, the county explained that, "[b]ased solely on [Cook's] compliance with (and reasonable reliance upon) all applicable law in establishing his use, [Cook] has established that all of his expenditures * * * were in good faith." The county concluded that, under Holmes, Cook had done enough to establish a vested right to complete his subdivision.

Friends of Yamhill County filed a petition for a writ of review in the Yamhill County Circuit Court, seeking review of the county's decision. See Or. Laws 2007, ch. 424, § 16, codified at ORS 195.318 (providing for judicial review of county decisions under section five of Measure 49). The circuit court issued the writ, ordering the county to return to the circuit court the record before the county. After reviewing that record, the circuit court upheld the county's ruling that Cook's proposed use complied with the state and county waivers, and it upheld the county's alternative holding that Cook had expended sufficient costs toward construction to obtain a vested right to complete the construction of homes on the lots.

On the first issue, the circuit court reasoned that there was substantial evidence in the record that, in 1970, a residential subdivision was a permissible conditional use of land zoned A-1. It followed, the court concluded, that the use complied with the waivers. On the second issue, the court rejected the county's primary ground for finding a vested right, which assumed that Measure 37 rights are transferrable. The circuit court reasoned that "[t]he fact that individual lots can be transferred after plat approval does not mean that the new owner has any vested right to use that lot or parcel in a nonconforming way." The circuit court recognized, however, that "a use fully established prior to the Measure 49 deadline or partially completed but vested under common law analysis is transferable as a nonconforming use." (Emphases omitted.) The circuit court accordingly focused on whether Cook had incurred sufficient costs toward the construction of the homes that he had a common law vested right to complete their construction. On that question, the circuit court agreed with the county's alternative conclusion that Cook had established a vested right to complete construction of the homes under the six-factor test set out in Holmes.

Friends of Yamhill County appealed the circuit court's judgment.9 On appeal, the Court of Appeals agreed with Friends of Yamhill County that the circuit court had erred in two respects. Friends of Yamhill County, 237 Or.App. at 178, 238 P.3d 1016. First, the Court of Appeals concluded that the county had erred in not determining whether a residential subdivision was a permissible use under the zoning ordinances in effect when Cook acquired his property and that the circuit court should have remanded for that determination. Id. at 171-72, 178, 238 P.3d 1016. Second, the Court of Appeals held that, in determining whether Cook had a vested right to complete construction of the homes, the county had failed "to determine the extent and general cost of the project to be vested and to give proper weight to the expenditure ratio factor" and that the circuit court had erred in upholding the county's determination. Id. at 178, 238 P.3d 1016.

In reaching the latter conclusion, the Court of Appeals recognized that Measure 49 provides that a landowner can complete a partially finished development when he or she has "a common law vested right" to do so. Id. at 174, 238 P.3d 1016. The court also recognized that this court's decision in Holmes sets out a six-factor test for determining when a common law vested right to complete a development will exist. Id. at 161, 238 P.3d 1016. The court reasoned, however, that, in the context of Measure 49, some of the Holmes factors are redundant while others are "more material." Id. at 175-77, 238 P.3d 1016 (emphasis in original). Specifically, the court concluded that, in the context of Measure 49, "a common law vested right" will turn primarily on two of the Holmes factors: (1) the ratio between the expenditures incurred and the cost of the project and (2) the cost and location of the project. Id. at 177, 238 P.3d 1016.10 Having concluded that the expenditure ratio is a "more material" factor in a Measure 49 vested rights determination, the court faulted the county for not defining the projected cost and thus not giving proper weight to that factor. Id. at 178, 238 P.3d 1016.

On review, Cook raises three issues. He argues initially that whether he had a common law vested right to complete his development presents a question of fact and that substantial evidence supports the county's factual finding on that issue. In his view, that should be the end of the matter. Second, Cook contends that the Court of Appeals erred in concluding that, as used in Measure 49, the statutory phrase "common law vested right" depends particularly on one of the Holmes factors—the ratio between a landowner's expenditures and the development's projected cost. Finally, he argues that the Court of Appeals erred in remanding this case to determine whether the Yamhill County zoning ordinances in effect in 1970 would have permitted him to build a residential subdivision on land zoned for agricultural use.

We begin with the second issue that Cook raises—whether the Court of Appeals erred in interpreting the statutory phrase "a common law vested right."11 Resolution of that issue entails three related but separate inquiries. The first requires us to identify the body of common law to which Measure 49 refers. The second requires us to identify the content of that body of law. The third is whether the Court of Appeals correctly determined that, in the context of Measure 49, only some common law factors play a role in deciding whether a landowner has a vested right.

The first question poses little difficulty. It is true, as the Court of Appeals noted, that Oregon does not follow the majority rule in determining when a landowner will have a common law vested right to complete a partially finished use of property. See Friends of Yamhill County, 237 Or.App. at 160, 238 P.3d 1016.12 However, we agree with the parties that, in using the phrase "a common law vested right," Measure 49 was referring to the common law of Oregon rather than the majority rule. We said as much in Corey when we explained that the phrase "a common law vested right" in Measure 49 refers to "broadly applicable legal precedents describing a property owner's rights when land use laws are enacted that make a partially finished project unlawful" and identified this court's decision in Holmes as an example of those precedents. See Corey, 344 Or. at 466, 184 P.3d 1109. We conclude that, without an explicit reference in Measure 49 to some other body of common law, the voters who approved Measure 49 intended to adopt Oregon common law as the standard for identifying a vested right.

Having identified Oregon law as the body of law to which Measure 49 refers, we turn to the content of that law. This court has used the phrase "vested right" in the context of land use regulation to describe a subconstitutional conclusion that a landowner is entitled either to continue a preexisting use or to complete a partially finished one. See Polk County v. Martin, 292 Or. 69, 74, 636 P.2d 952 (1981) (preexisting use); Holmes, 265 Or. at 197, 508 P.2d 190 (partially constructed project). In Polk County, the court used the phrase to refer to an existing use that, as a result of a zoning change, had become unlawful. See 292 Or. at 74, 636 P.2d 952. As the court explained in Polk County,

Id.

In addition to recognizing that a landowner may have a vested right to continue an existing use, this court also has held that a landowner may have a vested right to complete a use that was begun but not finished when the government enacted laws prohibiting the use. See id. at 82, 636 P.2d 952; Holmes, 265 Or. at 197, 508 P.2d 190. The question in the latter class of cases typically has been how much must a landowner have done before the right to complete the use will vest. In Holmes, this court answered that question by explaining that "in order for a landowner to have acquired a vested right to proceed with the construction, [either] the commencement of the construction must have been substantial, or substantial costs toward completion of the job must have been incurred." 265 Or. at 197, 508 P.2d 190.

Saying that the costs incurred or the amount of construction begun must be "substantial" narrows the inquiry somewhat, but it leaves unanswered the question of how much is substantial. In addressing that question, the Holmes court began by considering whether to follow a New York decision, Town of Hempstead v. Lynne, 32 Misc.2d 312, 222 N.Y.S.2d 526 (1961). See Holmes, 265 Or. at 198, 508 P.2d 190 (discussing that decision). The New York court had looked solely to "the ratio of expenses incurred to the total cost of the project" in deciding whether the expenses incurred were substantial, and it had reasoned that only those expenses that related "exclusive[ly]" to the proposed development should be considered in determining the ratio. See id.

This court took a different course. It reasoned that "the ratio test should be only one of the factors to be considered." Id. at 198, 508 P.2d 190. It explained that, in addition to the ratio of expenditures to projected cost,

Id. at 198-99, 508 P.2d 190.

Holmes departed from Town of Hempstead in three respects. First, it did not focus solely on the expenditure ratio. Second, in determining the expenses incurred, Town of Hempstead had considered only those expenditures made "for the exclusive purpose" of the proposed development. Holmes, 265 Or. at 198, 508 P.2d 190 (describing Town of Hempstead). Holmes instead identified the issue as whether "the expenditures have any relation to the completed project or could apply to various other uses of the land." Id. at 198-99, 508 P.2d 190. Third, the New York decision deducted the expenses that were not incurred exclusively for the proposed development from the numerator of the ratio, making the entire vesting decision turn on the resulting ratio. Holmes, by contrast, described the issue whether the expenditures "have any relation to the completed project or could apply to various other uses of the land" as one of five factors that "should be taken into consideration" in addition to the expenditure ratio. Id.

Holmes is the last word from this court on when a landowner will have a common law vested right to complete a partially finished use in the face of an adverse zoning change. It follows that, in determining whether a landowner has "a common law vested right" within the meaning of Measure 49, we look to Holmes both for guidance in determining what the common law of Oregon requires and also as the controlling precedent on that issue. We note, however, that almost 40 years have passed since this court decided Holmes and that, during that period, the amount of upfront costs that landowners must incur to build some projects has increased. See Salkin, 4 American Law of Zoning § 32:9 at 32-25 (explaining that, as a result of environmental and other regulations, landowners may have to invest substantial sums before beginning construction). We cannot lose sight of those changes in applying the factors identified in Holmes to current conditions.

The remaining issue is whether the Court of Appeals correctly held that, in the context of Measure 49, only some of the Holmes factors are material. On that issue, the Court of Appeals started from the premise that a landowner who complies with the terms of Measure 49 necessarily will have satisfied the second, third, and fourth Holmes factors.13 See Friends of Yamhill County, 237 Or.App. at 175-77, 238 P.3d 1016. More specifically, the Court of Appeals reasoned that compliance with the terms of Measure 49 necessarily means that a landowner's expenditures will relate to the proposed use and that all the landowners' expenditures will have been made in good faith. See id. It follows, the Court of Appeals reasoned, that those Holmes factors are redundant (and should be discounted) in a Measure 49 vested rights determination and that the remaining Holmes factors become "more material" in that determination. Id. at 177, 238 P.3d 1016 (emphasis in original).

One problem with the Court of Appeals' analysis is the premise on which it rests. In concluding that the fourth Holmes factor was redundant, the Court of Appeals began by noting that section 5(3) of Measure 49 requires that "the claimant's use of the property compl[y] with the waiver." Id. at 175, 238 P.3d 1016. The court reasoned that a landowner who satisfies that requirement will always satisfy the fourth Holmes factor. Id. We reach a different conclusion. The question under section 5(3) is whether the proposed use complies with the waiver. The question that the fourth Holmes factor asks is whether the expenditures the landowner has incurred relate to the proposed use and, if they do, whether those expenditures could be adapted to other permissible uses. Even if the proposed use complies with the waiver, it does not follow that all the expenditures either will relate to the use or could not be adapted to other uses. Contrary to the Court of Appeals' reasoning, the two inquiries are not coextensive; compliance with the terms of Measure 49 does not mean that the fourth Holmes factor will always be satisfied.

We reach the same conclusion regarding the landowner's good faith.14 Regarding that factor, the Court of Appeals observed that section 5(3) of Measure 49 asks whether a claimant had a common law vested right "on the effective date of this 2007 Act." See Friends of Yamhill County, 237 Or.App. at 176, 238 P.3d 1016. The court inferred from the fact that section 5(3) identifies the effective date of Measure 49 as the cut-off date for determining the existence of a vested right that the voters intended to give landowners a "green light" to incur as many costs as they could before that date and that all costs incurred before then necessarily would be incurred in good faith.

The text of section 5(3) permits that interpretation. However, the text also permits another interpretation. The phrase "on the effective date of this 2007 Act" could simply identify the cut-off date after which no further expenditures will be considered; under that interpretation, the phrase would not reflect a legislative judgment that any and all expenditures incurred before then will always count in determining the existence of a vested right. Not only is the latter interpretation permissible, but it is also gives full effect to the remainder of the section's text— "a common law vested right on the effective date of this 2007 Act." Or. Laws 2007, ch. 424, § 5(3). As noted, the phrase "a common law vested right" refers to a body of substantive common law that identifies which expenditures count in determining whether a landowner has a vested right to complete construction and which do not. Under the common law, expenditures made in good faith and expenditures that relate to the project count while expenditures made in bad faith and expenditures that could apply to other permissible uses of the land either do not count or are discounted in determining the existence of a vested right. See Holmes, 265 Or. at 198-99, 508 P.2d 190.

Interpreting the phrase "on the effective date of this 2007 Act" as merely identifying the cut-off date after which further expenditures will not be considered (rather than as a substantive judgment that all expenditures incurred before then will always count in a vested rights analysis) gives full effect to all the common law factors reflected in the phrase "a common law vested right." Conversely, the Court of Appeals' reading of the phrase "on the effective date of this 2007 Act" eliminates part of the content of the phrase "a common law vested right," contrary to the rule of statutory construction that we should give effect to all parts of a statute if possible. Cf. Vsetecka v. Safeway Stores, Inc., 337 Or. 502, 510, 98 P.3d 1116 (2004) (applying that rule). Although either interpretation is textually permissible, the better interpretation is the one that gives full effect to all the terms of section 5(3).

The context points in the same direction. Context includes "the preexisting common law and the statutory framework within which the law was enacted." Klamath Irrigation District v. United States, 348 Or. 15, 23, 227 P.3d 1145 (2010). The common law that preceded the adoption of Measure 49 made clear that "[t]he substantial expenditure requirement typically focuses only on those expenses incurred by the developer before the zoning ordinance [o]n which he relied was amended." Salkin, 4 American Law of Zoning § 32:4 at 32-18.15 That common law context suggests that the fact that section 5(3) states that the existence of a vested right will be measured as of the effective date of Measure 49 is not unusual. The effective date of a zoning change is ordinarily the date as of which vested rights are determined, and we hesitate to infer from the fact that section 5(3) follows that customary pattern that the voters intended to signal that every expenditure made before the effective date of Measure 49 will have been made in good faith.

Not only was it customary before the passage of Measure 49 to use the effective date of the zoning change as the temporal point for measuring the existence of a vested right, but courts either discounted or did not count expenditures made before the effective date of a zoning change when those expenditures were made in bad faith—i.e., when they were made for the purpose of circumventing the new zoning law. Id. § 32:5 at 32-20. Indeed, one of the sources that this court relied on in Holmes explained that the law would not help "`one who waits until after an ordinance has been enacted forbidding the proposed use and ... hastens to thwart the legislative act by making expenditures a few hours prior to the effective date of the ordinance.'" Elizabeth Lynne Pou, Comment, Good Faith Expenditures in Reliance on Building Permits as a Vested Right in North Carolina, 49 NC L Rev 197, 199-200 (1970) (quoting Warner v. W & O, Inc., 263 N.C. 37, 138 S.E.2d 782, 787 (1964)) (ellipses in law review article).

We conclude from the text and context of section 5(3) that stating that a common law vested right will be determined on the effective date of Measure 49 does not signal a legislative judgment that every expenditure made before then necessarily will have been made in good faith. Rather, we think that the measure leaves it to the trier of fact, as the common law did, to determine the good or bad faith of the landowner in making expenditures. Specifically, the trier of fact could find that expenditures made, after the voters adopted Measure 49, to "thwart the legislative act" were made in bad faith. Conversely, nothing precludes a trier of fact from finding that expenses planned before the voters approved Measure 49 but incurred after its passage but before its effective date were not made to thwart the measure. We need not attempt to catalogue the various ways in which a trier of fact could conclude that costs were incurred in either good or bad faith. It is sufficient to note that the Court of Appeals erred in concluding that any expenditures made before the effective date of Measure 49 necessarily were made in good faith.16

We accordingly conclude that the Court of Appeals erred in holding that compliance with the terms of section 5(3) means that a landowner's expenditures necessarily will relate to the proposed use and be made in good faith. We also conclude that the Court of Appeals erred in discounting some of the Holmes factors and finding, as a result, that other factors were "more material." Having reached those conclusions, we note that all the Holmes factors may not apply in a given case and that the extent to which they do apply will presumably vary with the circumstances of each case. We also note that, when a landowner seeks to establish a vested right because "substantial costs toward completion of the job * * * have been incurred," only one of the Holmes factors entails consideration of the "costs * * * incurred" —namely, "the ratio of expenses incurred to the total cost of the project." See Holmes, 265 Or. at 197, 508 P.2d 190 (listing that factor). That factor provides an objective measure of how far the landowner has proceeded towards completion of the construction. As such, we think it provides the necessary starting point in analyzing whether a landowner has incurred substantial costs toward completion of the job, although the other Holmes factors will bear on whether the costs incurred are substantial enough to establish a vested right under section 5(3).

Having concluded that the Court of Appeals erred in holding that only some of the Holmes factors will apply in a Measure 49 vested rights determination, we turn to the first issue that Cook raises—whether substantial evidence supports the county's determination that he had a vested right to complete his proposed 10-home subdivision. On that issue, Cook notes that this court stated in Holmes that "[t]he question of whether the landowner has proceeded far enough with the proposed construction to have acquired a vested right is an issue of fact to be decided on a case-by-case basis." See id. Cook reasons that, because there is substantial evidence to support the county's findings regarding the Holmes factors and because the county reasonably determined that those factors established that he had incurred substantial costs, the Court of Appeals should have affirmed the circuit court's judgment.17

Friends of Yamhill County, for its part, does not dispute that the county's determination that Cook had a vested right presents a factual issue. It argues, however, that a vested rights determination turns on the application of a multi-factor test and that, if the county's decision reveals that it misapplied the law in deciding that Cook's right to complete a 10-home subdivision had vested, then the decision must be reversed and remanded for reconsideration. As we understand Friends of Yamhill County's argument, it views the county's decision whether a vested right exists as similar to a jury's decision whether a person was negligent. It contends that both decisions present a question of fact for the trier of fact; however, if the trier of fact applies the wrong legal standard in deciding those issues, then the decision must be reversed and remanded.

In our view, Friends of Yamhill County has the better of the argument. Measure 49 authorizes a person adversely affected by a county vested rights decision to challenge that decision by means of a writ of review. See Or. Laws 2007, ch. 424, § 16, codified at ORS 195.318. A writ of review, in turn, permits a plaintiff to challenge a county's vested rights decision either because the county "[m]ade a finding or order not supported by substantial evidence in the whole record" or because it "[improperly construed the applicable law." ORS 34.040(1)(c) and (d).18 It follows that, if, as Friends of Yamhill County argues, the county improperly construed the law in determining that Cook had a vested right to complete his subdivision, then the Court of Appeals correctly reversed the circuit court's judgment and remanded for further proceedings before the county. Put differently, Cook cannot argue that substantial evidence supports the county's findings without first answering Friends of Yamhill County's argument (and the Court of Appeals' holding) that a legal error infected those findings. We accordingly turn to that issue.

Friends of Yamhill County argues that the county "improperly construed the applicable law" when it found it unnecessary to decide the ratio of the expenditures that Cook incurred to the cost of the project. On that issue, the county's vesting decision focused primarily on the proposition that the right that vested was the right to sell buildable lots. The county accordingly determined the ratio between the costs that Cook had incurred and the cost of developing buildable lots.19 Alternatively, the county noted that, if the relevant issue were the ratio between Cook's expenditures and the cost of actually building homes on the lots, then it was unclear "whether [it should] consider the cost of the types of homes the applicant is contemplating, the cost of a legal, habitable dwelling, an `average cost' or some other measure."

The county declined to resolve the latter issue, calling the expenditure ratio "speculative" and finding that, in any event, "[a]ll of the expenses in this case were legitimately incurred, in good faith, and are substantial." In the county's view, the fact that Cook's development was "fully sanctioned by, and coordinated with, Yamhill County is more important in this case than the speculative ratio test." The county then observed that, in any event:

We agree with the Court of Appeals that the county misapplied the governing law in failing to decide the ratio between the costs that Cook had incurred and the projected cost of constructing the residential subdivision. As noted, when a landowner seeks to establish a vested right by showing that he or she has incurred "substantial costs toward completion of the job," the expenditure ratio is the only Holmes factor that requires consideration of the costs incurred. It provides an objective measure of how far the landowner has proceeded towards completion of construction and thus serves as an initial gauge of whether the landowner has proceeded far enough that he or she has a vested right to complete construction.

To determine the ratio, the county should have found two historical facts: (1) the costs that Cook incurred to construct the planned development and (2) the estimated cost of the planned development. The county found the costs that Cook had incurred as of the effective date of Measure 49. The county erred, however, when it failed to find the estimated cost of building the homes. On that issue, Cook submitted evidence that he planned to put manufactured homes on the lots, while Friends of Yamhill County submitted evidence that Cook planned to build luxury homes on them. The county did not determine the type of homes that Cook planned to build, nor did it determine what the estimated cost of building those homes was.20 In that respect, the county incorrectly construed the applicable law, which required that the county find that fact.

Instead of making that finding, the county observed that:

That observation is problematic in two respects. First, the county did not find what the estimated cost of building the homes was and consequently did not determine how far Cook had proceeded towards completing construction of the homes. Rather, the county found only that there was "substantial evidence" to support a range of ratios. Substantial evidence, however, is the standard by which a court reviews a county's factual findings on a writ of review. See ORS 34.040(1)(c) (stating the standard of review for factual findings on a writ of review). It is not the standard by which the trier of fact makes a factual finding in the first place. The county's job as the trier of fact was to decide by a preponderance of the evidence what the estimated cost of constructing the planned homes was. The county did not do that.

Second, the county's statement that the amount spent and the steps taken were substantial, even assuming that each house would cost $450,000 to build, is not a sufficient substitute. This court faced a similar issue in Schoch v. Leupold & Stevens, 325 Or. 112, 934 P.2d 410 (1997). In that case, the workers' compensation board found that a specific amount was a reasonable attorney's fee and recited that it had particularly considered four of eight factors in reaching that conclusion. Id. at 119, 934 P.2d 410. In holding that the board's explanation was not sufficient to permit meaningful appellate review, this court reasoned that the board:

Id. This court held that, although the board legitimately could have reached the conclusion that it did, the board needed to identify and explain the factual premises of its decision more clearly. Id. at 120, 934 P.2d 410.

In this case, the county did not find what the cost of the project was, what the expenditure ratio was, nor did it explain why, in light of the other Holmes factors, an expenditure ratio based on a projected cost of $450,000 to construct each home would be a substantial expenditure. It instead posited, without explanation, that Cook's expenditures would be substantial, no matter what the facts were or what analysis applied. In this context, the county's assertion is not an adequate substitute for a finding regarding the estimated cost of Cook's development, a determination of the ratio between the costs that Cook had incurred and the projected cost of the development, and a reasoned explanation why those costs, in light of the other Holmes factors, were substantial enough to establish a vested right.

Having concluded that the county erred in not finding what the expenditure ratio was, we emphasize that the ratio provides only the starting point for the analysis. It is not the sole factor to be considered, nor will it necessarily be the dispositive factor; that is, there is not some specific percentage which must always be present before the right to complete construction will vest. For example, Holmes states that the "ultimate cost" also matters in the analysis.21 265 Or. at 199, 508 P.2d 190. As we understand Holmes, it lists the "ultimate cost" of the project separately from the expenditure ratio (which incorporates the cost of the project) because the weight to be given the expenditure may vary depending on the ultimate cost. More specifically, if the ultimate cost of developing a project is $1,000, a landowner who spends $200 toward its development will have incurred 20 percent of the projected cost. Few people, however, would think $200 a substantial expenditure, at least when determining whether a landowner's expenditure is substantial enough to complete and continue a prohibited use. Conversely, when the ultimate cost of a project runs into millions of dollars, an expenditure may be substantial even though it is only a small percentage of the projected cost.

We recognize, as Holmes did, that there is no bright line for determining when an expenditure will be substantial enough to establish a vested right. See id. at 197, 508 P.2d 190. However, we agree with the Court of Appeals that, in making that determination, the county needed to find the "ultimate cost" of completing construction and also the ratio between the costs that Cook had incurred and the cost of the project. Without those findings, the county was in no position to determine whether Cook's expenditures, in light of all the Holmes factors, were substantial.

We turn to the third issue that Cook raises. Cook argues that the Court of Appeals erred in concluding that the county needed to make additional findings on whether his proposed use complied with the terms of the Measure 37 waivers. Because this issue is likely to arise on remand, we address it briefly.

In seeking either compensation or a waiver under Measure 37, Cook stated that he intended to "divide the [property] into nine 2.69-acre parcels and one 12.4-acre parcel and develop a dwelling on each 2.69-acre parcel." In response, both the state and the county issued Measure 37 waivers. The state's waiver was more restrictive than the county's. It waived only those land use regulations that prevented the specific development that Cook sought to construct, and it did so "only to the extent [that] that use was permitted when [Cook] acquired the property on December 3, 1970." It follows, and neither party disputes, that the question whether Cook's "use complies with [his] waiver" turns on whether the county ordinances in effect when Cook acquired his property in 1970 permitted him to develop a residential subdivision on his land, which was zoned A-1 (farming use) at that time.22 See Or. Laws 2007, ch. 424, § 5(3) (to proceed under the authorization of a Measure 37 waiver, among other things, the landowner's use of the property must "compl[y] with the waiver").

In resolving that issue, the county reasoned that its prior approvals of Cook's preliminary and final subdivision plats constituted substantial evidence that a residential subdivision was a permissible use when Cook acquired the land in 1970. Alternatively, it reasoned that its approvals were final land use decisions that Friends of Yamhill County could not collaterally attack in this proceeding. The county accordingly ruled that Cook's proposed use complied with the Measure 37 waivers that both the county and the state had issued.

Before the circuit court, Friends of Yamhill County relied on a Yamhill County circuit court decision for the proposition that residential uses were not permitted in A-l zones, while Cook argued that a Land Use Board of Appeals (LUBA) decision established precisely the opposite proposition. Cook reasoned that, in any event, issue preclusion barred Friends of Yamhill County from relitigating LUBA's resolution of that issue. As noted, the circuit court ruled that substantial evidence supported the county's conclusion that residential subdivisions were permitted as a conditional use in an A-l zone in 1970, but it did not identify the evidence that supported the county's ruling.

The Court of Appeals, for its part, rejected the county's reasoning that the plat approvals were final decisions that Friends of Yamhill County could not collaterally challenge in this proceeding. Friends of Yamhill County, 237 Or.App. at 170, 238 P.3d 1016.23 The Court of Appeals then noted that "[t]he county decision did not make findings on the acquisition zoning and the limitations of that zoning" in 1970 and that, as a result, the county's decision "was insufficient to explain why the proposed vested use[] complied with the waivers." Id. at 171, 238 P.3d 1016. That deficiency, the Court of Appeals concluded, required a remand. Id.

On review, Cook does not appear to challenge the Court of Appeals' first ruling; that is, he does not argue that the plat approvals were final land use decisions that Friends of Yamhill County could not challenge in this proceeding. Rather, as we understand Cook's argument on review, he relies primarily on the LUBA decision both as support for his argument that residential subdivisions were a permissible use in A-l zones in 1970 and also to establish that issue preclusion bars Friends of Yamhill County from relitigating that issue in this proceeding.

Before turning to Cook's argument on review, we note that both the county and the circuit court treated this issue as if it were a question of fact. In doing so, they erred. The meaning of a county's ordinance, like the meaning of a statute, presents a question of law for the court. See Lincoln Loan Co. v. City of Portland, 317 Or. at 192, 199, 855 P.2d 151 (1993) (stating that principle); Anderson v. Peden, 284 Or. 313, 318, 587 P.2d 59 (1978) (same). Contrary to the basis for the county and the circuit court's conclusions that Cook's use complies with the waivers, the determination of what Yamhill County's ordinance permitted in 1970 is not a factual finding to be supported by substantial evidence in the record. The fact that the county approved Cook's preliminary and final subdivision plats, standing alone, provides no assurance that the county applied the correct legal standard in doing so—the county may have granted its approval erroneously. Cf. Loosli v. City of Salem, 345 Or. 303, 307, 193 P.3d 623 (2008) (local government official erroneously certified that an applicant's proposed use complied with local land use regulations). Accordingly, we conclude that the county erred in treating the question of Cook's compliance with his Measure 37 waiver as one that could be resolved by resort to factual evidence in the record, rather than by a legal interpretation of the applicable zoning ordinances.

Ordinarily, this court would be equally equipped to engage in that analysis. Cf. Anderson, 284 Or. at 315-18, 587 P.2d 59 (analyzing a disputed issue of ordinance interpretation). And the record in this case does contain some evidence from which a partial picture of the ordinance's terms might be reconstructed, including the uses permitted outright in an A-l zone, the uses permitted as conditional uses in that zone, and the omission of any minimum lot size requirement for that zone. The set of terms before this court is an incomplete one, however, and the parties have neither provided us with the complete text of the applicable ordinances, nor have they advanced any arguments bearing on their proper construction. As this case comes to us, we lack the ordinance's express wording, we lack any other historical evidence bearing on the meaning of the ordinance, and we lack any argument from the parties as to whether Cook's proposed development would, as a legal matter, have been permitted by that ordinance.

To be sure, LUBA's decision in Reeves v. Yamhill County, 55 Or. LUBA 452 (2007), provides some support for Cook's position.24 LUBA's decision, however, is not binding on this court; rather, the decision is relevant only for its persuasive value. However, without the complete text of the ordinance and any argument from the parties on how we should interpret it, we are in no position to assess how persuasive LUBA's reasoning is. More importantly, in Reeves, LUBA held only that former ORS 215.213 (1971) did not apply to an A-l zone in 1970 and thus did not preclude residential subdivisions in that zone. LUBA did not hold that the county's ordinance permitted residential subdivisions in an A-l zone. There is a difference between holding that a statute does not preclude a particular use and holding that the applicable zoning ordinances permit that use. Cook, as the party seeking to establish that his use complies with the waiver, must establish that the zoning ordinances in place when he acquired the property affirmatively permitted a residential subdivision in an A-l zone, not merely that former ORS 215.213 (1971) did not preclude such a use.

Put differently, even if we were to conclude that LUBA correctly decided the specific issue before it, the issue that LUBA decided is not a complete answer to the question that this case presents. That conclusion also answers Cook's claim that issue preclusion bars Friends of Yamhill County from relitigating whether former ORS 215.213 (1971) applied to A-l zones in 1970. Even if Friends of Yamhill County were precluded from relitigating that issue, it does not necessarily follow that residential subdivisions were a permitted use in A-l zones. We note that there may be reasons why issue preclusion does not apply in this context.25 Neither party, however, has addressed those issues, and it is sufficient to hold, as we do, that, even if issue preclusion applies, it does not establish that Cook's use complies with the waiver. For the foregoing reasons, we conclude that the Court of Appeals correctly held that the record was insufficient to determine whether Cook's proposed use complied with the zoning laws in place when he acquired his property and properly remanded this case for further proceedings on that issue.

The decision of the Court of Appeals is affirmed. The judgment of the circuit court is reversed, and the case is remanded to the circuit court for further proceedings consistent with this opinion.