Wolf v. Oregon Lottery Commission

Petitioners seek review of a rule adopted by the Oregon Lottery Commission (Lottery) establishing the compensation that the Lottery pays to retailers, mostly restaurants, taverns, and convenience stores, who make video lottery games available to the public. 1 Petitioners assert that the Lottery exceeded the legislature’s grant of authority when it promulgated the rule because the statute setting the Lottery’s task, ORS 461.445, required the Lottery to consider retailers’ “reasonable rate of return,” and nothing in the rule or in the rulemaking record indicates that the Lottery did so. We agree with petitioners. The rule is invalid.

Under ORS 183.400(4), we may invalidate an agency’s rule only if we find that, in adopting it, the agency violated the constitution, exceeded the agency’s statutory authority, or failed to comply with applicable rulemaking procedures. Petitioners do not contend that the Lottery violated rulemaking procedures. They do, however, contend that the rule exceeds the Lottery’s statutory authority and that it violates the provision of Article XV, section 4(3), of the Oregon Constitution that dictates how the Lottery must spend its proceeds. We begin and end with the subconstitutional question, Planned Parenthood Assn. v. Dept. of Human Res., 297 Or 562, 565, 687 P2d 785 (1984), because our disposition of that question obviates our need to address the constitutional challenge.

*673 I. THE LOTTERY’S RETAILER COMPENSATION RULE

The rule at issue, former OAR 177-040-0026 (May 27,2004), 2 declares that the Lottery’s payment to a “retailer,” that is, to a business that makes video lottery available to the public, equals a percentage of the proceeds of video lottery wagering (“game shares”) that takes place at the business less payment of prizes: “The compensation amount the Lottery shall pay a retailer for the sale of video lottery game shares is calculated on a percentage of net receipts during a business year.” The rule then announces two compensation *674 options, each setting rates based on net receipts. Under the first option, called the “3-Tier Option,” retailers who take in up to $175,000 in net receipts per year receive 32.5 percent of the proceeds; retailers with net receipts between $175,000 to $475,000 per year receive an additional 26 percent on proceeds exceeding $175,000; and retailers who sell $475,000 or more receive an additional 17 percent on net receipts exceeding $475,000. The “2-Tier Option,” likely preferred by retailers with a high volume of video lottery sales, pays the retailer 26 percent of net receipts up to $650,000 and 19 percent of net receipts exceeding $650,000.

II. INTERPRETATION OF ORS 461.445

To determine whether an agency’s rule is beyond the authority granted to the agency by statute, we must begin with “an examination of * * * the statutory provisions authorizing the rule[.]” ORS 183.400(3)(b). Here, the authorizing statute, ORS 461.445, provides:

“In establishing its schedule of payments to contractors, the Oregon State Lottery Commission shall undertake to develop a system that maximizes the net revenue to the state for the public purpose consistent with providing a reasonable rate of return for contractors.”

Our “examination” of the statute entails interpretation, and that task begins with the phrase “undertake to develop.” The Lottery argues that, by using that phrase, the legislature requires the Lottery only “to attempt to maximize state revenues to the extent it can do that while providing retailers a reasonable rate of return.” (Emphasis in original.) In the Lottery’s view, the “statute is hortatory.” In support of its position, the Lottery points out that, according to Webster’s Third New Int’l Dictionary 2491 (unabridged ed 2002), “undertake” is synonymous with “attempt.”

We are not persuaded. First, although Webster’s does point to the word “attempt” as a synonymous cross-reference for the word “undertake,” it points to a number of other synonyms, including, “guarantee,” “promise,” “contract,” and “covenant.” Further, Webster’s defines “undertake” to mean, variously, “take in hand,” “enter upon,” “set about,” “take upon oneself solemnly or expressly,” “put oneself under *675 obligation to perform,” and “accept as a charge.” Id. The dictionary, as usual, offers a smorgasbord of definitions. Some of them support the Lottery’s position and some undercut it. A more meaningful method of interpreting the term is to ask whether the legislature would enact operative legislation (as opposed to a general, introductory policy statement preceded by a “whereas” clause or its functional equivalent) containing explicit, substantive standards (such as “maximize [ ] the net revenue” and “reasonable rate of return”), couched in mandatory terms {“shall undertake”), if the legislature’s intention was merely to give advice. We conclude that, in context, the term “undertake” imposes a genuine mandate. The Lottery is charged with the duty to develop a system, and to do so subject to an express constraint: to maximize state revenue consistent with providing retailers a reasonable rate of return.

The parties dispute the precise nature of that constraint. The Lottery explained its understanding during the rulemaking process and does so again on judicial review: The constraint requires the Lottery to balance competing interests, giving due regard to each. Petitioners strenuously disagree:

“In order to meet the statutory mandate, the ultimate question the Lottery must answer is this: Do the commission rates maximize net revenue to the state? To answer that question, the Lottery needs to consider what a reasonable rate of return to retailers is, but only in the context of determining whether that rate will maximize net revenue to the state. * * * That is, the Legislature did not direct the Lottery to ‘maximize net revenue’ and ‘provide a reasonable rate of return to retailers.’ Rather, it directed the Lottery to set rates ‘consistent with’ providing a reasonable rate of return. The distinction is crucial. The two phrases are not on equal footing. The statute does not require or authorize the Lottery to attempt to find a ‘balance’ between maximizing revenue and providing a reasonable rate of return to retailers. To the contrary, the legislature simply recognized that in order to maximize revenue, retailers must receive a reasonable rate of return. Otherwise, retailers would not continue to offer video lottery and net revenues would decrease. The task for the agency is to identify the point *676 where a lower rate of return would decrease revenues to the state.”

(Emphasis in original.) Petitioners’ argument has some force. It presumes that the market is the ultimate arbiter of reasonableness. From that premise, it follows that, when the rate of return falls to a certain level, rational retailers will choose in such numbers to stop providing their services to the Lottery that any increase in revenue resulting from lower retailer commissions will be more than offset by the decrease in volume.

There are, however, two problems with petitioners’ argument. First, it begs the issue: it presumes that the legislature intended “reasonable” compensation to be the lowest compensation that would keep retailers in business — which is precisely the proposition that petitioners are trying to prove. Second, it renders superfluous the phrase “consistent with providing a reasonable rate of return for contractors.” Under petitioners’ interpretation, a statute simply directing the Lottery to establish a compensation system that maximizes revenues to the state would have effected the same outcome as the statute that the legislature enacted. As the Lottery points out, if providing reasonable rates of return to contractors “were already captured in a directive to maximize state revenue, * * * the legislature would have had no reason to describe it separately in the statute.” Instead, the legislature chose to give explicit and separate weight to retailer compensation. We therefore conclude that ORS 461.445 requires the Lottery to establish a retailer compensation system that balances two competing objectives: maximizing state revenue and providing retailers with a reasonable rate of return. In doing so, the Lottery is not obligated to establish a rate of return that is the lowest one that the market will sustain.

We turn, then, to the meaning of the phrase “reasonable rate of return.” We note first that the term “reasonable” is usually a delegative term signaling the legislature’s intention to give the Lottery authority to “complete a value judgment that the legislature itself has only indicated.” Springfield Education Assn. v. School Dist., 290 Or 217, 228, 621 P2d 547 (1980) (quoting McPherson v. Employment *677 Division, 285 Or 541, 550, 591 P2d 1381 (1979) (describing delegative terms)); Oregon Assn. of Rehab. Prof. v. Dept. of Ins., 99 Or App 613, 616, 783 P2d 1014 (1989) (illustrating significance of legislature’s use of “reasonable” in an authorizing statute). Our review of agency action implementing such a term is deferential; it is limited to a determination whether that action is “within the broadly stated legislative policy.” Springfield Education Assn., 290 Or at 230. In the present case, however, the question whether the Lottery’s compensation schedule accommodates a reasonable rate of return consistent with the legislative policy embodied in ORS 461.445 arises only if we determine that the Lottery’s system is, in fact, geared to any rate of return at all. If, as petitioners contend, the Lottery established a compensation schedule that is unrelated to rate of return, then obviously there can be no inquiry into whether the rate is reasonable.

Our initial task, then, is to determine whether the Lottery complied with the legislative mandate to calculate and consider a “rate of return.” In the context of a statutory scheme governing distribution of revenues derived from a commercial operation, the phrase is a term of art. We therefore assume that the legislature intended the phrase to have the specialized meaning that it conveys within that context. Leupold & Stevens, Inc. v. City of Beaverton, 206 Or App 368, 376, 138 P3d 23, rev den, 341 Or 579 (2006). A leading basic economics text provides the following definition and example:

“The rate of return on capital is the net return (or net rental per dollar of investment) received per year. It is a pure number — percent per year or per unit of time.

“For example, say that you own a truck worth $20,000. After all expenses are paid (maintenance, insurance, etc.), you earn a net rental of $2000 for a given year. We then say that the rate of return on capital for that year was $2000/ $20,000 = 10 percent per annum.”

Paul A. Samuelson and William D. Nordhaus, Economics 650 (12th ed 1985). In the context of regulation, “The rate of return is defined as the ratio of net profits to total invested capital.” Richard G. Lipsey, et al., Economics 309 (9th ed 1990). According to Black’s Law Dictionary 1289 (8th ed *678 2004), “rate of return” is “[t]he annual income from an investment, expressed as a percentage of the investment.” Seen as a fraction, rate of return has a numerator consisting of net profits and a denominator consisting of total capital investment. It appears, then, that the phrase “rate of return” in the video lottery context means the profits received by a retailer from the proceeds of video lottery games — that is, the revenue minus the operating cost of providing those games (for example, labor) — expressed as a percentage of the total capital investment (for example, furniture or remodeling, or the opportunity cost of using space for lottery games instead of other revenue-generating uses). It follows that ORS 461.445 requires that the Lottery, in developing a compensation system for retailers, obtain and consider data not only on retailers’ revenue, but also on their operating costs and capital investments.

Statutory context adds weight to that conclusion. A provision of the bill that authorized the establishment of video lottery, House Bill (HB) 3151 (1991), indicates that the legislature understood that, in making video lottery available to the public, retailers would incur expenses, including the costs for labor and maintenance, as well as fiirnishings and opportunity costs. Section 13 of that bill states:

“(1) The Oregon State Lottery Commission shall contract for an outside review of the cost of operation of video lottery games and the division of video lottery proceeds between total compensation paid to contractors and moneys available to the state for allocation. This report shall examine the costs and rates of return on video lottery machines and of the provision of space and services for the operation of such machines. This analysis shall be completed and submitted to the lottery commission not later than December 1, 1992.* * *

“(2) After considering the findings and recommendations of this report, the lottery commission shall modify its provisions for the payments to contractors.”

Or Laws 1991, ch 962, § 13, compiled as a note after ORS 461.544 (1991). 3 That provision required the Lottery to measure retailers’ compensation for providing video lottery games *679 in terms of revenue, operating costs, and capital expenditures (“space * * * for the operation” of machines).

The legislative record further reinforces the conclusion that the Lottery’s statutory mandate requires consideration of retailers’ profit, operating costs, and capital expenditures. As noted above, the legislature in 1991 enacted HB 3151, which directed the Lottery to offer video lottery games. In the course of their deliberations, members of the Senate Trade and Economic Development Committee sought an amendment to declare how retailers would be compensated. Initially, the drafters envisioned a system in which retailers would purchase or lease video lottery machines, so that a retailer’s “rate of return” would be keyed to an investment that included buying video lottery machines. See Tape Recording, Senate Trade and Economic Development Committee, HB 3151, May 31, 1991, Tape 135, Side A (exchange between witness and Sen Wayne Fawbush discussing “rate of return” for retailers that privately own video lottery machines and feasibility of state ownership). In the context of discussing how compensation should be handled, the committee chair, Senator Fawbush, suggested that the Lottery Commission should conduct an audit to determine a “reasonable rate of return” for retailers. Id. (statement of Sen Fawbush). When asked whether the state would guarantee that rate of return to retailers who invested in the business of selling video lottery shares, Senator Fawbush stated that the return would be linked to the retailers’ costs, and he discussed the possibility of allowing the state to own the machines instead. He stated, “What we would probably request of the Commission if we go ahead with this is that they do an independent audit and evaluation outside of their organization * * * to figure out the actual costs and what a fair rate of return would be and pay that. That’s what everybody wants.” Id. (emphasis added).

At the next hearing on the bill, the committee adopted an amendment that would later be codified as ORS 461.445 and another provision directing the Lottery to conduct the audit that Senator Fawbush had suggested. Tape *680 Recording, Senate Trade and Economic Development Committee, HB 3151, June 10,1991, Tape 145, Side A. The audit provision, which required the Lottery to conduct an external study of the costs retailers incurred in making video lottery available to the public, including the costs of space and labor, became section 13 of HB 3151, quoted above.

Shortly thereafter, the Senate Judiciary Committee considered the amended bill. There, Jim Davey, then-director of the Lottery, was asked about the costs of implementing the video lottery games. He testified that he anticipated that retailers would purchase 10,000 machines statewide at a cost of $5,000 each, and he predicted that the investment would be made “by independent businessmen in Oregon who would supply the games and want to share in the revenue.” Tape Recording, Senate Judiciary Committee, HB 3151, June 14, 1991, Tape 234, Side B (statement of Jim Davey). He explained that the results of the audit would assist in determining the amount of proceeds from video lottery share sales that the state would retain. Id. At that point, no decision had been made concerning whether the state would own the machines, but the committees were aware that state ownership would reduce costs for retailers.

At a subsequent hearing, concerns about fraud led to further discussion about directing the state to own all video lottery machines. Tape Recording, Senate Judiciary Committee, HB 3151, June 18, 1991, Tape 237, Side A (remarks of Sen Jeannette Hamby). Senator Hamby proposed amendments to the Senate Judiciary Committee that would require the state to own all video lottery machines. Testimony, Senate Judiciary Committee, HB 3151, June 18,1991, Ex H (Sen Jeannette Hamby proposed amendments); Tape Recording, Senate Judiciary Committee, HB 3151, June 18, 1991, Tape 238, Side A (remarks of Sen Jeannette Hamby). Senator Hamby asked Jim Davey to “identify the difference in anticipated revenues between state-purchased [and] owned [and] operated machines versus the proposal that is before us today.” Davey explained:

“[T]he language that is in the bill to maximize revenue to the state, and we are also directed to conduct an independent audit * * *, come back to the Commission, and recommend how that system should be applied to maximize revenues. One of the things this audit would look at is would it *681 be more profitable for the state to eliminate the middleman, so to speak. The way we estimated revenue, we would contract with machine operators * * * who would invest in the machines, and tavern owners, and the revenue currently is anticipated to be split [between] operators and tavern owners. * * * An alternative to that, which is not currently being done in the U.S., would be for the state to operate the system in its entirety.”

Tape Recording, Senate Judiciary Committee, HB 3151, June 18, 1991, Tape 237, Side A. Ultimately, the committee did not adopt Senator Hamby’s amendment requiring state ownership of lottery machines, leaving that question to the Lottery. Tape Recording, Senate Judiciary Committee, HB 3151, June 18, 1991, Tape 238, Side B. The bill was subsequently enacted without further amendment to the policy statement concerning compensation to retailers (which would become ORS 461.445) or to the audit provision.

From the entire legislative record, we infer that the drafters knew that contractors would incur outlay and expenses in offering video lottery games to the public, whether the state owned the video lottery machines or not. The committees were aware of the possibility that the state would own the machines and that a decision establishing such a system would reduce the costs to retailers. Even so, they maintained the mandate to devise a compensation scheme setting a reasonable rate of return for retailers, calculated in light of an audit that examined their costs. The legislative history therefore confirms that, in using the term “rate of return,” the legislature intended to assign to the Lottery the task of formulating a compensation system for retailers taking into account their revenues, operating costs, and total capital investment, in such a way as to provide a “reasonable rate of return.” Our remaining inquiry is whether the Lottery did so.

III. THE LOTTERY AND ITS STATUTORY MANDATE

Before considering whether the Lottery exceeded its delegated authority when it promulgated OAR 177-040-0026, we must address a preliminary matter: the role that the rulemaking record can play in our inquiry. The Lottery appears to argue that the record is beyond our purview. In *682 support of that position, the Lottery cites ORS 183.400(3), which provides:

“Judicial review of a rule shall be limited to an examination of:

“(a) The rule under review;

“(b) The statutory provisions authorizing the rule; and

“(c) Copies of all documents necessary to demonstrate compliance with applicable rulemaking procedures.”

The Lottery also cites AFSCME Local 2623 v. Dept. of Corrections, 315 Or 74, 79, 843 P2d 409 (1992), where the Supreme Court stated:

“Aside from questions that might arise concerning the facts surrounding the process of adopting a rule — questions not raised in this case — -judicial review under ORS 183.400 is limited to the face of the rule and the law pertinent to it. Numerous individual fact situations can arise under any rule, but judicial review of the rule as applied to each of those situations is reserved to other forums.”

The Lottery reads the statute and the case to stand for the proposition that, in a facial challenge to a rule such as the one in this case, the record is relevant only to the extent that it establishes (or fails to establish) that the rule “[w]as adopted without compliance with applicable rulemaking procedures.” ORS 183.400(4)(c). Except as necessary to determine compliance with rulemaking procedures, the Lottery contends, we cannot consider the rulemaking record. We disagree. Neither the statutes nor the cases construing them support the Lottery’s position.

As we have noted above, ORS 183.400(4) requires the reviewing court to invalidate a rule not only if it was not adopted in compliance with rulemaking procedures, but also when it violates constitutional provisions, ORS 183.400(4)(a), and, as is relevant to the present discussion, when it “[e]xceeds the statutory authority of the agency.” ORS 183.400(4)(b). In deciding whether a rule is valid, the court may conduct “an examination of’ the rule under review and the statutory provisions authorizing it. ORS 183.400(3)(a), (b). To examine a statute or a rule is to interpret it, and interpretation may require deployment of the usual arsenal of *683 interpretive strategies, including examination of a rule or statute’s sources, history, or circumstances surrounding its adoption.

AFSCME Local 2623 is not to the contrary. That case did not involve a challenge to the propriety of reviewing a rulemaking record; it involved a facial challenge to a rule based on specific fact situations in which the rule might be applied. Indeed, in declaring that judicial review is “limited to the face of the rule and the law pertinent to it,” the court explicitly excepted “the facts surrounding the process of adopting a rule.” AFSCME Local 2623, 315 Or at 79. As the context of the court’s discussion makes clear, the limitation that the court imposed precludes examination of a rule’s application in particular situations; it does not address examination of the rule’s enactment:

“We emphasize at the outset the limited scope of the Court of Appeals’ review (and ours) under ORS 183.400. Aside from questions that might arise concerning the facts surrounding the process of adopting a rule — questions not raised in this case — -judicial review under ORS 183.400 is limited to the face of the rule and the law pertinent to it. Numerous individual fact situations can arise under any rule, but judicial review of the rule as applied to each of those situations is reserved to other forums. ORS 183.400(1). See, e.g., ORS 183.482, ORS 183.484 (providing for judicial review of agency orders in various fact-specific situations). Petitioners’ petition for review in this case refers to actions alleged to be occurring pursuant to the rules at issue here, but the legality of any particular application of the rules is premature, and not subject to review under ORS 183.400.”

Id.; see also Oregon Newspaper Publishers v. Dept. of Corrections, 329 Or 115, 118-19, 988 P2d 359 (1999) (citing AFSCME Local 2623 for the proposition that “as applied” challenges cannot be brought under ORS 183.400); GTE Northwest, Inc. v. Public Utility Commission, 321 Or 458, 464, 900 P2d 495 (1995), cert den, 517 US 1155 (1996) (noting that, in AFSCME Local 2623, the court held only that “as applied” challenges could not be reviewed under ORS 183.400). Thus, in declaring that judicial review was limited to “the face of the rule,” the court meant the face of the rule as *684 opposed to an application of the rule; it did not mean the face of the rule as opposed to the rulemaking record.

There is, however, a line of cases dealing with the role that a rulemaking record can play in a facial challenge under ORS 183.400. That line begins with Int’l Cncl Shopping Cntrs v. Env. Quality Comm., 41 Or App 161, 597 P2d 847, rev den, 287 Or 508 (1979). In that case, the petitioners argued that one of the respondent’s rules was invalid because the respondent did not provide a statement of the factual basis on which the rules were based and “a statement of the reasoning which led from the factual basis to the rule.” Id. at 163. The court held that failure to provide a record containing such statements was not justification for invalidating the rules because, under the predecessor to ORS 183.335(13), rules “need not be based upon or supported by an evidentiary record” unless some statute so requires.

Shortly thereafter, in Kids Against the Cut v. Wage and Hour Comm., 41 Or App 179, 597 P2d 1264 (1979), the court distinguished an agency’s (nonexistent) need to rationally justify a rule on the basis of facts adduced “on the record,” from an agency’s very real need to produce, and the court’s ability to review, a record demonstrating that the agency, in promulgating the rule in question, had complied with substantive requirements imposed by the statute authorizing the rule. The petitioners in Kids Against the Cut challenged a rule reducing the minimum wage for persons under the age of 18. Id. at 181. The statute authorizing the commission to promulgate rules allowing subminimum wages stated that the commission could do so “ ‘when the commission has determined that the application of [the minimum wage statute] would substantially curtail opportunities for employment for specific types of persons! ” Id. (quoting ORS 653.030; emphasis in Kids Against the Cut). The court held that, when a statute imposes “limitations on state agencies” such as the one requiring the commission to link its rule establishing subminimum wages to the results of a factual determination, the reviewing court could legitimately consider the rulemaking record because the limitation would be “meaningless unless the rulemaking record reveals in some form” that the agency complied with it. Id. at 181-82. Although, presumably, the court could not substitute its *685 judgment for the agency’s as to what “curtail [ment] of opportunities” was “substantial,” it nonetheless could (and did) review the record to determine whether the agency had conducted the specified inquiry and considered its results. Accord Fund for Animals v. Dept. of Fish & Wildlife, 94 Or App 211, 214, 765 P2d 215 (1988), rev dismissed, 307 Or 611 (1989) (court must invalidate a rule when examination of rulemaking record shows that agency did not comply with substantive prerequisites; because respondent performed statutorily required investigation, rule was not invalid).

Further, the legislative mandate to engage in fact-finding and consider the results need not be explicitly stated as such, nor does the fact that the mandated factfinding is impracticable or even impossible excuse the failure to engage in it. WaterWatch v. Water Resources Commission, 199 Or App 598, 112 P3d 443 (2005), illustrates these principles. In that case, the petitioners argued that the Water Resources Commission exceeded its statutory authority under ORS 390.835(1), which states Oregon policy governing the use of scenic waterways. That subsection provides, in part:

“It is declared that the highest and best uses of the waters within scenic waterways are recreation, fish and wildlife uses. The free-flowing character of these waters shall be maintained in quantities necessary for recreation, fish and wildlife uses.”

The statute also required the director of the Water Resources Commission to deny applications for groundwater use when that use “will measurably reduce the surface water flows necessary” to implement the policy set out in ORS 390.835(1), unless it imposed, where appropriate, mitigation measures to “ensure the maintenance of the free-flowing character of the scenic waterway in quantities necessary for recreation, fish and wildlife.” ORS 390.835(9)(a), (10).

At issue was the validity of rules that required the “moderation” of impacts on surface water flows due to groundwater appropriations. Those rules, this court concluded, did not accomplish the statutory directive to “mitigate” the damage to surface flows due to groundwater appropriations because they did not ensure that the “character of the[ ] waters [would] be maintained in quantities necessary *686 for recreation, fish and wildlife uses,” as required in ORS 390.835(1). We based that holding in part on the conclusion that, in order to execute the mandate of ORS 390.835(1), the commission was implicitly required to “determine the level of stream flow necessary to support recreation, fish, and wildlife uses, [and] once that level is determined, [the level necessary to ensure that] it is to be maintained.” WaterWatch, 199 Or App at 609 (emphasis in original). Unless it determined the necessary level of surface water, the agency could not carry out its directive to maintain that level. We looked to the administrative record in order to determine whether the rules accomplished the statutory directive. We concluded that the rules were invalid because, even if the necessary level of stream flow was known to the agency, the record revealed that the agency could not adequately measure the impacts of groundwater appropriations in order to maintain that level. Id. at 615; see also Fed. of Seafood Hrvstrs. v. Fish & Wildlife Comm., 291 Or 452, 460, 632 P2d 777 (1981) (“If the findings can never be made, it is the legislature’s prerogative to rewrite the statute.”).

The statutes and cases, then, establish that we cannot invalidate a rule because it lacks a rational basis in the rulemaking record, but we can — indeed, we must — invalidate a rule when the record demonstrates that the agency has not engaged in the predicate factfinding that is required, either expressly or impliedly, by the statute authorizing the agency to promulgate the rule, and has not considered the results of that factfinding. Put another way: If a statute requires an agency to base its rule on facts, we will examine the record to determine whether the agency has acquired the facts and taken them into consideration. If it has not, the agency has exceeded its statutory authority. Kids Against the Cut, 41 Or App at 184. We will not, however, substitute our judgment for the agency’s with respect to whether the facts actually support the rule.

In the present case, the Lottery’s legislative mandate is to “maximize [ ] the net revenue to the state * * * consistent with a reasonable rate of return,” ORS 461.445. We now turn to the question whether the Lottery followed the legislature’s instructions.

*687 IV. THE RULEMAKING RECORD

OAR 177-040-0026 took shape during the two-year period before June 2004, the date when six-year contracts between the Lottery and video lottery retailers expired. The opportunity to draft new contracts presented the occasion to review previous retailer compensation rates and revise them through rulemaking if appropriate. In anticipation of that occasion, the Lottery’s director conducted an analysis of retailer compensation. She announced her recommendation in a memorandum on March 31, 2004, after 15 months of public meetings with retailers around the state and a review of several studies commissioned from private firms.

Her memorandum, entitled “Final Retailer Contract Recommendation,” sets out the bases for the proposed compensation system. At the outset, she acknowledged, in a section captioned “The Legal Standard,” that, in order to comply with the mandate of ORS 461.445, “the Commission must find that the compensation system it ultimately adopts maximizes the net revenue to the state consistent with providing a reasonable rate of return for Lottery retailers.” Also at the outset, she reported her reliance on one of the several commissioned studies of retailer compensation. She wrote:

“Throughout this memorandum, I have relied on the commission study conducted in October 1997 by KPMG Peat Marwick LLP (KPMG) for a number of reasons. First, it is the most recent comprehensive study on this issue. Second, all Lottery retailers had an opportunity to provide input into the study. Third, and most importantly, I generally found its analysis of lottery compensation issues persuasive.”

Because it formed the basis of the director’s recommendation (and, through that medium, the commission’s rulemaking), the KPMG study’s content and methodology merit attention.

The authors of the study sent a survey to all 3,198 lottery retailers then in existence; 232, or 7.3 percent, responded. In addition, KPMG conducted follow-up interviews with 100 responding retailers by telephone and with another 50 during site visits. Taking into consideration all of the responding retailers, KPMG reported that “very few *688 retailers were able to provide valid cost analysis of their lottery activities.” The study explained:

“Most of the Video Lottery retailers did not provide valid cost information. Common issues with the cost information provided were the following:

“• Compensation for administrative time was based on what owners thought their time should be worth, as opposed to what they were actually spending.

“• Capital costs for fixed assets in their business were listed as operating expenses in a lump sum, rather than being allocated over the useful life of the assets.

“• Space costs were not based on actual expenses incurred, but rather what the retailer thought the space should be worth.”

Despite the dearth of reliable data, the study concluded that “many retailers have a margin of 50% to 75% on their Video Lottery revenues.” By “margin,” the author of the study presumably meant “profit margin,” that is, revenue minus costs expressed as a percentage of costs. The study did not elaborate on the phrase “many retailers.”

The study then presented policy recommendations, expressed only in terms of a “reasonable rate range,” without reference to a “reasonable rate of return.” Based at least partially on retailers’ expectations (rather than retailers’ costs), the study stated that “[t]he reasonable rate range for Video Lottery games should incorporate the basic commission rate of 35% that has been in place since the initial days of Video Lottery.” However, the authors also reported:

“There is rationale for decreasing the commission rates of video game retailers to a rate closer to 30% than 35%. The primary rationale is that most retailers have had five years to recover initial capital costs associated with becoming a video retailer. A second rationale is that retailers appear to have a margin of lottery revenue over costs of 50% to 75% and this range is considered on the high end for commercial retail operations.”

The study did not explain how the recommended “rate range” established a “reasonable rate of return,” or, for that matter, *689 any rate of return at all. Such a recommendation would necessarily have included figures for capital investment — the denominator of the “rate of return” formula — and the report does not contain any such calculations. Rather, the report seems to derive its recommended commission rate as follows: The original compensation to retailers was 35 percent of net receipts. That compensation generated a profit margin that, by the study’s best guess, ranged between 50 percent and 75 percent. Such a range was “on the high end for commercial retail operations,” so a slightly lower rate, “closer to 30%,” would generate a profit that was closer to the median for commercial retail operations.

As noted above, the director and her staff drafted a memorandum based primarily on the KPMG study and formally presented it to the Lottery Commission on March 31, 2004 (it had been informally circulated in advance of the meeting). The memorandum summarized a number of studies leading up to the KPMG report, and contained the following recommendation:

“I recommend that the Commission implement a new video lottery compensation system that: (1) immediately lowers the overall rate paid to video lottery retailers from an average of 32% to an average between 28% and 29%; (2) reduces the number of tiers; and (3) provides retailers the choice of two compensation options. I find that this system will maximize the net revenue to the state for the public purpose consistent with providing a reasonable rate of return for retailers.”

In an oral presentation accompanying the memorandum, the director did not address the relationship between the recommended compensation system and retailers’ costs, capital investment, or “rate of return.” She did, however, explain that the proposal was an attempt to accommodate the expectations of retailers, who had become accustomed to rates between 32 and 35 percent:

“I think when the Commission makes a decision on this, just as the staff and I have made a decision on this, there isn’t any way that you can get around the fact that we have a 12-year history that has set compensation rates at somewhere between 35 percent and 32 percent. Many retailers, who have been with us from the beginning, have had a *690 12-year history of being compensated at 35 percent. Overall rates have dropped to an average of 32 percent over the course of 12 years. * * * [I]t’s tough when you’re not writing on a clean slate to just simply ignore history and to say, ‘well, we’re not going to look at history — we’re just going to go ahead and come up with a brand new model,’ notwithstanding what that may do to retailers, notwithstanding what that may do to the revenue that this state has come to expect from the Lottery.”

At the conclusion of the director’s presentation, the commission took testimony from the public. Some of that testimony was not successfully transcribed, but what survives came entirely from retailers or their representatives, all of whom urged the Lottery to retain or increase its existing compensation levels. Two items of written testimony were presented, one from a private citizen (the former Secretary of State) and one from an economist who was formerly the Executive Officer of the Legislative Task Force on Lottery Oversight. Both writers proposed rates significantly lower than the ones proposed by the director.

Each of the four commissioners then commented. All expressed appreciation for the integrity and thoroughness with which the Lottery director and staff had carried out its mandate; all but one, Commissioner Solomon, explained that retailer expectations figured significantly in their decision to support the director’s recommendation; and none of the commissioners except one, Solomon, mentioned retailer rate of return. A vote followed, and the director’s recommendation, with minor changes, passed, three to one, with Solomon casting the lone “no” vote.

V. CONCLUSION: THE RULE EXCEEDS THE LOTTERY’S AUTHORITY

Mindful that the effective operation of executive agencies depends to some significant degree on the limited scope of our review of their rulemaking, see ORS 183.400, we nonetheless conclude that, in this instance, the Lottery did not adequately comply with the legislative mandate described above: to maximize revenue to the state consistent *691 with a reasonable rate of return. Or Laws 1991, ch 962, § 13. 4 Clearly, the director and the commissioners correctly understood their statutory task, repeating it so often that one commissioner noted, “We can all say it in our sleep by now.” Equally clearly, however, the Lottery did not perform that task. Doing so would have required consideration of rate of return, which, in turn, required consideration of operating costs and capital investment. The KPMG study on which the director and the commission relied cautioned that, among the 7.6 percent of the retailers who responded to a survey, “very few retailers were able to provide valid cost analysis of their lottery activities” and “[m]ost of the Video Lottery retailers did not provide valid cost information.” More significantly, neither the KPMG study, nor any other source, provided information about capital investment. The record, then, affirmatively shows that, in promulgating OAR 177-040-0026, the Lottery never made the connection between the rates that the rule established and “reasonable rate of return.”

In a recent memorandum involving commission rates for video line games, the director reported:

“One complication in this issue concerns the studies that were done in the past. Most of those studies have been rejected by the Legislature as well as the Lottery Commission over the years. In examining these studies I have made several observations. There is a lack of reliable data on retailer costs. This is a recurring theme in all of the studies. All of the existing studies have focused exclusively on analyzing commissions, and the resulting revenue streams, for Video Lottery poker.”

(Underscoring in original.) In requiring the Lottery to create a compensation system that takes into account an objectively determined reasonable rate of return, the legislature may have imposed a difficult task, if not an impossible one. As the Supreme Court has noted, however, “ [i] f the findings can *692 never be made, it is the legislature’s prerogative to rewrite the statute.” Fed. of Seafood Hrvstrs., 291 Or at 460. In the meantime, because the rule that the Lottery enacted establishes a compensation that has no discernable relationship to rate of return, we find the rule to be outside of the Lottery’s authority and therefore invalid.

Former OAR 177-040-0026 (May 27, 2004) held invalid.