Lipscomb v. State Bd. of Higher Ed.

*474 LINDE, J. In 1921, the Governor’s veto power under the Oregon Constitution was amended to include, besides single items in appropriation bills, “any provision in new bills declaring an emergency, without thereby affecting any other provision of such bill.” The present dispute concerns whether the quoted words empower the Governor to veto only an emergency clause or any provision of a bill if the bill contains such a clause. In 1983, the Legislative Assembly passed Senate Bill 137, which made certain changes in the Public Employes’ Retirement System (PERS). One section amended an existing provision (ORS 237.003(8)) so as to exclude some types of earnings of college teachers from the definitions of “salary” or “other advantages” for purposes of determining employment retirement credits and employer contributions. 1983 SB 137, § 1, codified as ORS 237.003(1) through (12). The bill contained a provision declaring an “emergency.” The Governor signed Senate Bill 137, including the emergency clause, but his message to the Legislative Assembly stated that he disapproved of and therefore vetoed three provisions of the amending bill. On advice of Legislative Counsel, the legislative leadership at a subsequent special session treated the vetoes as legally ineffective and did not take them up for reconsideration and enactment over the veto. Legislative Counsel published the provisions in the session laws, Oregon Laws 1983, chapter 830, section 1, and in the Oregon Revised Statutes, ORS 237.003(8)(c)(E), (F) and (H) (1983 Replacement Part), but the executive branch treated the provisions as vetoed and continued to make contributions to PERS on the employe earnings that the bill sought to exclude. Plaintiffs, as “residents” and “taxpayers,” alleged the foregoing facts in an action against officials of the State Board and Department of Higher Education and of PERS, seeking orders declaring the purported vetoes invalid and the contributions to PERS improper and enjoining the defendant agencies to stop further contributions and to return those already made to the Department of Higher Education or to the State Treasurer. Plaintiffs also asked for attorney fees. Defendants filed an answer admitting most of the historical facts *475 pleaded but denying their alleged legal consequences. Defendants also denied allegations that plaintiffs were “aggrieved and adversely affected” by defendants’ acts or that they were entitled to attorney fees. Subsequently defendant William Whitelaw, a college teacher, was permitted to join defendants as a member of the class affected by the provisions at issue and filed the same answer as that of the other defendants. After both sides moved for summary judgment, the circuit court entered a partial summary judgment in favor of plaintiffs on all issues except for reserving for later determination what monetary relief should be awarded. The court also found that there was no just reason to delay an appeal. ORCP 67B. The Court of Appeals affirmed, holding that the 1921 amendment to Article V, section 15a, meant only to empower the Governor to veto the emergency clause of a bill. Lipscomb v. State Bd. of Higher Ed., 85 Or App 241, 736 P2d 571 (1987). Having allowed review because of the constitutional importance of the issue, we affirm the decision of the Court of Appeals. “Taxpayer” standing. Although defendants chose not to raise it, this court asked the parties to brief the question of plaintiffs’ standing as taxpayers to challenge defendants’ payments to PERS on the earnings excluded by the provisions of ORS 237.003(8)(c). In response, plaintiffs cite ORS 294.100, but, as defendants point out, that section refers only to suits by “any taxpayer of [a] district.” 1 Defendants, in turn, object to the court’s raising the question on its own motion on grounds that standing is not a jurisdictional matter. Defendants state that plaintiffs’ standing to initiate the action as taxpayers must be judged by the fiscal impact on them alleged on the face of the complaint. While maintaining that “it is indeed doubtful that plaintiffs will establish their standing as taxpayers” when the reserved issue of monetary relief is *476 reached, defendants did not try to demonstrate this by affidavits in support of their motion for summary judgment, apparently preferring not to question plaintiffs’ standing until after resolution of the constitutional issue. In Hanson v. Mosser, 247 Or 1, 427 P2d 97 (1967), cited by both sides, plaintiffs as low bidders on a state contract sought to enjoin a purchase from a higher bidder, and the court allowed them to sue as taxpayers “whose tax burden will be augmented by unlawful expenditure of public funds.” Id. at 11. Putting aside some of the earlier cases which involved local governments and property taxpayers, 2 the adequacy of the present complaint rests on plaintiffs’ allegation that by March 1985, the challenged contributions had cost Oregon taxpayers “in excess of the sum of $85,000 to date, exclusive of interest, * * * and said contributions are continuing.” Although the complaint did not allege the tax impact on these plaintiffs, and they are not entitled to rely on the cumulative cost to all “taxpayers of the State of Oregon” without a class action, defendants have accepted the allegation as adequate. We therefore turn to the merits. Principles of judicial review. The defendant officials make several arguments addressed to the court’s role in the present dispute. They propose a distinction between cases in which institutional issues of government and those in which individual rights are at stake. They contend that the constitution places in the Legislative Assembly the power to correct any abuse of the Governor’s veto power by overriding the veto, *477 and that the court should defer to the Governor’s understanding of his constitutional powers as long as that understanding is arguably correct. While the argument stops short of asserting that the dispute is nonjusticiable, it is reminiscent of the so-called “political question doctrine.” 3 Such a doctrine was mentioned in Putnam v. Norblad, 134 Or 433, 440-41, 293 P 940 (1930), which was an original proceeding for a writ of mandamus to order the Governor to call an election to fill a vacancy in the Senate. In denying the writ, the court distinguished between the Governor’s “ministerial” and “discretionary” duties, equating “political” with “discretionary” decisions. 4 Here *478 defendants argue that the court need not decide whether a governor’s veto was within his power because the Constitution commits to the Legislative Assembly the power to override the veto. But legislators do not vote on overriding a veto purely on constitutional grounds. A vote to override a veto of a specific measure involves the legislator’s view of its merits, not to mention partisan considerations; it is not a clean vote on the constitutional issue. Moreover, passage of a bill over a veto requires a two-thirds majority. Or Const Art V, § 15b. The legal effect of an arguably unconstitutional veto should not depend on whether a Governor can muster the support of one member more than one-third of those voting in one chamber to sustain the veto. 5 The distinction between issues of constitutional government and issues of individual rights does not withstand examination. Compliance with constitutional processes is the predicate for any government action; when an action is ultra vires, there is no need to decide whether a properly authorized action would violate a guarantee of individual rights. Cf. State ex rel Chapman v. Appling, 220 Or 41, 348 P2d 759 (1960) (statute concerning compensation of certain state officials held unconstitutional). When the institutional issue is raised by persons claiming that they are affected by the challenged action, individual rights are alleged to be at stake. The defendants presumably would expect careful constitutional scrutiny of a veto that is challenged by a defendant facing imprisonment, and we see no difference in principle when plaintiffs object to improperly increased taxes, as alleged in this case. We are not persuaded to distinguish between rights of a few persons under the Constitution and those shared by them with the community as a whole. Governors, legislators, and other public officials are *479 responsible in the first instance for determining their constitutional duties, as defendants say. See Cooper v. Eugene Sch. Dist. No. 4J, 301 Or 358, 364, 723 P2d 298 (1986). From this, defendants argue that the court should give “meaningful deference” to the Governor’s construction of his powers under Article V, section 15a. “Deference” is a word of many shadings, see 1000 Friends of Oregon v. LCDC (Lane Co.), 305 Or 384, 752 P2d 271 (1988); defendants presumably would find “meaningful” deference only when the Governor’s (or other political officer’s) construction prevails. In the constitutional relationships between the legislative and executive branches, a longstanding understanding and practice shared by both branches doubtless deserves respectful consideration, though it is not conclusive. Political institutions like any others may adapt to erroneous practices that should not be sustained. See Planned Parenthood Assn. v. Dept. of Human Res., 297 Or 562, 569-70, 687 P2d 785 (1984) (disapproving agency’s deference to Emergency Board on statutory issue). Congress and presidents, for instance, had struggled over and lived with the “legislative veto” (the practice of subjecting authorization of executive actions to future disapproval by one or both houses of Congress), for a generation before it was held unconstitutional in INS v. Chadha, 462 US 919, 103 S Ct 2764, 77 L Ed 2d 317 (1983). The veto at issue here does not represent an established practice based on a shared understanding of the Governor and the Legislative Assembly. To the contrary, it was the first time that a Governor had invoked Article V, section 15a, to disapprove a clause of a nonappropriation bill other than the emergency clause, and the legislature refused to recognize this disapproval, with the consequences already described. Those are not signs of a well-established shared understanding of the political constitution that a court would be cautious to upset. Origins of Article V, section 15a. When a constitutional amendment sets out to change the allocation of power between the political departments of government, it is necessary to understand the political background that motivated the amendment. Some amendments may be designed to correct or clarify a text that has proved difficult to administer or in the view of legislators and voters has been misinterpreted, or to replace obsolete provisions. The 1921 amendment of *480 Article V, section 15a, was not of that kind. It gave the Governor a measure of power over legislation that he did not previously have. Identifying the reasons for the amendment bears on interpreting what this new power was meant to be. The legal effect of adding a clause declaring an “emergency” to a bill is to allow legislation to take effect sooner than the 90 days after the end of the session provided in Article IV, section 28. 6 The political consequence is to prevent a referendum of the act upon petition of the voters, because that power is reserved only for “any Act, or part thereof, of the Legislative Assembly that does not become effective earlier than 90 days after the end of the session at which the Act is passed.” Or Const Art IV, § 1(3)(a). Contemporaneous sources show this background of the 1921 amendment. In 1921, the constitutional innovation of the popular initiative and referendum was still recent, and the use of declarations of emergencies that prevented referendum petitions was controversial. 7 The proposal, Senate Joint Resolution 13, that became the 1921 amendment to Article V, section 15a, was introduced by Senator B. L. Eddy on February 17, 1921, about a week before the end of the legislative session. In the absence of recorded debates, we must rely on press reports. On that day, the Senate had a heated debate about a pair of bills to use a fund collected from all state taxpayers for construction of a highway by a special district of coastal counties. The Oregon Statesman reported that several Senators criticized the declaration of an emergency attached to the bill, Senator Eddy objecting that the emergency clause “prohibits the people having anything to say as to whether [the highway bills] shall become effective.” Two Highway Bills Passed By Senators, The Oregon Statesman, Feb. 18,1921, at *481 1, col 3. The emergency clause preventing a statewide referendum on the coastal highway bill was not the only source of controversy about emergency clauses. Also on February 17, 1921, the Senate passed two other bills involving the highway commission that had first been passed at a special session in 1920 with declarations of an “emergency,” but which Governor Olcott had disapproved “not because he was opposed to them but because he did not consider them emergency measures,” as The Oregon Statesman explained in reporting the previous day’s House repassage of the bills. Id., Feb. 17,1921, at 1, col 6-7. 8 Considered descriptions of SJR 13 appeared during the spring. A committee of legislators prepared an official explanation in support of SJR 13 for the official Voters’ Pamphlet, describing SJR 13 as “a constitutional amendment to permit the governor to veto the emergency clause attached to a bill, without affecting the rest of the bill.” The argument stated: “The amendment now proposed would give him the power to veto the emergency clause of any bill, leaving the rest of the bill to stand. At present, if the emergency clause is objectionable, the governor can only veto the whole bill or allow it to become a law.” *482 Voters’ Pamphlet for June 7, 1921, Election, at 11. After explaining the existing provisions on emergency clauses and on the referendum, the committee continued: “When the legislature passes a bill and attaches to it a valid emergency clause, providing that the bill shall take effect upon its passage, there can be no referendum on the bill, that is to say, it cannot be referred to the people for their approval or rejection. It becomes a law at once. It is often necessary, in order to protect the public interests, that this should be the case. There is, however, a tendency to misuse the emergency clause, and to attach it when not necessary or proper. <<* * * * * “If the amendment now proposed shall be adopted the governor will have no reason to veto a good bill because an emergency clause has been unwisely added to it. He can veto that clause and allow the bill to stand on its merits, and subject to a referendum if the people so desire. In the absence of the referendum it would take effect at the expiration of ninety days from the adjournment of the legislative session. “The proposed amendment would enable the governor to bring his judgment to bear as to the necessity of the emergency clause, and it would prevent him from making the emergency clause an excuse or pretext for vetoing the whole bill. As matters stand now, there is opportunity for the governor to make the emergency clause the ostensible reason for a veto, which is really based on other grounds.” 9 Id. Editorial support quoted in the Oregon Voter, a weekly magazine devoted to state government and elections, *483 before the June 7 special election was based on this understanding. “ ‘Adoption of the amendment gives the governor the power to remove the appendix without affecting the body of the measure,’ analyses Pendleton East Oregonian. ‘The emergency clause veto is a needed supplement to the single item veto,’ says Astoria Budget.” From the McMinnville Telephone Register: “There will be on the June ballot a constitutional amendment, which will grant to the governor the power to veto the emergency clause of any bill passed by the legislature. This is a splendid measure and should have the support of the voters. This measure will correct an abuse which has been done to the purposes of the emergency clause. a* * * * * “[T]his clause was intended for use only for cases involving peace, health or safety to the people; but sometimes it has been seized upon and used to keep a measure from being referred to the people, even when no emergency exists. “This emergency veto will give the governor power to veto the emergency clause and not affect the body of the measure. As it is now, he must veto the entire bill or permit it to become a law. This emergency veto is a needed supplement to the single item veto and when it is passed another of law evils will be relegated to the scrap heap.” Oregon Voter, June 4, 1921, at 27. The Oregon Voter itself favored the measure because it would permit the governor to veto the emergency clause only, thereby preventing its excessive use, particularly its use “solely for the purpose of making it impossible to use the referendum.” Id. at 10. Reviewing the measures on the special election ballot, the Portland Oregonian wrote on June 5: *484 Portland Oregonian, June 5, 1921, at 1, col 6 and cont. at 11, col 1. *483 “The emergency clause veto constitutional amendment, while not having received any great amount of publicity, apparently is favored by the majority of voters throughout the state. This measure, if approved, will vest in the governor power to veto the emergency clause without affecting any other provisions of such bills.” 10 *484 “Plain meaning” of Article V, section 15a. The contemporaneous materials leave little doubt what the sponsors and the public understood SJR 13 to mean at the time of its enactment. 11 While plaintiffs rely on this history of Article V, section 15a, defendants urge us to disregard the historical evidence of its purpose and scope. They argue that resort to legislative history is improper when the meaning of the text is “plain” or “unambiguous.” This and similar formulations are often recited, 12 but in practice they do not and should not confine the court to *485 historically blind exegesis. A case dealing with the allocation of legislative and executive power by the state’s charter of government cannot be treated as if it involved the use of parol evidence to vary the words of a private contract. When one side to a dispute over the meaning of a public law urges a court not to look at or consider materials presented by the other side for its reading of the law, this only invites doubt whether the materials might show that the “plain meaning” is not so plain after all. That is the case here. 13 In practice, also, courts rarely see disputes over interpretation when the opposing party cannot show a possible alternative reading of the words, which it claims to be correct in context. That, too, is true in this case. Article V, section 15a, is not as unambiguous as defendants claim. It provides (with the 1921 amendment italicized): “The Governor shall have power to veto single items in appropriation bills, and any provision in new bills declaring an emergency, without thereby affecting any other provision of such bill.” Fixing upon the word “any” before “provision,” defendants insist that it must mean that in bills declaring an emergency, the governor shall have power to veto any provision. Plaintiffs, on the other hand, maintain that the words “declaring an emergency” were meant to modify “provision,” not “bills.” Defendants, in turn, respond that the amendment’s drafter could easily write “any provision declaring an emergency,” if this was meant, and that plaintiffs’ reading makes surplusage of the words “in new bills.” The arguments show, not that one reading of the text is more correct or even more plausible than the other, but that the constitutional issue cannot be decided simply by parsing the words of the amendment. That approach puts a greater burden on draftsmanship than the legislative institutions, then or now, can bear, let alone the initiative process that produces some of our laws. See OEA v. Phillips, 302 Or 87, 106-07, 727 P2d 602 (1986) (Linde, J., concurring). Nor are we dealing with the work of the kind of committee on style that *486 produced the text of the United States Constitution. The 1921 Legislative Assembly had no regular legislative counsel or committee staffs, and SJR 13 was sent to committee and reported back to each chamber so quickly as to suggest little if any hearing or scrutiny for potential ambiguity. 14 No doubt there were other and clearer ways to phrase what Senator. Eddy and his colleagues had in mind. We have no published explanation why they included the words “in new bills,” since any “bill” is “new” when it is introduced, but perhaps they meant to exclude bills already passed earlier in the same session, which was almost at an end. As long as the sponsors and supporters of SJR 13 knew what they intended to accomplish and why, they had little time and probably little patience for meticulous editorial refinements. Conclusion. Contemporaneous materials widely available to the voters in 1921, particularly the explanation by a committee of legislators in the official Voters’ Pamphlet, leave no doubt that the amendment to Article V, section 15a, was intended to authorize the Governor to veto a declaration of emergency in a bill so as to protect the opportunity of voters to petition for a referendum. Although the wording of the amendment also could suggest a much broader power to veto individual provisions of nonappropriation bills, it is not believable that such a broad power would have gone unmentioned and undebated if anyone had understood the amendment as defendants now contend. The Governor’s veto power is not the kind of provision that must find new application in changing technological, economic, or social conditions. We *487 agree with the Legislative Assembly and the courts below that the purported veto of the disputed provisions of SB 137 were without legal effect. The decisions of the circuit court and of the Court of Appeals are affirmed. Footnotes [1] ORS 294.100(2) provides: “(2) Any public official who expends any public money in excess of the amounts, or for any other or different purpose or purposes than authorized by law, shall be civilly liable for the return of the money by suit of the district attorney of the district where the offense is committed, or at the suit of any taxpayer of such district.” [2] E.g., Childs v. Marion County, 163 Or 411, 97 P2d 955 (1940); Miller v. Henry, 62 Or 4, 124 P 197 (1912); Sears v. James, 47 Or 50, 82 P 14 (1905); Burness v. Multnomah County, 37 Or 460, 60 P 1005 (1900); Avery v. Job, 25 Or 512, 36 P 293 (1894); White v. Commissioners, 13 Or 317, 10 P 484 (1886). Gruber v. Lincoln Hospital District, 285 Or 3, 8-9, 588 P2d 1281 (1979), distinguished between “taxpayer standing” for fiscal and for other alleged maladministration: “When a suit under ORS 28.020 is grounded on ‘taxpayer’ standing, these fiscal consequences presumably represent the effect on plaintiff required by that statute. “The present complaint contains no such allegations. * * * When the potential fiscal implications of a public contract are not apparent on its face, they should not be left to judicial speculation; they need to be asserted, if a plaintiff relies on his interest as a taxpayer rather than some other ‘rights, status, or other legal relations’ to proceed under ORS 28.020.” See also McKinney v. Watson, 74 Or 220, 145 P 266 (1915) (taxpayer’s allegations insufficient to show increased taxes from transfer of license receipts from corporation fund). [3] For background and debate on judicial abstention from deciding “political” questions, see, e.g., Redish, Judicial Review and the “Political Question, ” 79 NWU L Rev 1031 (1985) (doctrine should be abandoned); Henkin, Is There a "Political Question" Doctrine?, 85 Yale L J 597 (1976) (no such “doctrine” exists); L. Tribe, American Constitutional Law 96-107 (2d ed 1987) (doctrine applies only when disputed issue does not give rise to enforceable rights); A. Bickel, The Least Dangerous Branch 186 (2d ed 1986) (courts should affirm legitimacy of “discretionary functions of the political institutions, which are unprincipled on principle,” by not making rules for them); Wechsler, Toward Neutral Principles of Constitutional Law, 73 Harv L Rev 1, 6-10 (1959) (court should invoke doctrine only when constitution has committed decision to another branch); and J. Choper, Judicial Review and the National Political Process 263 (1980) (courts should not decide constitutional questions concerning the respective powers of the legislature and the executive vis-a-vis one another). [4] “It is generally agreed that the courts have no authority or power to interfere by mandamus with the Governor upon questions involving his judgment and discretion. He cannot be compelled by mandamus to perform duties which are partly discretionary and partly ministerial. As to the power of the courts to enforce the performance of ministerial duties by the Governor, there is a difference of opinion: 38 C.J. 662, § 207. In this state it has been held that where the duty to be performed is purely ministerial and not political and not otherwise objectional, [sic] the courts, in a proper case, may exercise their authority. The general rule is that the executive of the state, in the proper discharge of his duties under the constitution, is as independent of the courts as he is of the legislature. The authorities agree in holding that where the duty sought to be enforced is not merely ministerial, but one involving discretion, it cannot be enforced by the writ: 1 Cooley’s Const. Lim. (8th Ed.), pp. 221, 222.” Putnam v. Norblad, 134 Or 433, 441-42, 293 P 940 (1930). Other language in the opinion speaks of “political matters,” but the words are juxtaposed to “ministerial”; they do not imply that any duty related to what is commonly called “politics” cannot also be nondiscretionary and the subject of adjudication by a court. See, e.g., State ex rel Sajo v. Paulus, 297 Or 646, 688 P2d 367 (1984) (reviewing method of qualifying signatures on initiative petitions). The phrase “policy question” would be preferable to “political question” to describe decisions beyond judicial determination. Putnam speaks of the “proper” discharge of official duties. Of course, no official can invoke either “policy” or “politics” to avoid review of actions not authorized by law, though in some states courts will regard the signatures of responsible legislative *478 officers on an enrolled bill as conclusive on compliance with required procedures of its enactment. See Lloyd, Judicial Control of Legislative Procedure, 4 Syracuse L Rev 6 (1952); Treadgold, Constitutional Provisions Regulating the Mechanics of Statutory Enactment in Oregon—Effect of Enrollment, 27 Or L Rev 46 (1947); cf. Field v. Clark, 143 US 649, 12 S Ct 495, 36 L Ed 294 (1892); United States v. Ballin, 144 US 1, 12 S Ct 507, 36 L Ed 321 (1892). [5] See, e.g., Eckhardt and Black, The Tides of Power: Conversations on the American Constitution 23 (1976) (Congress could but will not override vetoes on nonpartisan, institutional principles). [6] Article IV, section 28, of the Oregon Constitution provides: “No act shall take effect, until ninety days from the end of the session at which the same shall have been passed, except in case of emergency; which emergency shall be declared in the preamble, or in the body of the law.” [7] A declaration of emergency still may not be attached to a tax measure. Or Const Art IX, § la. The legislature itself may provide both speed and a referendum by setting a special election on a measure. See, e.g., SJR 30 (1983 First Special Session) (calling special election on constitutional amendment regarding taxation), see Hart v. Paulus, 296 Or 352, 676 P2d 1384 (1984). A referred measure is not subject to veto by the Governor. Or Const Art IV, § l(3)(e). [8] The Oregon Statesman briefly reported that Senator Eddy “introduced a joint resolution calling for a constitutional amendment empowering the governor to veto single items in emergency bills without effecting [sic] the remainder of the measure. The governor now has power to veto single items in appropriation bills without effecting [sic] the other parts of the measure.” The Oregon Statesman, Feb. 18, 1921 at 1, col 2. The Portland Oregonian also erroneously reported Senator Eddy’s introduction of SJR 13: “BILL ENLARGES VETO POWER “Constitutional Amendment Asked in Senate Measure “STATE HOUSE, Salem, Or., Feb. 17. —(Special.)— A proposed constitutional amendment introduced by Senator Eddy, if approved by the voters at the next special or general election, will give the governor authority to veto single items in appropriation bills. “He also may veto any provision in new bills declaring an emergency without thereby affecting any other provision of such proposed law.” The Portland Oregonian, Feb. 18, 1921 at 5, col 2. The lead paragraph is clearly in error; that veto power had been enjoyed by the Governor since a 1916 constitutional amendment. The second paragraph correctly describes SJR 13 but implies that the governor already has that power. In any event, since the second paragraph is nearly a word-for-word quote of the phrase at issue, it sheds little light upon the question now before the court. [9] The defensive tone of further passages in the committee’s argument may explain why a legislative majority was ready to increase the Governor’s veto power: “The chief reason for the amendment is to prevent undue encroachment upon the referendum power of the people. Complaint has been made that the legislature seeks at times to prevent measures from being submitted to the people, and accomplishes this object by the use of the emergency clause. We do not think this often happens, and yet there is at times a disposition in the legislature to use the emergency clause where it is unnecessary. The fact that the governor could veto an emergency clause and allow the remainder of a bill to stand would not only protect the referendum power in fact, but would also tend to allay the suspicion that sometimes arises as to the motives of the legislature in using the emergency clause. The amendment is in the interest of good government and of public confidence and repose.” Voters’ Pamphlet, supra, at 11-12. The committee’s statement also was published in the Oregon Voter on April 2,1921 at 17-19. [10] On election day, the same newspaper’s lead story described the measure as providing “that the governor can veto an item to cover an emergency clause without affecting other items in the bill,” Portland Oregonian, June 7, 1921, at 1, col 8. [11] We have found only one item besides the brief news report previously quoted (n 8, supra,) that anyone interpreted SJR 13 as defendants now suggest. A speaker, Thomas B. Kay, reporting to the Salem Rotary Club on all the special election measures, was reported to have said that the proposal would “allow the governor to approve or reject certain parts of any legislative measure, without having either to swallow vicious provisions by approval or kill essential good things because other parts of the bill were bad,” citing the highway bill as an egregious example. Ballot Measures Heard at Lunch, The Oregon Statesman, June 7, 1921, at p 4, col 6-7. Besides the two views of Article V, section 15a, argued by the present parties, there is in fact a third possible view of what the measure sought to achieve: the Governor might have been empowered to veto a declaration of an emergency with respect to some provisions of a bill but not others, thus saving the possibility of a referendum on some provisions without sacrificing an early effective date for any provisions that he considered genuinely urgent. But no party has argued for this alternative, and the historical record points to a simple veto of the emergency clause on entire bills. [12] Defendants quote from Northwest Natural Gas Co. v. Frank, 293 Or 374, 381, 648 P2d 1284 (1982): “As a court, our role is to interpret the statutes and constitutional provisions. We do not redraft these provisions; we interpret them as the legislature has drafted them. It is axiomatic that in a case of statutory and constitutional construction, this court must give preeminent attention to the language which the legislature and the people have adopted. The statutes are the law, and while the legislative history may provide invaluable insights into the legislative process, it remains supplemental to the statutes as adopted.” Plaintiffs counter with State ex rel Cox v. Wilson, 277 Or 747, 750, 562 P2d 172 (1977): “ ‘Frequently, however, even when the plain meaning did not produce absurd results but merely an unreasonable one “plainly at variance with the policy of the legislation as a whole” this Court has followed that purpose, rather than the literal words. When aid to construction of the meaning of words, as used in the statute, is available, there certainly can be no “rule of law” which forbids its use, however clear the words may appear on “superficial examination.”* * *’ ” The court was quoting from United States v. Amer. Trucking Ass’ns, 310 US 534, 543-44, 60 S Ct 1059, 84 L Ed 1345 (1940), omitting footnotes. Neither brief refers to the other’s quotation. [13] Even a numerical reference, by nature unambiguous when viewed in isolation, may prove erroneous and require interpretation in its legislative context. See, e.g., Bush v. Greyhound Lines, Inc., 295 Or 619, 669 P2d 324 (1983); Warren v. Marion County, et al, 222 Or 307, 353 P2d 257 (1960). [14] The Oregon Voter’s March 5,1921 issue followed its report of the introduction of SJR 13 with this item (at page 4): “ERRORS IN WORDING “Many of the laws enacted by the 1921 legislature contain errors in wording, some of them materially affecting the meaning of the language. Some errors crept into constitutional amendments and bills to be submitted to the people. While some errors are made at every session, the carelessness this session in this regard was peculiarly atrocious.” The criticism was directed more at incompetent typists than at careless draftsmanship. It continued: “It is not impossible that litigation over the meaning of some of these misspelled, misplaced, substituted or omitted words will cost many times what a thoroughly competent typing and proofreading force would have cost if employed by the legislature.” Id.