Opinion
Civil No. 3486.
Filed February 24, 1934.
1. CONSTITUTIONAL LAW. — Constitutions should be liberally construed to give effect to their general meaning, rather than narrowly in accordance with technical rules of grammar.
2. COURTS. — Rule of " stare decisis" has for its object the salutary effect of uniformity, certainty and stability of the law, and is the application of doctrine of estoppel to court decisions, though it should not prevail where departure therefrom is necessary to avoid perpetuation of pernicious error.
3. COURTS. — Rule of stare decisis should not be disregarded by majority vote of court except in cases of gravest importance and where previous decision conclusively appears to be erroneous.
4. STATUTES. — Where last conference committee's emergency bill to authorize tax anticipation bonds did not receive vote of two-thirds of all members elected to house, but did receive vote of majority of members elected to house and senate, respectively, and was signed by Governor and no referendum was had, bill became operative as law ninety days after close of session, since emergency clause only was void (Laws 1933, chap. 16; Const., art. 4, pt. 1, § 1, subsec. 3).
5. MANDAMUS. — It was state treasurer's duty, enforceable by mandamus, to countersign tax anticipation bonds authorized by statute and resolution of loan commissioners pursuant thereto (Laws 1933, chap. 16).
See 7 Cal. Jur. 632; 7 R.C.L. 1008 (3 Perm. Supp., p. 2116).
Original proceeding in Mandamus. Peremptory writ ordered issued.
Mr. Arthur T. La Prade, Attorney General, and Mr. Chas. L. Strouss and Mr. Frank W. Beer, Assistant Attorneys General, for Plaintiff.
Messrs. Kibbey, Bennett, Gust, Smith Rosenfeld, for Respondent.
This is an original proceeding in this court by the state of Arizona, on the relation of its Attorney General, hereinafter called petitioner, asking for a writ of mandamus against William M. Cox, as treasurer of the state, hereinafter called respondent, to compel him to countersign certain tax anticipation bonds, duly authorized by a resolution of the loan commissioners of the state. The respondent answered, setting up certain facts, and petitioner demurred to the answer as being insufficient. The legal question before the court is whether petitioner's demurrer should be sustained.
There is no dispute as to the facts, and the real issue is as to whether or not chapter 16 of the Session Laws of 1933 is a valid enactment or not, for, if it is, there is no doubt it is the imperative duty of respondent to sign the bonds in question. The validity of the chapter is attacked on the ground that it was not constitutionally adopted. In order to determine this question, it is necessary that we recite briefly certain facts in regard to its passage.
The original act, on file in the office of the Secretary of State, is entitled:
"An Act to permit the issuance of bonds in anticipation of the collection of state taxes; to authorize such bond issue for the last half of the year 1932, all of the year 1933, and the first half of the year 1934; and declaring an emergency."
Then follow eight sections setting forth the circumstances under which, and the manner in which, tax anticipation bonds may be issued. Section 9 of the act is the ordinary repealing clause of previous laws, and section 10 thereof reads as follows:
"To preserve the public peace, health and safety, it is necessary that this act shall become immediately operative. It is therefore declared to be an emergency measure, and shall take effect upon its passage in the manner provided by law."
The journals of the house and senate show that the bill first passed the house by a vote of 53 ayes to 0 nays, was amended by the senate and passed, as amended, by a vote of 13 ayes to 6 nays, went to conference twice, and the report of the last conference committee was adopted by the senate by a vote of 19 ayes and 0 nays, and by the house by a vote of 42 ayes and 12 nays.
It is the contention of respondent (a) that the act contains an emergency clause, and that the journal of the house shows it did not receive two-thirds of the votes of all the members elected thereto, so that it could not become effective as an emergency measure; and (b) that, since on its face it carried the emergency clause, under no circumstances could it become effective as an ordinary measure ninety days after it was signed by the Governor. In support of this position respondent cites the recent case of Cox v. Stults Eagle Drug Co., 42 Ariz. 1, 21 P.2d 914, 917. The facts in the present case are identical with those arising in the case just cited, and the points of law involved are the same.
There is no question, if this court adheres to the rules of law laid down in that case, that chapter 16, supra, is ineffective for any purpose, for therein we held that a measure carrying an emergency clause must receive a vote of two-thirds of all the members elected to each house of the legislature, not merely on its original passage by each house, but upon any amendments which might be added by one house after it had passed the other. In other words, that the constitutional two-thirds majority was required for the bill in its final form as it went to the Governor. We also held explicitly that, even though the bill did receive the votes of a majority of the members elected to each house, since it carried the emergency clause on its face, under no circumstances, and in no manner, could it become effective as an ordinary measure.
The petitioner does not question the ruling of the court in Cox v. Stults, supra, so far as it determines that chapter 16, supra, did not take effect as an emergency measure, and that the journals of the legislature are admissible as evidence, and, indeed, conclusive as to what the actions of the house and senate on the chapter actually were, but it contends that the holding that a bill which took the course in the legislature which was taken by chapter 16 does not take effect as an ordinary measure ninety days after the adjournment of the legislature is unsupported by authority and unsound in reason.
The reasoning of the majority opinion on this point is based upon two points. The first is the exact language of the Constitution in subsection 3, section 1, part 1, article 4, which provides that ". . . no such emergency measure shall be considered passed by the Legislature unless it shall . . . be approved by the affirmative votes of two-thirds of the members elected to each House of the Legislature, taken by roll call of ayes and nays. . . ."
The court, after quoting this language, proceeds to say:
". . . If this provision read that `no such measure shall be considered passed as an emergency,' the contention of appellants, that it would become operative ninety days after the close of the session, might have merit, but inasmuch as it uses the all-inclusive term, `no such emergency measure,' when referring to those not considered passed without the two-thirds vote, there is nothing more to say. . . ."
It is the general rule that, because Constitutions are for the purpose of laying down broad general principles, and not the expression of minute details of law, their terms are to be construed liberally, for the purpose of giving effect to the general meaning and spirit of the instrument, rather than as limited by technical rules of grammar. Cohens v. Virginia, 6 Wheat. 262, 5 L.Ed. 257; Gherna v. State, 16 Ariz. 344, 146 P. 494, Ann. Cas. 1916D 94. This court, in Clark v. Boyce, 20 Ariz. 544, 185 P. 136, 140, says:
"The prime effort in construing Constitutions or statutes is to ascertain the intention of those who framed them, and `to give effect to the intention therein expressed when such intention can be collected therefrom, words may be modified, altered or supplied so as to obviate any repugnancy to or inconsistency with such intention, although in so doing particular provisions may not be read or construed according to their literal meaning.' . . ."
On a careful reading and comparison of all the provisions of the Constitution dealing with the referendum and the enactment of legislation, their purpose is obvious. It is to allow the people to have the right to review the action of the legislature on any act which is passed by the usual simple majority of both houses and signed by the Governor. In order to do this, it is provided that such an act shall not take effect until ninety days after the legislature has adjourned, thus giving ample opportunity for the voters to study it, and, if they desire, to require that it be submitted to them at the next general election, for their approval or disapproval. The makers of the Constitution, realizing, however, that some necessity might arise, which would require a law to take effect immediately, have provided a special manner in which this need can be met. It is by the use of the so-called "emergency clause." It will be observed that this emergency clause never affects in any manner the substance of the proposed law. It refers only to the time when it shall take effect. It would seem, then, that when the legislature has observed every constitutional provision in regard to the substance of a law, but has failed to comply with one which relates to nothing but the time the law shall take effect, the spirit of the Constitution would be obeyed by holding that the measure is valid as an ordinary law, because every requisite of such a law has been fully carried out, but that it shall not take effect immediately, because the constitutional provisions which authorize such a result have not been followed.
The second point in the majority opinion is that "members frequently support measures containing the emergency clause when they would not do so without it." We think while it is obvious that a member of the legislature, who believed in the principle of a bill, and was willing to vote for it on its merits, might not be willing to take from the people the right to review his action, it would be rare, indeed, that a member, who was so thoroughly convinced of the vital necessity of a measure that he was willing, not only to support it on its merits, but also to remove it from the referendum provision of the Constitution, would not be willing to support it as an ordinary measure, on the theory that "half a loaf was better than no bread."
There are no cases cited in Cox v. Stults, supra, in support of the majority opinion on the point under discussion, while there is a wealth of authority which upholds the contrary view. This court itself has held against the rule laid down therein, in the case of Santa Cruz County v. McKnight, 20 Ariz. 103, 177 P. 256, 259. In that case, the question of when chapter 80 of the Session Laws of 1917 took effect was before the court. This chapter carried the emergency clause, and received the necessary number of votes to make it effective as an emergency measure, if it were not vetoed by the Governor. He neither signed nor vetoed it, and we held, under those circumstances, that while it could not take effect as an emergency measure, it did take effect ninety days after the adjournment of the legislature.
It is true that in the later case of Clark v. Boyce, supra, we held that the failure of the Governor either to sign or veto an emergency measure was in effect an approval thereof, but it is obvious that this last opinion was in no way in conflict with the holding in Santa Cruz County v. McKnight, supra, that an act which failed to take effect as an emergency measure, because of some defect in the proceedings required by the Constitution for such a measure, would be perfectly valid as an ordinary act, if it satisfied the provisions of the fundamental law in regard to such an act.
A similar question has been before the appellate courts of a number of states, which have provisions in their Constitutions similar to ours in substance, if not in form, providing for a referendum on ordinary measures taking effect at some period, generally ninety days, after the legislature adjourned, but allowing measures which carried a so-called "emergency clause" and were adopted in a certain special manner to take effect immediately and to be exempt from such referendum.
In the case of Riley v. Carico, 27 Okla. 33, 110 P. 738, 740, it was contended that a certain bill which carried the emergency clause was not one of the class to which such a clause could be constitutionally applied, and was therefore void. The court, referring to the matter, said:
". . . Such an act cannot be put into immediate effect by declaring an emergency. . . . The fact that the action of the Legislature in declaring an emergency was void did not invalidate the act, but resulted in its taking effect 90 days after the adjournment of the Legislature at which it was passed."
A similar question arose in the case of Gayman v. Mullen, 58 Okla. 477, 161 P. 1051, and the court held that, notwithstanding the act on its face was declared to be an emergency measure, and was unconstitutional as such, it took effect as an ordinary measure ninety days after the adjournment of the legislature.
The Supreme Court of Oregon in State v. Coos County, 115 Or. 300, 237 P. 678, 679, in reply to the argument used against a measure which carried the emergency clause, that it was of a nature which could not be passed as an emergency matter, said:
". . . Even if this section had any application to the act in question, such application would not go to the extent of rendering the whole act void on the ground contended by the defendants, inasmuch as, even without an emergency clause, the act would have gone into effect before this writ was issued or application made therefor. Even if the emergency clause was unauthorized, the act itself would be valid, and was in full effect 90 days after the adjournment of the Legislature. . . ."
The Supreme Court of Arkansas in Hargrove v. Arnold, 181 Ark. 537, 26 S.W.2d 581, referring to the time a certain act of the legislature took effect, said:
"We take judicial notice of the records of both branches of the General Assembly from which we know that the Legislature of 1929 adjourned March 14, and that, while the act in question contained an emergency clause, no separate vote or roll call was had thereon, and, therefore, said emergency clause was never adopted and the act did not go into effect until ninety days after the adjournment of the Legislature."
The Supreme Court of Texas in Popham v. Patterson, 121 Tex. 615, 51 S.W.2d 680, 683, referring to an act which carried the emergency clause, but did not receive the vote requisite to make such clause effective, said:
". . . As the act did not get the required recorded vote to put it into immediate effect, this notice did not become operative until the act became a law, which was 90 days after the adjournment of the Legislature. . . ."
A similar question was also before the Supreme Court of South Dakota in the case of In re Opinion of the Judges, 58 S.D. 72, 234 N.W. 671. The act which they were required to construe contained an emergency clause. In South Dakota, under its Constitution, the Supreme Court must finally decide whether, as a matter of fact, there really was an emergency, notwithstanding the fact that the legislature had declared it to exist, and the court, in the case cited, holding that no emergency existed, declared that the law was valid, but only took effect when the constitutional period had elapsed.
So far as we know, there is no case considering a like question which has held to the contrary. If, therefore, the question was one of first impression, it would seem that on reason, and by the unanimous weight of authority, an act, which has been passed with all the constitutional formalities required for an ordinary law, but which has not received the vote necessary to make it effective immediately, does become effective as a valid law ninety days after the adjournment of the legislature which passed it, unless it is referred to a vote of the people in the constitutional manner. In order so to hold, however, we must disregard the rule of stare decisis and overrule our holding in Cox v. Stults, supra, for there is no possible way in which the facts in this case can be distinguished from the facts in that.
The rule of stare decisis has for its object the salutary effect of uniformity, certainty and stability of the law, and, stated in simple form, is really nothing more than the application of the doctrine of estoppel to court decisions. It finds its support in the sound principle that, when courts have announced, for the guidance and government of individuals and the public, certain controlling principles of law, they ought not, after the principles have been promulgated and after these constructions have been published, to withdraw or overrule them. While it is perhaps of the greater importance, as to far-reaching principles, that the court should be right than merely in harmony with previous decisions, it is nevertheless also important that there be stability in the courts in adhering to decisions deliberately made and after ample consideration. Parties should not be encouraged to seek a re-examination of determined principles and to speculate on a fluctuation of the law with every change in the expounders of it, and substantial justice may often be better promoted by adhering to an erroneous decision than by overthrowing a rule once established. While this rule applies with greater force in the case of private affairs, where contractual and property rights have grown up in reliance upon the original decision, it nevertheless is of importance in questions where the rights of the public, as a whole, are concerned. This rule, however, is subject to exceptions.
While, under our judicial system, all courts have a strong respect for precedent, this respect is a reasonable one which balks at the perpetuation of error, and the doctrine of stare decisis should not prevail when a departure therefrom is necessary to avoid the perpetuation of pernicious error. But, except in the cases of the gravest importance, where the law, as previously announced, has the most far-reaching and vital effect, and where it conclusively appears to be erroneous, it is only when a tribunal is unanimous in its later opinion that the rule of stare decisis should be disregarded, for the overruling of a former decision by a mere majority would be to put the law in a state of doubt and uncertainty, which would be worse than a concededly wrong settlement of it. As was said by Chief Justice GEORGE in the case of Lombard v. Lombard, 57 Miss. 171, in discussing the propriety of overruling an opinion previously given by a divided court:
"That the former decision was made by a divided court rather increases than diminishes the duty to adhere to it, when a departure from it must result alone from my non-concurrence in the views of my predecessor, the other two judges participating in the former judgment adhering to the opinions then formed and expressed by them. Moreover, a majority of this court are invested by the Constitution with all the powers which can be exercised by a full bench. This Court has no distinct substantive power to declare a rule of law. It has power to render judgments, and, as an incident to this power, may, in the exercise of it, declare a rule, which becomes a precedent for similar cases. The power of a majority to render judgments is as ample as is that of the full bench; and necessarily the incidental power of declaring the rule which governs the case must exist also. How much a division among the judges shall detract from the force and authority of the decision as a precedent, it is difficult to define. That a subsequent full bench, all concurring in the impropriety of the first ruling, would depart from it with less hesitancy in cases where a departure is allowable than they would overrule a unanimous decision, is probable. But this does not help the difficulty here; for, if the former decision be overruled, it will be by a majority only, one judge being thoroughly convinced that it is correct. And so, if a ruling by a majority is not to be regarded as settling the law, the overruling of that decision by a mere majority will leave the law in that state of doubt and uncertainty, which is worse than a concededly wrong settlement of it."
In this case, however, there are several reasons why we should not hesitate to set aside the holding in Cox v. Stults, supra. In the first place, the question does not directly involve private rights, and no rights are likely to have, as yet, grown up by reason of that decision, so that an altering of the rule laid down therein will not affect property rights already accrued. In the second place, that decision itself disregarded the previous unanimous opinion of this court in the case of Santa Cruz County v. McKnight, supra. In the third place, it is utterly unsupported by any authority, directly or indirectly. In the fourth place, it is avowedly based on an application of the strictest rules of grammatical construction, instead of the universally accepted principle that Constitutions should be liberally construed to effect their broad general purposes, and upon a presumption of human conduct under given circumstances, which is contrary to the ordinary customs of the average man.
For the foregoing reasons, reluctant as we are to depart from the rule of stare decisis, we think this is one of the cases which require such a procedure, and we therefore revert to the rule previously laid down by us in Santa Cruz County v. McKnight, supra, and hold that a bill, which has been approved by the legislature and the Governor in every manner required by the Constitution for it to take effect as an ordinary measure ninety days after the adjournment of the legislature, is a valid law, and does so take effect, notwithstanding there has been an attempt made to cause it to take effect immediately, through the means of an emergency clause, which is ineffective because it was not approved in the manner required by the Constitution. Such being the case, and more than ninety days having elapsed after the adjournment of the legislature which adopted chapter 16, supra, without a referendum being invoked against its provisions, it is in full force and effect, and it is the imperative duty of the respondent to sign the tax anticipation bonds involved in this proceeding.
The alternative writ of mandamus is made peremptory.
ROSS, C.J., concurs.
I regret that it is not possible to concur with my colleagues in the conclusion reached by them in this case. Their opinion is certainly not in accord with either the letter or, as I see it, the spirit of the Constitution.
The holding in the case of Cox v. Stults Eagle Drug Co., 42 Ariz. 1, 21 P.2d 914, that an emergency measure which receives the affirmative vote of fewer than two-thirds of the members elected to each house does not become effective either as an emergency or an ordinary act, is exactly what the framers of the Constitution intended. However, this point was briefed but slightly in that case, the stress there being placed on the main proposition, but the pertinent language of the Constitution itself, when rightly construed, could lead to no other conclusion. I have re-examined the question presented in this case with the hope, since a matter of such importance depends upon it, that it might appear that the opposite conclusion could just as properly have been reached in that case, but nothing has been advanced, or even suggested, indicating that the view expressed then was other than correct.
A proper construction of subsection 3, section 1, part 1, article 4 of the Constitution of this state, especially that portion of it reading as follows, can lead to no other result:
"Provided, that no such emergency measure shall be considered passed by the Legislature unless it shall state in a separate section why it is necessary that it shall become immediately operative, and shall be approved by the affirmative votes of two-thirds of the members elected to each house of the legislature, taken by roll-call of ayes and nays, and also approved by the Governor."
This language is so clear that it needs no construction; its meaning is unmistakable. It states positively and definitely that no emergency "measure" shall be considered passed, unless it receives the affirmative votes of two-thirds of the members elected to each house, and it should be kept in mind that the word "measure," as used throughout this section, refers to the entire act voted on by the legislature — from its title to its last word — and not the emergency clause only. And by the term "passed" is meant that the act has received in the proper way the affirmative votes of two-thirds of the members of each house of the legislature; hence, the expression, "no such emergency measure shall be considered passed by the Legislature," clearly signifies that it shall not be treated or regarded as having become law at all. And just how, in view of this positive statement to the contrary, it can be held that it did become law, in the absence of a two-thirds vote, and operative later on, I do not understand. It could be interpreted to mean this only by reading into it additional language, such, for instance, as this: "No such emergency measure shall be considered passed at that time," or "no such measure shall be considered passed as an emergency," or language indicating that a failure to receive a two-thirds vote would not interfere with its becoming a law but merely postpone its operative date, but nothing of this kind is there.
Another matter entitled to great weight in construing the meaning of this provision is the fact that the legislative branch of the government has given it this construction since its first session, twenty-two years ago. During all these years it has had among its membership many good lawyers and excellent citizens, some of whom helped write the Constitution, but no one of them, so far as my knowledge goes, has ever claimed that it should be given such a meaning. The legislature has proceeded at all times upon the theory that when an emergency act fails to receive a two-thirds vote, that clause should be stricken and the act passed as an ordinary measure. This construction is a practical one by the branch of government for whose guidance it was placed in the Constitution and should be given great, if not controlling, weight. To construe it as meaning that an emergency measure receiving less than two-thirds but more than a majority of the vote of each house becomes a law when approved by the Governor but not operative until ninety days after adjournment places the legislature in the anomalous position of enacting as an ordinary measure a law it desires to pass only as an emergency, a thing that was never contemplated by the framers of the Constitution. If the legislature desires to pass a measure as an emergency without running the risk of its receiving a vote of less than two-thirds of the members of either house and in consequence of this becoming an ordinary act, it was clearly intended that it should have this right. In addition, no one will be sure of his ground in examining the Session Laws, unless he first investigate the legislative journals and ascertain when a measure containing the emergency did in fact become operative.
The majority opinion is based upon the proposition that any member of the legislature who supports an emergency measure believes in the legislation it proposes and will support it in an ordinary act, a proposition the fallacy of which is illustrated by the situation that arose last year over the sales tax legislation, and this is merely one case that is typical. No one at all familiar with the situation confronting the legislature at that time would, I take it, contend for a minute that that body would have even attempted to pass the sales tax as an ordinary measure, because the members, as practically everyone acquainted with the circumstances was aware, knew that such an act had to become effective immediately or not at all, since it was so regarded at that time by the public that it would have been referred to the people if an opportunity to do so had presented itself, and not voted on until November, 1934. It seems clear that it was just as much the purpose and spirit of the emergency provision to give the legislature a right to pass on emergency measures disassociated from ordinary acts as it was to enable it to pass on ordinary acts disassociated from emergency measures. Even though it be true that the emergency deals only with the time an act takes effect the Constitution says specifically that a measure containing it shall not take effect at all (be considered passed) unless it receives a two-thirds vote.
The majority give three reasons for reversing the ruling in the Cox-Stults case, but preliminary to them they state that inasmuch as the question does not directly involve private rights and since in all probability none of this character has been acquired under it there is no reason why it should not be reversed if the facts and the law render it necessary to do so. This is clearly correct and if it appeared that the holding was wrong I would not hesitate a moment to set it aside.
But I desire to consider briefly the reasons given for this action, the first one being that no authorities were cited in support of the ruling in that case. This statement is made in the face of the fact that the majority must know that there is not another Constitution in the land having a provision like, or even similar to, the Constitution of Arizona which states as plainly as the English language can make it that "no such emergency measure shall be considered passed by the Legislature unless it shall . . . be approved by the affirmative votes of two-thirds of the members elected to each House of the Legislature," an expression so clear that a schoolboy can understand its meaning. What further authority is needed than the Constitution itself, when its language is so certain and unambiguous that there is no occasion for interpretation? It is the only authority on either side of the subject to be found anywhere, and this being true decisions from the courts of Oklahoma, Arkansas, Texas, North Dakota or Oregon have about as much relevancy to the question of the use of the emergency in Arizona as does the rule in Shelley's Case. What their courts may say about its function in their respective states would be entitled to as much weight here as the decisions from states in which contributory negligence as a matter of law still defeats a claim for damages would have in Arizona on that question. There is just as much reason for citing opinions from such jurisdictions on the one question as the other. Neither could possibly have the slightest relevancy.
The next ground advanced for reversal is that the opinion is based on a technical rather than a liberal construction and that this is contrary to the rule which governs the interpretation of the provisions of a Constitution. It is undoubtedly true that they should be construed to accomplish the evident purpose of the framers of that document, but it is not my understanding of liberality of construction that it goes to the extent of permitting a court under the guise of liberally interpreting language to set aside and absolutely nullify by judicial construction a provision that is so plain it cannot be misunderstood, and especially where the language of the Constitution itself provides that "the provisions of this constitution are mandatory, unless by express words they are declared to be otherwise." Section 32, art. 2, Constitution.
The people reserved in the Constitution the right to pass on all measures enacted by the legislature except certain ones needing earlier operation, and to allow the opportunity to refer these to a vote of the people the Constitution provides that none of the acts passed shall become effective until ninety days after the adjournment of the legislature, except those whose earlier operation is required. It became necessary, therefore, to formulate a plan by which such measures could become earlier operative, and the one the framers of the Constitution adopted to bring this about was the emergency clause which says: "Provided, that no such emergency measure shall be considered passed by the Legislature unless it shall . . . be approved by the affirmative votes of two-thirds of the members elected to each House of the Legislature." The letter and very spirit of the proposition is that since these emergency measures are to become immediately operative and the people's right to vote on them taken away they should not "be considered passed by the Legislature," that is, become law, unless two-thirds of the members of each house should vote for them. Before a measure can become effective at all, either immediately or at the expiration of ninety days, it is necessary that it first be passed — at least this has always been my view of the law — and if it does not receive that vote, I do not know how it can be considered passed, and the majority opinion does not tell me, though it proceeds upon the theory that this may be done. The language is not that two-thirds shall vote for it before it shall become operative but before it shall become a law at all, though according to the opinion of the majority it becomes effective ninety days after the legislature adjourns notwithstanding it has never passed. The opinion fails to point out, however, by what process of legerdemain this result may be brought about. There is no way it can be done except by walking over the body of this constitutional provision as it lies slain by the sword of judicial construction.
The third reason for setting aside the Cox-Stults holding is that the court in that case disregarded a unanimous decision rendered in Santa Cruz County v. McKnight in December, 1918. In that case it became necessary to pass on the contention whether chapter 80, Session Laws of 1917, which contained an emergency clause and passed both houses of the legislature by a two-thirds vote, was a bar to appellee's claim for salary, and the court said:
"This chapter did not take effect until 90 days after the adjournment of the session of the Legislature at which it was passed; it not having received the necessary approval of the Governor to make it an emergency measure."
Just ten months later, however, the exact situation presented itself in Clark v. Boyce and the court there held that the same measure did become effective as an emergency act, the Governor's tacit approval being sufficient for this purpose. This was a direct reversal of the ruling in the McKnight case that it did not become effective as an emergency but did become operative ninety days after adjournment. If, as the court correctly decided in the later case, an act, under such circumstances, becomes operative immediately, it does not do so ninety days later, since a measure cannot become effective on two different dates. A ruling which results from an admitted error and is made without discussion of the constitutional provision involved is not a very firm basis upon which to rest the rule of stare decisis. And the fact that the statement of the court in the McKnight case which again becomes the rule in Arizona was made without any serious deliberation is shown by the language of the court itself in overruling the holding in that cause. It said in Clark v. Boyce:
"In Santa Cruz County v. McKnight . . . and Ross v. Cochise County, 20 Ariz. 167, 177 P. 931, a contrary view as to when an emergency act takes effect, to what we here announce, was expressed. In neither of these cases was the question formally briefed or argued to the court, and the question was not given that consideration to which it was entitled. Besides, in neither case was it necessary to a decision."
It occurs to me that a ruling made in this way, and especially one having the effect of setting aside an important provision of the Constitution, could not be re-established as a guide for the legislature and the courts of this state with any very great degree of assurance as to its soundness.
The implication in the majority opinion that the construction given the emergency provision of the Constitution in the Cox-Stults case is a pernicious one has no legal effect and is, to say the least, of a tone that calls for no comment.
There is, to my mind, no question but that the framers of the Constitution intended that a measure containing an emergency clause should become effective immediately or not at all. Therefore, the act providing for anticipation bonds never became a law.