(dissenting). I make no judgment, public or private, as to whether "choice" is good public policy. That issue is not presented nor is it appropriate for us to so decide.1 But no one can disagree *566that "choice" is major public policy involving fundamental educational decisions. And no one can disagree that it merited full legislative consideration.
It did not receive such consideration. In fact, it received no consideration whatsoever in the senate.
"Choice" was never debated in the senate. It never received a public hearing in the senate. No expression of public sentiment was ever sought by the senate nor received. There was no separate vote taken on it in the senate. It passed the senate as part of the budget bill four legislative days after the senate received it as a separate piece of legislation from the assembly. See 1989 Wisconsin Assembly Bulletin, 169; 1989 Senate Bulletin 148-149. The committee in the senate to which the original bill was referred never even dealt with it.
Yet the majority opinion inexplicably concludes that "choice" was "greatly debated in legislative committee public hearings and by the entire legislature." Majority op. at 512 (citation omitted) (footnote omitted). "[W]e find no evidence in this case that suggests the program was smuggled or logrolled through the legislature without the benefit of deliberate legislative consideration . . .. Clearly, the legislature 'intelligently participate^) in considering' this program." Id. at 522-523.
The evidence, contrary to the assertions of the majority opinion, is overwhelming that the senate never "intelligently participate^] in considering" this program. On Wednesday, March 15, 1990, the Wisconsin Assembly passed Assembly Bill 601, The Milwaukee Parental Choice Program, and sent it to the Wisconsin *567Senate. It was immediately referred to the Senate Educational Financing Committee where no action was ever taken. Five days later (which includes Saturday and Sunday), on Monday, March 20, 1990, this bill was tucked into the budget bill by the Joint Finance Committee. One day later, on Tuesday, March 21, 1990, the budget passed the senate. The "choice" plan was part of that bill. See 1989 Wisconsin Assembly Bulletin, 169; 1989 Senate Bulletin 148-149.
The majority, having concluded that "choice" was debated extensively by the legislature, affords it a presumption of constitutionality. The majority then, after analyzing only one of two classes that the legislation creates, concludes that it is not a private or local bill within the meaning of article IV, section 18.
I agree with Chief Justice Heffernan that the legislation fails the "private or local" constitutional prohibitions of article IV, section 18 with respect to the classification of school children who live in the city of Milwaukee. That is the first classification created by the law, and that is what the majority addresses. What most observers are unaware of, and what the majority does not address, is that the bill creates a second classification which also violates art. IV, sec. 18: private schools located within the city limits of Milwaukee. These are the only eligible recipients of the state's $2.5 million annual outlay for this program. This is an annual outlay, payable only to a small group of eligible private schools, and will continue to be paid for out of state taxpayers' funds, unless and until it is repealed by the legislature. If the legislature wanted this to be law, it could constitutionally do so only as a separate piece of legislation, considered separately by each house of the legislature, and not as part of a "must pass" omnibus budget bill. Including private legislation in a "must pass" omnibus *568budget bill, particularly when that legislation receives no consideration in one house of the legislature other than the vote on the budget itself, is precisely what leads to the internal logrolling in the legislature which members of the majority opinion have in the past found so deplorable.2 Accordingly, I dissent.
The legislation in question provides that only school children in school districts located within cities of the first class may participate; their "choice" is limited to private schools located within the city of Milwaukee. Thus, the legislation adopts two classifications: 1) school children residing in cities of the first class and attending school districts within cities of the first class; and, 2) private schools located within cities of the first class. The majority opinion addresses only the first classification and finds it constitutionally unobjectionable because, in essence, cities of the first class have the most educational problems (the first prong of the classification tests, e.g. real differences); and because this legislation is "experimental" in nature (the second prong of the classification tests, e.g. germaneness).
Missing in the majority's analysis, completely missing, is any meaningful discussion whatsoever with *569respect to the second classification that this legislation also adopts; private schools located within cities of the first class. Had the majority subjected this second classification to the very same classification tests they applied to the first classification, it could not pass constitutional muster.
The first prong of the classification tests provides that the classification employed must be based on substantial distinctions which make one class really different from another. How are private schools located within cities of the first class "really different" from all other private schools located in the state of Wisconsin? To ask the question is to answer it; there are no differences. None are posited by the petitioners, none are discernible. Yet under this legislation a private school located within the city of Milwaukee can be the recipient of the state's largesse, a private school located just outside the city limits cannot. One can only conclude that the authors of this legislation intended to benefit only private schools located within the city, and there are no reasons given to support that discrimination.
The second prong of the test provides that the classification adopted must be germane to the purpose of the law. The majority opinion argues quite cogently that this is "experimental" legislation. The petitioners argued this same point extensively in their briefs and at oral argument. Assuming both petitioners and the majority are correct in that hypothesis, then how is it that only private schools located in the city of Milwaukee can test that experiment? Why not private schools located in the suburbs of Milwaukee, or any other private school? The classification adopted, private schools located in the city of Milwaukee, is simply not germane to the avowed purpose of educational experimentation. Any other private school, located anywhere in the state, is equally capable *570of providing the documentation needed to assess this experiment. Again, just as in the first test, one can only conclude that the authors of this legislation intended to benefit only private schools located within the city, and there are no reasons given nor discernible that support that restriction.
The legislation as drafted puts the emphasis on the first classification. But the above analysis becomes clearer if one simply re-states the legislation and puts the emphasis on the second classification. Assume the legislation said: "Any nonsectarian private school located in the city (of the first class) shall receive $2,500 per year for each student who resides in the city (of the first class) and attends the private school providing that all of the following apply: (Here, the bill would state all the criteria listed in the actual legislation)." With this re-drafting, everything ends up the same as the original legislation. But now it becomes clear why this legislation is constitutionally objectionable. "Why should private schools in Milwaukee be treated preferentially?" one would legitimately ask. "Why should they get this $2.5 million annually and not us?" private schools in suburbs of Milwaukee and other cities in Wisconsin would ask. "What is it about them that makes them different from us?" The answers are obvious. There are no reasons.
I do not doubt the sincerity of the authors of this legislation with respect to their belief that "choice" is good public policy. It may be, it may not be. I make no judgment as to that. Perhaps school children who reside in Milwaukee will be major beneficiaries of such a program. But this legislation also targets another beneficiary, a very small group of private schools located only in the city of Milwaukee who will collect the amount of $2.5 million annually. This benefit is not subject to debate. It is their's until the legislature decides otherwise.
*571There is a principle at stake here which has been cited numerous times in our previous cases; legislation which benefits only a few must rise or fall on its own merits, and not be a part of a "must-pass" bill. The basis for this principle was recently stated in a concurring and dissenting opinion in Milwaukee Brewers v. DH&SS, authored by Justice Ceci:
The prison siting legislation, buried deep within the budget bill, represents the very worst of the logrolling and railroading practices which have become all too commonplace in the legislature.
The very design of art. IV, sec. 18 is disregarded in the legislative practice whereby a provision such as the prison-siting legislation is included in a budget bill. Such a practice breeds unaccountable representation: it necessarily forces a legislator to vote once on two separate matters. A legislator is forced to vote on a matter of statewide importance and prominence — the budget in this instance — the same way in which he or she will vote on a wholly unrelated subject — here, the siting of a prison in the Menomonee Valley. An affirmative or negative vote on the overall bill necessitates the vote extending to all subject matter within the bill. I find such a practice to be deplorable and untrue to the spirit of art. IV, sec. 18. Certainly the representatives' respective constituencies, which may well have different opinions about the logrolled issues, deserve to have their views be fully represented by separate voting on separate issues.
We do not require that the general electorate vote a straight party ticket; we should not tolerate legislative practices which dictate that only a single vote be cast on wholly separate issues. Such a practice is, at best, a modified form of logrolling, which is *572prohibited by statute. Such a practice destroys the accountability of our representatives and reduces the legislature to an internally acquiescent institution, unresponsive to the constituency it purports to represent. Milwaukee Brewers v. DH&SS, 130 Wis. 2d 79, 156-157, 387 N.W.2d 254 (1986) (footnote omitted).
This principle was also discussed by a different justice in the same opinion in his dissenting and concurring opinion:
The [majority's] test still requires legislators to vote for a comprehensive budget bill with its many concerns and fiscal necessities without voting directly on matters of private or local effect. Accountability is sacrificed, not because legislators are unaware of the private or local provisions of the budget bill, but because they cannot vote their convictions on such provisions without affecting the entire budget bill. Contrary to the majority's conclusion, therefore, a legislator could credibly claim to oppose a local or private provision, despite voting for the entire budget bill. Milwaukee Brewers, 130 Wis. 2d at 145.
The principles stated in these prior opinions have however been ignored by their authors who inexplicably have joined the majority in this case. The majority opinion here glosses over these principles by pointing out that a similar separate bill had been passed by the Assembly and "all" the Senate did was include it in the omnibus budget bill. But that gloss completely disregards the legislative history of "choice" in the senate. As explained above, this was never debated in the senate, it never received a public hearing in the senate, there was no expression of public will. It passed the senate as part of the budget four legislative days after it was received from the assembly.
*573The majority's gloss also ignores another obvious implication. If there were sufficient votes in the Senate to pass the bill as a separate piece of legislation, the Senate would have done so, thereby avoiding any possible constitutional challenge under this section. Given that they did not do so, it is clear that the votes were not there to pass it as a separate piece of legislation. It needed to be tucked into the budget in order to snare otherwise negative votes of senators who felt they had to vote for the omnibus budget bill because of other policy items in the budget they supported which had a higher priority than this "choice" legislation.3 That is precisely what Section 18 seeks to prohibit; it is precisely what our former opinions attempted to address.
I turn next to the discussion in the majority opinion regarding the presumption of constitutionality that should or should not attach to this legislation. The majority adopts a middle ground which will only serve to confuse. Better had they simply stated that either a presumption of constitutionality always attaches to this type of legislation or it does not. From their opinion, one can only guess as to how much deliberation is sufficient for the presumption to attach.
I conclude a presumption of constitutionality should never attach to legislation that is challenged as being procedurally unconstitutional, as opposed to legislation *574that is challenged as being substantively unconstitutional. The challenge under article IV, section 18 is a procedural challenge. The other challenges to this legislation are substantive, and therefore deserve the presumption. But the two challenges are completely distinct and should be treated as such.
The procedural challenge here asserts that the legislature failed to follow essential procedural steps mandated by our constitution. The challenge, in essence, is that the legislation on its face is private or local and was included in a multisubject bill, and is therefore violative of article IV, section 18. No one disagrees that on its face the legislation is private or local and was part of a mul-tisubject bill. Why, then, is such legislation entitled to a presumption that it is constitutional? On its face, it clearly is not. Neither logic, common sense, nor precedent requires a presumption of correctness when on its face it is not correct. Our only obligation in such a situation is to determine whether the legislation fits within one of the narrowly circumscribed exceptions that have been carved out by this court. If anything, logic would tell us that legislation that on its face is unconstitutional starts with a presumption that it is not constitutional. But I do not argue for that proposition; I urge only that in such a situation, no presumption should attach either way. When legislation that is private or local on its face and could have been passed as a separate piece of legislation, with all the legislative scrutiny that entails, is instead passed as a part of a multisubject bill, it does not warrant a presumption of constitutionality. Because of the potential for abuse that is present in such a situation, it deserves careful scrutiny with no presumption attaching.
Perhaps an example might make this clearer. Assume that an Assembly Bill granted a liquor license to *575John Jones of Middleton. Assume also that this bill received the precise treatment that the legislation at issue in this case received. That is, assume it received a public hearing in the Assembly committee with much public testimony, and then assume it passed the Assembly; that it went to the Senate, but the Senate did not consider it separately; and assume that it was then included in the multisubject budget bill and then passed. Assume then the liquor license legislation (that was passed into law as part of the budget) is then challenged as being a "private or local" law. Under the majority's view, this legislation would be entitled to a presumption of constitutionality. That strikes me as being absurd. Yet it is not different from the legislation before us. Other examples of classification legislation could be equally forthcoming. Legislation that on its face is "private or local" (as is classification legislation as well as specific entity legislation), that is part of a multisubject bill, that is challenged as being violative of section 18, has no presumption of constitutionality.
The majority's conclusion rests on their belief that this legislation deserves the presumption because of the attention this issue received in the process. Putting aside the problem addressed earlier in this dissent that the senate never even debated it, that conclusion invites confusion. What, in the future, will constitute sufficient "attention" so as to deserve the presumption? The scenarios under which a bill that passes one house but gets sidetracked in the other, and then appears in the budget bill, are almost infinite in number. And yet the majority gives the same presumption of constitutionality to that situation as attaches when both houses pass the bill. The presumption does not apply, if for no other reason than the simple fact that when a bill passes one house but fails to pass in the other, and then magically appears in *576the budget, that usually means there were not sufficient votes to pass it on its own merits. Nothing of significance in the legislative process "just happens."
The majority does not need their presumption analysis to reach the result they reached. They should discard it in favor of the black letter rule which this court adopted three years ago. Only confusion can result.
In conclusion, the result reached by the majority leaves the law regarding article IV, section 18 in serious disarray in two major respects. First, with respect to classification legislation, the result in this case cannot stand alongside the recently decided case of Brookfield v. Milwaukee Sewerage, 144 Wis. 2d 896, 426 N.W.2d 591 (1988). Although Brookfield involved Milwaukee sewers and this case involves Milwaukee schools, the principles are precisely the same, the results are diametrically opposed. Members of the public, practitioners, and perhaps most importantly the legislature, cannot now state nor predict with any degree of certainty what the law is regarding whether a piece of legislation is "private or local." The second serious problem involves the presumption of constitutionality and when it attaches. The majority's test of legislative consideration simply cannot stand the test of time. The majority's test has no certainty, and therefore no predictability.
In a concurring opinion of less than one page, the refrain "Let's give choice a chance!", or similar verbiage, is repeated four times. The issue here is not whether we agree with the policy choice made by the legislature, the issue is the process by which the policy was enacted into law. The policy of whether "choice" should be law in this state is a legislative decision. It is no more *566appropriate for judges to applaud a policy decision of the legislature in this context than it is to disparage it. When the court challenge is based on process, it is totally inappropriate and judicially indefensible for judges to base their decision on whether they agree with the policy or not.
The majority opinion, in footnote 8 expresses concern at what it perceives to be this dissent's "indictment of the legislature's integrity." That is utter codswallop! The challenge here is to the process by which a bill becomes law. The conclusion of this dissent that the process was constitutionally defective is no more an indictment of the legislature's integrity than were a number of past decisions of this court, some of which the author of this majority opinion participated in and agreed with, that found other legislation "private or local" and therefore violative of article IV, section 18. See, e.g., Soo Line R. Co. v. Transportation Dep't, 101 Wis. 2d 64, 303 N.W.2d 626 (1981); Brookfield v. Milwaukee Sewerage, 144 Wis. 2d 896, 426 N.W.2d 591 (1988).
The majority suggests, in a footnote that responds to this part of the dissent, that "a plausible alternative explanation could include the Senate's concern that a worthy piece of legislation may be thwarted by the close of a legislative session." Majority op. at 522. If that is a plausible alternative explanation, it is equally repugnant. Is the majority suggesting that "worthy" legislation can escape the constitutional imperatives of article IV, section 18 if such legislation comes up at the end of the legislative session?