(dissenting). The issue is whether the Legislature may impose a tax on a specific activity in a specific location in order to generate revenue to assist financially the specific county and the specific municipality in which that activity occurs. In this case, I would hold that it may not. I dissent because I fail to see how the Legislature’s purpose in relieving financially distressed Wayne County is reasonably related to imposing an excise tax on airport parking at private lots located solely in the Detroit Metropolitan Airport area. Accordingly, I would hold that the Airport Parking Tax Act, MCL 207.371 et seq.; MSA 7.559(101) et seq., is unconstitutional under the Headlee Amendment, Const 1963, art 9, § 31,1 because it was levied without local voter approval.
i
Generally, there are two means of interpreting a constitutional limitation on the government’s authority to impose taxes. The first would be to determine whether the statute violates the literal language of Const 1963, art 9, § 31, which proscribes units of local government from levying or increasing taxes without voter approval. Under this approach, the airport parking tax arguably could be upheld under *546Const 1963, art 9, § 31, because it was imposed by the Legislature and not by a unit of local government. The majority seems to reject this approach by stating that “the entity imposing the tax in question may not conclusively resolve the Headlee question.” Ante at 534. If the majority is in fact rejecting this approach, then I agree with it because this Court has long looked beyond literal constitutional language. See, e.g., Lockwood v Comm’r of Revenue, 357 Mich 517; 98 NW2d 753 (1959) (a legislatively imposed tax that was styled as a use tax was held to be unconstitutional as an improper sales tax).
Indeed, the proper approach is to look beyond form to substance and to determine whether the practical function of the tax violates the intent of the constitutional provision. Lockwood explained:
There is no middle ground. Either we construe the constitutional limitation literally or we construe it to accomplish its manifest objective and intent.
Which is to be our choice? Actually, there is no choice but one. The unanimous opinion of this Court, written by Mr. Justice Cooley close to a hundred years ago, sets our course in its enunciation of immutable principles:
“Constitutions do not change with the varying tides of public opinion and desire; the will of the people therein recorded is the same inflexible law until changed by their own deliberative action; and it cannot be permissible to the courts that in order to aid evasions and circumventions, they shall subject these instruments, which in the main only undertake to lay down broad general principles, to a literal and technical construction, as if they were great public enemies standing in the way of progress, and the duty of every good citizen was to get around their provisions whenever practicable, and give them a damaging thrust whenever convenient. They must construe them as the people did in their *547adoption, if the means of arriving at that construction are within their power.”
The literal construction of the words, without regard to their obvious purpose of protection, is to make the constitutional safeguard no more than a shabby hoax, a barrier of words, easily destroyed by other words. This canon of constitutional construction we reject. A constitutional limitation must be construed to effectuate, not to abolish, the protection sought by it to be afforded. [Id. at 554-557 (citation omitted).]
Although the majority seems to be applying this approach, I believe that it is choosing to ignore the realities of who benefits from the tax revenue generated, the means and the extent to which the tax is limited in the scope of its applicability, and how the two provisions are wholly unrelated.
n
Although this is the first time that we have considered whether a statutory tax is really a local tax for purposes of a Headlee Amendment challenge, the law is replete with cases determining whether statutory taxes based on classifications2 are local or state taxes under various state constitutional provisions. See anno: Validity of statutory classifications based on *548population — Tax statutes, 98 ALR3d 1083.3 The Headlee Amendment shares a common theme with other constitutional provisions.
In Durant v State Bd of Ed, 424 Mich 364, 383; 381 NW2d 662 (1985), this Court described the purpose of the Headlee Amendment:
[The voters] were striving to gain more control over their own level of taxing and over the expenditures of the state. It is evident that while the voters were concerned about the general level of state taxation, they were also concerned with ensuring control of local funding and taxation by the people most affected, the local taxpayers. The Headlee Amendment is the voters’ effort to link funding, taxes, and control.
The goal of local control underlying the Headlee Amendment parallels the people’s intent underlying Const 1963, art 4, § 29, which provides:
The legislature shall pass no local or special act in any case where a general act can be made applicable, and whether a general act can be made applicable shall be a judicial question. No local or special act shall take effect until approved by two-thirds of the members elected to and serving in each house and by a majority of the electors voting thereon in the district affected.
*549Chief Justice Kavanagh explained the historical purpose underlying Const 1963, art 4, § 29, in Advisory Opinion on Constitutionality of 1975 PA 301, 400 Mich 270, 286-287; 254 NW2d 528 (1977):
The history and rationale of this provision, which first appeared in the Constitution of 1908, was set forth by Justice Brooke in Attorney General ex rel Dingeman v Lacy, 180 Mich 329, 337-338; 146 NW 871 (1914):
“Considering the history of legislation under the Constitution of 1850, it is apparent that there had grown up a pernicious practice on the part of the legislature in passing local acts. The practice was bad in two very important particulars. In the first place, much of the legislation thus enacted constituted a direct and unwarranted interference in purely local affairs and an invasion of the principles of local self-government. In the second place, such legislation affecting as it did certain limited localities in the State, the senators and representatives from unaffected districts were usually complaisant, and agreed to its enactment without the exercise of that intelligence and judgment which all legislation is entitled to receive from all the members of the legislature. This course led to many abuses (principally in amendments to city charters), some of which found their way into the courts, and were there redressed so far as the Constitution then in force would permit.
“With these evils in mind, the Constitution of 1909 [sic] was formulated and adopted by the people. From a reading of the provisions above quoted and others of a similar character, it is, we think, entirely clear that it was the settled purpose of the framers of the new instrument and of the people who adopted it to forever insure to the people the right to control their affairs purely local, and to secure for all general legislation grave attention and the application of the collective wisdom of the legislators.”
Given the analogous purposes underlying the Headlee Amendment and Const 1963, art 4, § 29, I believe we should approach the instant case in the *550same manner that this Court has approached a Const 1963, art 4, § 29 challenge. Under that provision, the challenge has been that a facially neutral statute, which in practicality only applied to Wayne County or Detroit, was really an unconstitutional local act. The test this Court has used was summarized in Dearborn v Wayne Co Bd of Supervisors, 275 Mich 151, 155-156; 266 NW 304 (1936):
The legislature has enacted a number of laws upon a population basis and without referendum requirement, which, in prcesenti, could apply only to the county of Wayne or the city of Detroit. Some of them have been before this court. The principles upon which they have been sustained as general laws or defeated as local acts are well established in this State and elsewhere.
The first test to be applied is whether population has a reasonable relation to the purpose of the statute. In Mulloy v Wayne Co Bd of Supervisors, 246 Mich 632, 635 [225 NW 615 (1929)], the distinction is pointed out:
“Clearly, because of its provision as to population, the act applies to Wayne county only. If it is a reasonable and logical basis of classification, considering the subject of legislation, unquestionably a specified population may be made the test of the applicability of a general legislative act; and under such conditions the act will not be construed to be invalid as local legislation. But where the subject of legislation is such that population has no obvious relation to the purpose sought to be accomplished, an attempt to make the application of the legislative act dependent on population is unwarranted and amounts to local legislation. Attorney General ex rel Dingeman v Lacy, 180 Mich 329.”
The second test of a general law, based upon population, is that it shall apply to all other municipalities if and when they attain the statutory population. It must have — “an open end through which cities are automatically brought within *551its operation when they attain the required population.” [Citations omitted.]
This Court reaffirmed the Dingeman v Lacy test in Avis Rent-A-Car System, Inc v City of Romulus, 400 Mich 337; 254 NW2d 555 (1977), a case factually similar to the instant case. There, the Avis and Hertz rental car companies had received grants of concessions at Detroit Metropolitan Airport pursuant to MCL 259.133; MSA 10.233. They also had leased land from the airport and had built service buildings. 1953 PA 189 related to taxing lessees of what would otherwise be tax-exempt property (owned by the airport). The act provided an exception for concessions at public airports that were available for general public usage. Romulus tried to tax Avis and Hertz for their use of the land at the airport, but lost in suits brought by the companies. Romulus then sought the assistance of the Legislature, which passed 1970 PA 174, limiting the exception to concessions at airports “only in ‘counties of over 1,000,000.’ ” Avis, 400 Mich 343, quoting 1970 PA 174. The Court restated the longstanding rule:
[A] population classification “must have a reasonable relation to the subject matter of the legislation, and must furnish some fairly apparent reason for legislation differing from that applicable to other municipalities having a substantial difference in population.” A population classification “can never be sustained where it is, as in the case at bar, a manifest subterfuge.” [Id. at 345.]
The Avis Court held that the 1,000,000 population limitation was unconstitutional. It reasoned:
1970 PA 174 limits the application of the tax exemption granted by 1953 PA 189 by requiring the concessionaire to *552meet “basic tests.” This is appropriate. However, the legislation has a discriminatory thrust. It is applicable only to airports and only to airports in counties with more than 1,000,000. We require that a reasonable relationship between those qualifications and the limiting of a tax exemption be demonstrated. It has not been demonstrated in this case.
The legislative purpose in 1970 PA 174 was to withdraw a tax exemption from concessions at airports located in counties over 1,000,000 people (only Wayne County). The legislative conclusion must have been that these airports will attract concessions without the lure of a tax exemption. The attraction, however, is not a function of county population. It depends on the volume of traffic flowing through the aiiport. Under this legislation, the exemption is withdrawn not only from concessions at Metro but also could apply to the other two Wayne County airports— Detroit City and Grosse He Municipal Airport. It is difficult to “reconcile the legislative intent . . . with the legislative scheme.”
The trial court in Hertz noted that the defendants had supplied “figures on enplanement and deplanements which show that the overwhelming majority of Michigan’s passenger air traffic taires place at Metropolitan Airport.” Defendants said this was because Metro “is in the state’s most populous county.” The court said that the aiiport traffic was not related to or dependent on Wayne County’s population. The trial court in Avis said “there is no correlation between the county in which an airport is located and the traffic which passes through the airport.”
We do not see a reasonable relationship between the withdrawing of a tax exemption from aiiport concessions and the size of the county where the aiiport is located. We believe the legislation is a local act in an area where a general act could be made applicable. [Id. at 347-348.]
The reasonable-relationship test has long been used by many state courts with respect to determining *553whether a statute with a population classification was really a local act or local tax.4 Such statutes have been sustained where the court found a rational reason for not extending the statute to less populated municipalities. Alexander v Chicago, 14 Ill 2d 261, 268; 151 NE2d 319 (1958) (population density was found to be rationally related to park maintenance and services); Hassler v Engberg, 233 Minn 487, 513-514; 48 NW2d 343 (1951) (population density and proximity of buildings may affect fire risks). In contrast, similar statutes have been struck down where the court did not find such a reason. Sivort Co v State, 136 Fla 179, 184; 186 So 671 (1939) (“the basis for the attempted classification finds no reasonable justification in connection with the subject matter of the Act”). Koons v Atlantic City Bd of Comm’rs, 134 NJL 329, 332-333; 47 A2d 589 (1946), explained:
[T]he law-making authority may subdivide the common-law municipalities into subordinate classes; and the validity of such legislative classification depends upon the existence of distinguishing qualities and attributes related to the subject matter of the legislation. The characteristic constituting the basis of the classification must be reasonably appropriate to the object of the law. Unless it rests upon distinctions that are substantial and not merely illusory, the classification is wanting in the virtue of constitutional generality. The test is whether the statutory class has a logical and reasonable basis, free from artificiality and arbitrariness, embracing all and omitting none naturally falling into that category. Is *554the legislation of such a character as that it is equally appropriate to all forming the statutory class, and is that class embracive of all in like situation and circumstances, and therefore natural members of the class? If, viewed in the light of the legislative design, the necessity and propriety of the classification reasonably appears, it is not within the constitutional interdict.
Green v Arlington Co Bd, 193 Va 284, 287-288; 68 SE2d 516 (1952), echoed a similar test:
The test of reasonableness of classification is said to be whether it embraces all of the classes to which it relates. The basis of the classification involved must have a direct relation to the purpose of the law, and must present a distinction which renders one class, in truth, distinct or different from another class. . . . Laws may be said to apply to a class only, and that class may be in point of fact a small one, provided the classification itself be a reasonable and not an arbitrary one, and the law be made to apply to all of the persons belonging to the class without distinction. [Citation and internal quotations omitted.]
These cases would support a two-pronged test for determining whether the instant enplanement classification is a cloaked local tax: (1) the classification must be open-ended so that other counties, municipalities, and airports, in a situation similar to that which is the aim of the statute, fall within the class, and (2) the classification itself must bear a reasonable relationship to the puipose of the statute. I turn first to the relationship between the Legislature’s purpose in enacting the Airport Parking Tax Act and the number of passengers enplaning at Detroit Metropolitan Airport.
*555m
The airport parking tax was part of a package of bills aimed at addressing Wayne County’s financial situation, at Wayne County’s request. House Legislative Analysis, HB 5164, 5166, 5168, 5170, 5171, 5198-5200 (January 22, 1988). The problem that the Legislature was seeking to address was described as follows:
Wayne County is seeking the state’s help with its massive budget problems. County officials, faced with the possibility of bankruptcy, are threatening drastic cuts and layoffs in essential services, such as law enforcement, health care, libraries, and parks. The county is contemplating closing a floor of the county jail and cutting the prosecutor’s staff, moves that would frustrate anti-drug trafficking efforts just underway. According to House Taxation Committee staff reports, Wayne County owes the state $49 million from a debt settlement agreement in 1984 and some $60 million accumulated since then. A $15 million shortfall is expected in the 1988 resident county hospitalization (rch) budget. Another $52-72 million is owed to non-state sources (e.g., tax anticipation notes issued to meet payroll). The county’s budget has a recurring $10 million operating deficit. The causes of the problems are manifold and complex, but there is little doubt that the cost of providing health care to indigents is a major contributor to the county’s fiscal woes. The rch program was expected to cost $32 million this year (slightly over $19 million to be paid by the state), but instead will cost about $50 million! (Wayne County officials say the second largest rch program, in Oakland County, cost $2.5 million.) Alone among Michigan’s counties, Wayne County is required to run an open-ended indigent health care program where demand determines costs. While the county needs long-term budget reforms, including a way to control health care costs, and new sources of revenue with growth potential, it also needs immediate relief from its overwhelming difficulties. [Id. at 1.]
*556The legislative history reveals that the revenue from the airport parking tax was generated to help Wayne County avoid financial disaster. I would fully agree that the state has a vital interest in Wayne County’s financial stability. But, “[a] laudable purpose does not validate unconstitutional means.” Dearborn, 275 Mich 158.
I find that none of the described problems that the Legislature sought to address are related to Detroit Metropolitan Airport parking. Indigent health care costs, which are a function of the number of indigent residents, are wholly unconnected to the volume of passengers enplaning at Detroit Metropolitan Airport, just as the volume of concessions at an airport is unrelated to the population of the county in which the airport sits. Avis, 400 Mich 348. I do not believe that the majority can genuinely say that this airport parking tax serves a state purpose related to serving a major airport.5
Further, the statute provides:
(1) Beginning January 1, 1988, through December 31, 1988, on the first day of each month, the state treasurer shall make a distribution from the fund to each qualified county in an amount equal to the total amount on deposit in the fund multiplied by a fraction the numerator of which is the population of that qualified county during the immedi*557ately preceding year and the denominator of which is the total population of all qualified counties during the immediately preceding year.
(2) Beginning January 1, 1989, on the first day of each month, the state treasurer shall make a distribution from the fund to each city within which a regional airport facility is wholly located in an amount equal to 20% of the total amount on deposit in the fund minus the amount of revenue generated from any airport parking facility located within the boundaries of the regional airport facility divided by the total number of cities within which a regional airport facility is wholly located. Any surplus funds remaining after the distribution shall be distributed in accordance with (1) above. After the distribution is made under subsection (1), on the first day of each month, the state treasurer shall make a distribution from the fund to each qualified county in an amount equal to the total amount on deposit in the fund after the distribution under subsection (1).
(3) A distribution made under subsection (1) or (2) shall be deposited in the general fund of the city or qualified county.
(4) The distribution provided by subsection (2) shall not be made if all taxing units are authorized by law to impose taxes and the collection is made of taxes imposed under Act No. 189 of the Public Acts of 1953, being sections 211.181 to 211.182 of the Michigan Compiled Laws, on concessions at a regional airport facility. [MCL 207.377; MSA 7.559(107) (emphasis added).]161
The money is in no way earmarked for maintaining the airport that produced the parking that is the subject of the tax. Instead, it is distributed to Wayne County’s and Romulus’ general funds. Moreover, assuming arguendo that another county or city could ever qualify under the statute, the amount that would *558be distributed would be a function of the population of the county — not the number of enplanements. Similarly, the amount distributed to qualified cities would be a function of both the number of cities and of their populations — again, not the number of enplanements.
In any event, by design, the Legislature intentionally drafted the airport parking tax statute to apply only to parking in the Detroit Metropolitan area.7 Yet, even if the majority was correct that Wayne County’s financial distress was related to serving its airports,8 there is no apparent reason for excluding from the *559taxed class the private parking facilities at airports in Wayne County other than Detroit Metropolitan Airport.
In contrast to the majority’s analysis, Lockwood would have us apply an “acid test” to the airport parking tax:
When we put these taxing acts, as we must, to the acid test of examination in terms of their “practical operation,” when we, again as is our duty, “look through forms and behind labels to substance,” what do we find? [Id. at 558.]
I find the following. The Legislature wanted to provide Wayne County with financial assistance, at Wayne County’s request. However, to do so directly would require two-thirds approval by both houses of the Legislature.9 Under the pre-Headlee Amendment constitution, the Legislature could not levy a local tax without two-thirds approval of both houses. Const 1963, art 4, § 29. Under the Headlee Amendment, Wayne County itself could not have imposed a local tax without direct voter approval. Const 1963, art 9, § 31. Therefore, in an attempt to circumvent the requirements of local voter approval and two-thirds approval of both houses, the Legislature drafted a tax classification scheme based on enplanements. Applying the acid test of examination to this tax, I would find a manifest subterfuge by which the Legislature sought to raise money for a local purpose. Moreover, the means chosen are not rationally related to the budgetary problems faced by Wayne County. Further, even if there were a rational connection, the enplane*560ment classification has excluded private parking at other airports that would seem to naturally fall within the burden of a county and city providing services to airports. As such, I find that the classification itself is arbitrary.
Absent a reasonable relation between the purpose of the tax and the classification employed for imposing the tax, the tax cannot be held to have a constitutional state purpose. By upholding this tax, the majority has created a loophole, that is, an opening through which the intent of the drafters of the constitution and of the people adopting the Headlee Amendment in protecting local control over local concerns can be defeated.
IV
I would hold that this tax is unconstitutional in violation of the Headlee Amendment’s purpose of ensuring local taxpayer control over local taxes. I would reverse the decision of the Court of Appeals.
Levin, J., concurred with Cavanagh, J.Units of Local Government are hereby prohibited from levying any tax not authorized by law or charter when this section is ratified . . . without the approval of a majority of the qualified electors of that unit of Local Government voting thereon.
The aiiport parking tax statute applies to private parking within five miles of “an airport that services 4,000,000 or more enplanements annually.” MCL 207.372(h); MSA 7.559(102)(h). I believe this enplanement classification functions much like a population classification, which is a device often used by state legislatures. However, it has long been recognized that population classifications can be used “as a mere cloak for local laws . . . .” Birmingham Electric Co v Harry, 215 Ala 458, 458; 111 So 41 (1926).
In those cases involving the issue of whether a population classification contained in a tax statute renders such a statute invalid, the courts have discussed a variety of different grounds of attack upon such a statute. One frequent argument as to the validity of such statutes has concerned state constitutional provisions prohibiting local or special laws, either altogether or in certain cases only and subject to certain qualifications and exceptions. Generally, it appears well settled that a statute under which counties or municipalities are classified on the basis of population does not fall within a constitutional inhibition against local or special laws, where the classification has reasonable relation to the purposes and objects of the legislation. [Id. at 1087.]
See, e.g., Miller v El Paso Co, 136 Tex 370, 375; 150 SW2d 1000 (1941):
Resort to population brackets for the purpose of classifying subjects for legislation is permissible where the spread of population is broad enough to include or segregate a substantial class, and where the population bears some real relation to the subject of legislation and affords a fair basis for the classification.
To further underscore that the airport parking tax is not a function of maintaining an airport, I note that the statute provides:
The distribution received by a qualified county under this act does not constitute revenue from the use or operation of the airport located within that county and is not subject to any pledge of, lien upon, or use restriction of revenue received or derived by that county from the use or operation of that airport. [MCL 207.382; MSA 7.559(112) (emphasis added).]
MCL 211.181; MSA 7.7(5) applies to lessees or users of tax-exempt property: i.e., the rental car parking lots operating on public airport property.
House Bill 5170 would create, effective January 1, 1988, an excise tax on the operators of airport parking facilities at the rate of 30 percent of the charge for parking. (The tax would apply only to Detroit Metropolitan Airport.) Parking facilities within the boundaries of the airport or within five miles of the boundaries would be subject to the tax. The revenue would go to a special state fund and be distributed monthly. During 1988, the distributions would go entirely to Wayne County. As of 1989, 20 percent would go to the city in which the airport is located (Romulus) and the remainder to the county. A county could assign or pledge some or all of its portion to pay off obligations under the Fiscal Stabilization Act or Shared Credit Rating Act. The distribution to the city would not be made if the city is collecting taxes on airport concessions under Public Act 189 of 1953. [House Analysis, supra at 3 (emphasis added).]
The majority finds that the airport parking tax is a state tax because it is “styled as a state tax, has the structural attributes of a state tax, and serves a state purpose.” Ante at 535.1 agree with the majority that this tax “looks” like a state tax, in that it was passed by the Legislature, that it facially appears to apply to all qualified counties and municipalities, and that it is administered by a state collection entity. However, the majority has “conclude[d] that the tax clearly serves a state purpose,” id. at 537, without specifying what that “clear” state purpose is. The majority seems to suggest that the revenue is intended to support counties and cities in providing services to major airports. Id. at 539-540. If that were true, then I would agree that an airport parking tax would have a valid state purpose because the state has a strong interest in maintaining airports. However, the revenue generated by the airport parking tax does not go to servicing airports. The revenue goes directly into nonearmarked general funds.
Const 1963, art 4, § 30 requires local appropriations to be approved by two-thirds of both houses of the Legislature.