The issue before the Court is the propriety under N.J.S.A. 19:37-1 of submitting certain public questions to the voters of five counties, Bergen, Monmouth, Ocean, Passaic, and Somerset. The public questions for the proposed non-binding referendum deal with the recently adopted tax and school aid laws— they seek to determine voter support for their repeal. Necessarily affected, although not before us, are similar public questions that five other counties and numerous municipalities seek to place on the ballot. We hold that all of the proposed questions are impermissible under the statute.
The issue in this case is not whether New Jersey should have a process for non-binding referenda on state issues or whether that process would be in the public interest. Those are issues of governmental policy committed solely to the legislature. Within this scheme, our function is to decide not what is good or wise electoral process but what process the legislature has established. We recognize the strength of the counties’ interests and consequent efforts to put these questions on the ballot in order to afford their voters the chance to be heard on this issue. The law, however, does not allow what these counties seek.
The laws of this state, N.J.S.A. 19:37-1, provide a method for municipalities and counties to determine their voters’ views on proposed action within their governmental power, i.e., their *259views on local governmental issues. The method is through a public question that puts the local issue on the ballot, giving the voters the opportunity to vote yes or no. The statutory language authorizes public questions “pertaining to the government or internal affairs” of the municipality or county. While it is a non-binding referendum — the municipality or county may act contrary to the vote — it gives local government the benefit of ascertaining the extent of popular support for a particular course of action.
The laws of this state do not provide any similar method for the State, i.e., the legislature and the governor, to determine its voters’ views — all of the voters of the state — on proposed action within the State’s power, e.g., proposed State legislation. There is no general non-binding referendum law applicable to State governmental powers. Public questions designed to determine statewide voter sentiment on such issues are not authorized by law.
These two facts — the existence of the non-binding referendum law at the local level and its absence at the State level — provide the framework for this case. They lead to the proposition, derived from the explicit language of the statute and accepted by all, that the public question on the ballot must relate to a matter within the power of the local government proposing it, something over which it can act, and to the further proposition that the local public question cannot be invoked to accomplish what the law has failed to provide: a method to determine voters’ views on State questions, on matters solely within the power of State government.
We recently applied those propositions to the clear-cut issue involved in Board of Chosen Freeholders of Mercer County v. Szaferman, 117 N.J. 94, 563 A.2d 1132 (1989). There the public question sought to be placed on the county ballot to determine its voters’ views asked whether the county should pass a resolution advising the legislature to adopt and repeal various laws concerning automobile insurance. Concluding that the *260underlying subject matter — automobile insurance — was one over which the counties had no power to act, we ruled that such a question could not be submitted to the county voters under N.J.S.A. 19:37-1 for their views, and further that the fact that the county had the unquestioned power to adopt such a resolution, a resolution simply advising the legislature of the counties’ position on this State issue, did not transform the counties' lack of substantive power over insurance into an issue within the counties’ jurisdiction suitable for a public question.
Three of the public questions before us (Monmouth, Passaic, and Somerset) pose precisely the same legal issue resolved in Szaferman. They would ask the county’s voters if the county should adopt a resolution “urging the New Jersey Legislature to repeal” (or “a resolution supporting a repeal of”) the recently adopted tax increases and the new school aid law. We note immediately that their only difference with the question involved in Szaferman is that here the legislature is “urged” and there it was “advised.” In effect these questions seek county voters’ views on matters that all parties concede are committed solely to the state legislature — income and sales taxes and State aid to local school districts, matters over which the counties have no power whatsoever. The only connection of the counties to the question that is asserted is the same as that in Szaferman: the county does have the power to pass a resolution. It is of no more avail here than in Szaferman, the difference for this purpose between “urging” and “advising” being immaterial. We hold that such public questions may not be placed on the ballot for submission to the counties’ voters.
The two other proposed public questions before us are not, in our view, significantly different in terms of the propriety of submitting them to county voters. Their only difference is that instead of asking if the county should urge the legislature to repeal these laws, they ask if the county should seek their repeal through lawsuits or other means, “legislative, legal and administrative.” We hold that when the underlying issue is the *261repeal of legislation of statewide applicability, the counties’ total lack of power over the subject matter, otherwise fatal to the proposed submission of a public question to the counties’ voters, is not cured by asserting the unquestioned county power to sue, to lobby, or to take “other action to repeal such legislation.” The basis for our holding is the statute itself and the consistent decisions interpreting it to allow only those matters within the power of the county to be submitted to its voters. State taxes and school aid are not within that category. We note also that to rule otherwise could in effect transform the local non-binding referendum law so that, as in Szaferman, it could be used as a law authorizing a statewide non-binding referendum on State legislation. We deem it likely that if such public questions were permitted, all interested counties and municipalities, instead of asking their voters if they should “advise” the legislature, would ask if they should sue or lobby. As we noted in Szaferman, supra, 117 N.J. at 104, 563 A.2d 1132, in the same context (the public question involved an important statewide legislative issue): “[W]e are confident that the legislature never intended the non-binding-referendum procedure to be used to test public opinion in the abstract or to ascertain the public’s views on controversial or timely issues outside the province of the governing body soliciting them.” That prohibition against non-binding local referenda on state issues cannot be turned into permission by what may amount to nothing more than a change in the words of the proposed question, a change that leaves the main purpose unaffected: determining voter sentiment on the statewide issue. Such a device would attempt to accomplish that which the legislature has so far failed to permit.1
*262I.
In July 1990 various laws were passed increasing the State sales and income taxes. L. 1990, c. 40; L. 1990, c. 61. At about the same time the law dealing with State aid to local school districts was substantially revised, L. 1990, c. 52, providing in the future some districts with more aid than before and others, with less. The total tax increase is asserted in the various resolutions of the counties to amount to $2.8 billion per year.
Citizens in various parts of the state have campaigned for the repeal of these laws. Concurrently some school districts, perceiving a future lessening of State aid, have advocated repeal of the new school aid law, as have municipalities and counties, as well as repeal of the increase in taxes. Ranged on the other side are firm supporters of this new legislation.
In July and August of this year, ten counties passed resolutions pursuant to N.J.S.A. 19:37-1 authorizing a public question on the repeal of these laws. Eight of the resolutions, at least in *263their operative language, were substantially the same. They provided: “Shall the-County Board of Chosen Freeholders adopt a resolution condemning the disastrous effects of Governor Florio’s tax package on the county and urging the New Jersey Legislature to repeal the new school aid formula and increases in the sales tax and income tax?” Bergen County, one of the eight, thereafter repealed this resolution and substituted in its place: “Shall the Board of Chosen Freeholders take legal and other action for the purpose of changing the disastrous effects of Governor Florio’s tax package on the county, and on our public school systems throughout our county?” Ocean’s resolution, the ninth, reads as follows: “Shall the Ocean County Board of Chosen Freeholders actively pursue all available legislative, legal and administrative remedies to challenge the adoption and seek the repeal of Governor James J. Florio’s recently enacted $2.8 billion increase in State taxes?” Salem County, the tenth, adopted a resolution similar to Bergen’s: “Shall the Salem County Board of Chosen Freeholders expend county funds to take legal and other action challenging the recently enacted income and sales tax increases and school aid distribution formula enacted into law by the State Legislature.” In addition to these similarities some of the introductory clauses of the resolutions are substantially identical. Practically all of them strongly condemn the laws and seek their repeal. Several note that the county previously, without the benefit of its voters’ views on any public question, “called upon Governor Florio and the state legislature not to adopt these new taxes and school aid formula.”
We mention these similarities and the sequence of events to note that these non-binding referenda have a statewide aspect. We do not have the paradigm of Szaferman — all twenty-one counties with identical public questions seeking voters’ sentiment on a statewide issue. There are, however, ten counties with resolutions, twenty to twenty-five municipalities, and a fair degree of similarity in the resolutions. These various *264factors are indicia of the kind of referendum barred in Szaferman.
Two actions were commenced seeking to enjoin the counties (and the county clerks of other counties where certain municipalities proposed similar resolutions) from putting these questions on the ballot. They were consolidated and, after argument, decided on September 15. Judge Humphreys, in an oral opinion, ruled that pursuant to Szaferman, the questions were improper, and enjoined putting them on the ballot. Noting the apparent identity of the issues to Szaferman in view of the form of the proposed public questions — whether the counties should “urge” repeal as compared to “advising” repeal — the court concluded that resolutions asking for authority to sue were not materially different. Given the time constraints of the statute, the appeal by five of the counties (Monmouth, Passaic, Ocean, Bergen, and Somerset) was heard by the Appellate Division on September 19 and decided the same day. That court unanimously affirmed the trial court’s decision on the resolutions “urging” or “supporting” legislative repeal (Monmouth, Passaic, and Somerset), but divided on Ocean’s and Bergen's public questions. Ocean’s question seeks voter approval to “actively pursue all available legislative, legal and administrative remedies,” Bergen’s seeks approval to “take legal and other action.” The majority ruled that the decision of a county to commit its personnel and resources to such an effort was within its power and qualitatively different from its power simply to adopt a resolution urging the legislature to act. The court’s guiding principle was the settled rule that laws authorizing public questions should be liberally construed to encourage voter participation in the political process. It concluded that the Ocean and Bergen resolutions were proper. The dissenting judge disagreed, relying on the trial court’s reasoning. In addition, he pointed out that if such public questions were permitted, every State legislative issue, regardless of the lack of county power over the subject, could be submitted to the county’s voters. Based on the dissent, plain*265tiffs appealed the Ocean and Bergen decisions as a matter of right; the other three appealing counties sought certification, which we granted on September 22. Oral argument was heard on September 24. This opinion followed.
II.
The only real issue before us is whether a public question, improper because the subject matter is one committed solely to the legislature and totally outside of the counties’ power, is rendered permissible if the - action sought, instead of “advising” or “urging” legislative action, is voter approval to “take legal and other action” or to “pursue all available legislative, legal and administrative remedies.” We note the similarity between the advisory type of resolution (“shall the county adopt a resolution advising the legislature ... ”) and one that authorizes the county to lobby the legislature. Both seek to persuade the legislature, through the county's influence, to repeal the law; and while lobbying can involve the expenditure of substantial resources, it can also be as inexpensive as a resolution. The same may be said of a lawsuit. The point is that the distinction between this case and Szaferman drawn by the majority of the Appellate Division — the commitment of county resources implicit in legislative, administrative, or judicial action to effect repeal — may in fact be hardly any difference at all. Certainly on the record before us we have no way of knowing what resources, if any, would be committed if any county decided to sue or lobby. If the possibility of sending someone to Trenton were enough to validate a county public question on a State issue, we would not have denied the public question vote in Szaferman simply because only a letter might be dispatched.
Plaintiffs suggest that legally there is no material difference in a resolution urging legislative action and one authorizing a lawsuit since the latter may be viewed as one urging judicial action. In both cases, the underlying subject matter, taxes and *266school laws, is beyond the counties’ power, and in both another part of government — the State legislature or the State judiciary — is invoked to exercise its power.
We find both of these analyses — the similarity between the actions authorized, ie., “urging” as compared to suing, and the similarity in terms of seeking the action of other parts of government, ie., the legislature and the judiciary, persuasive but not dispositive. There may be lawsuits that clearly contemplate a very substantial commitment of county resources, and the fact that the county power invoked involves use of the judiciary should not of itself automatically disqualify the public question.
We base our decision instead on the nature of the issue involved, the total lack of county power over it, and the clear potential of misusing the local-issue referendum procedure by converting it into an impermissible statewide non-binding referendum. If resolutions seeking voter approval of lawsuits are permissible public questions, then the impermissible use of the local public question law — a referendum on a statewide question — would quickly be achieved, for we assume that the various governmental entities involved would revise the wording of the resolutions to call for lawsuits. Indeed, in this very ease several have already done so in the days following the Appellate Division’s decision. And Bergen did so before then. Thus, there would be no realistic confinement of the local public question law to the purpose intended by the legislature, ie., providing a process to obtain an expression of voters’ view on local issues. All important State legislative issues would be potentially the subject of referenda, limited only by the number of counties and municipalities that wish to cooperate. That potential is clear from the automobile insurance case in which all twenty-one counties joined. It is also clear here. We do not suggest that counties would feign a wish to learn voter sentiment concerning lawsuits, legislative action, and administrative action. All we suggest is that the potential for misuse is patent. The motive for placing a public question of this kind on *267the ballot is so strong and the possibility of a -lawsuit or lobbying so arguably realistic that many counties would use this form of question.2
To the extent that we permit public questions to indirectly seek what is truly sought — the voters’ view on what the legislature should do — the indirection involved in those public questions, concerning lawsuits, lobbying, administrative action, “and other action,” and the like, may not clearly indicate what the voters’ views are. The law concerning all public questions, N.J.S.A. 19:3-6, requires that they “shall be presented in simple language that can be easily understood by the voter,” and that they “shall clearly set forth the true purpose of the matter being voted upon.” We have no doubt of the counties’ good faith in framing their questions. The imprecision of some of the questions is undoubtedly the product of many factors, but not the least of them is the indirection forced on counties seeking a referendum on repeal of a state tax but required by law to state the question in terms of matters over which the counties have power.3
*268Whatever their form, the fact that these statewide questions may be posed only in a limited part of the state, here ten counties, plus some municipalities, including some within those counties posing the same question, erodes the good these questions might achieve. Altogether, the lack of a direct route to achieving what many think is a desirable goal — a statewide non-binding referendum — may, if the indirect route is allowed, lead not to its benefits but to an unpredictable patchwork of potentially misleading, imprecise questions giving mixed signals about the voters in counties that pose the question and telling nothing about the voters in counties that do not, here a substantial portion of the state’s voters. In these very proceedings, it was entirely possible that the only voters who would be given the opportunity to express their views would be those from two counties, or perhaps from five, or at most from ten, an anomalous result for a referendum on a statewide issue. These deficiencies are most apparent when compared to what many seek but what the law does not permit — a true non-binding statewide referendum on an important state issue, a referendum in which all of the voters of the state can vote on the same question, clearly, precisely, and directly stated.
III.
The counties argue most forcefully that which is their true concern: that these laws may severely impact on them, on their citizens, on their ability to deal with governmental problems within their power; that these laws may badly hurt them. Certainly the counties’ effectively available local tax base may be affected if their voters are subjected to increased State taxes. And certainly those municipalities (more than the counties) whose school districts’ tax levy may have to be substantially increased may be so affected.
Whatever may be the answer to these concerns, the fact remains that the statute authorizes local public questions only on matters over which the local units have power, not on *269impacts over which they are without power. The legislature has not authorized county and municipal public questions concerning state legislation based on its impact on county or municipal functions, no matter how severe. As in Szaferman, where we assumed there would be some impact, there an increase in direct county expenditures for insurance, the impact, whether minimal or “disastrous,” does not make the legislation a subject matter over which the county has power to act, the sine qua non of a permissible public question. That the municipalities or counties may have the strongest reason to want to make their voters’ views known when the impact of the legislation is perceived to be severe is legally irrelevant given the clear language of the public question statute. The legislature confined that law to public issues about which counties have power to act; it.never extended it to issues about which the counties may have reason to complain.
IV.
Finally, the counties assert that their proposed public questions fit within certain categories defined in case law that render them permissible. The categories are questions involving powers that, while belonging to others, require municipal or county action for their implementation or that allow for cooperative governmental action between the county (or municipality) and some other unit of government. The cases relied on are Gamrin v. Mayor and Council of Englewood, 76 N.J.Super. 555, 185 A.2d 55 (Law Div.1962), and Rowson v. Township Committee of Mantua, 171 N.J.Super. 129, 408 A.2d 137 (App.Div.1979).
In Gamrin the question proposed by the municipality was whether to approve the transfer of students from various elementary schools within the district to one school, a move that apparently would entail considerable expense. The underlying issue was committed solely to the Board of Education, .the municipality having no power to decide whether such a transfer *270should be made. But the decision to transfer the students would be a nullity if the Board of School Estimate, including by statute three members — and thereby the controlling vote — of the municipal governing body, refused to provide the needed funds. Viewing the Board of School Estimate as, in effect, the alter ego of the municipality, the trial court concluded that the municipal power, in effect a veto power, over the question was sufficient to warrant submission of the public question to its voters. With the issue pending before the municipally-controlled Board of School Estimate, the municipality sought its voters’ views on the transfer. The court apparently deemed the question as the same as asking the voters if the money should be spent for that purpose, a subject over which the municipality — through the Board of School Estimate — clearly had power. This kind of “implementation” power is in no way involved in this case. The tax increase is beyond the counties’ power to implement or veto, as is the new school aid formula. The fact that local school districts, or in the case of county vocational education districts or county colleges through county boards of school estimate, can affect the quality of education by their actions, including the amount of local money they raise, is not the same as having a veto over the school aid law that would bring it within the county’s power. In Gamrin, based on the vote on the public question, the governing body could effectively control the underlying subject matter, the decision whether to transfer the students. That is not the case here. All the governing body of the county can do is try, within the limits of the very law it seeks to repeal, to affect its implementation. The direct question, whether to repeal, is one over which the county has no power whatsoever. In Gamrin, the direct question, whether to move the students, was one over which the municipality, through the Board of School Estimate, had total power.
Rowson involved a public question in which both the municipality and the utilities authority had statutory power over the subject matter involved, although, as in Gamrin, not over the *271specific issue stated in the public question. The question was whether the utilities authority should acquire certain additional facilities and make other expenditures, substantial ones, needed to use those additional facilities. While the municipality had no power to approve or veto that decision, the statute gave it substantial power to aid, or not to aid, the utilities authority in the venture, and especially to aid it financially. In a prior undertaking, the municipality had guaranteed more than $500,000 in authority notes pursuant to this power. The question, although not expressed in those terms, was interpreted by the court as designed to aid the municipality in exercising its unquestioned substantial financial powers over the subject.
We need not now review the correctness of Gamrin and Rowson. The considerations involved when a truly local public question is at issue may be different from those in the case before us. We conclude only that whatever “implementation” and “cooperation” may be implicit in the state school aid law provide no justification for the submission of a public question to repeal it.
To the extent that Ocean County contends its question involves the counties’ power to implement the State school aid law, or to cooperate with the State in that implementation, we note that its resolution nowhere mentions the school aid law but is limited solely to repealing the increase in State taxes. To the extent that Bergen County relies on the same justification for its public question, presumably referring to the counties’ only substantive power in that area (the county educational institutions, e.g., the county vocational school and the county college, through the county board of school estimate), we note that its resolution refers to the “public school systems” throughout the county. More than that, any claim by the counties that their public questions can be sustained as based on the need to know whether and to what extent to implement the school aid law or to cooperate in that implementation is unpersuasive. The validity of the public question cannot depend on some theoretical *272justification for asking it. It must rest on something that not only the county later on, but the voter at the time of voting, regarded as germane to his or her vote, some proposed action on which the voter reasonably thought he or she was expressing a view. It is simply not reasonable to assume that any voter, when asked whether the school aid law should be repealed, or whether suit should be brought for that purpose, would think his views were being sought on how to implement or cooperate with that law. Nor is it reasonable to assume that the counties had that in mind when they posed these public questions. Whatever indulgence we might give to these kinds of contentions in the context of a purely local issue, in the spirit of liberally construing this election law, is not appropriate where the question clearly concerns a state issue, beyond the county’s power, and therefore beyond what the legislature intended in the local public question law.
Of all the previous cases, Botkin v. The Mayor and Borough Council of Westwood, 52 N.J.Super. 416, 145 A.2d 618 (App.Div.), appeal dismissed, 28 N.J. 218, 146 A.2d 121 (1958), best provides the crucial distinction between this case, involving a statewide issue, and those where local issues are involved. In Botkin the defendant municipality wanted to “deconsolidate” a consolidated school district, one made up of two formerly independent school districts, one of which had been the defendant’s school district. The problem was that there was no law authorizing deconsolidation under any circumstances. The municipality expended great effort to investigate how to accomplish deconsolidation, in effect whether and how to obtain legislation authorizing it. Having concluded that such legislation was essential and that it wanted such legislation, it sought voter approval to go. ahead, to some extent because it had reason to believe that legislative approval depended on a showing of voter support for deconsolidation, and also because, in good faith, it was not itself ready to make the commitment to seek legislation without such support. At the same time — and critical to the resolution of the case — a borough council election *273was in process in which “deconsolidation” had become a potent issue. In that context, the municipality requested a public question vote on whether it should seek legislation authorizing deconsolidation; or so all sides viewed the meaning of the question. 52 N.J.Super. at 436, 145 A.2d 618.
The Appellate Division majority held that the question was impermissible under the statute. Justice (then judge) Hall, writing for the majority, gave two reasons for the result: first, that the municipality had no power even to seek legislation concerning this subject, and, second, and related, that to place that question on' the ballot would inevitably and improperly inject municipal polities, through the vehicle of the forthcoming election, into the affairs of the consolidated board of education. The dissent concluded that the municipality had the power and sufficient interest to warrant use of N.J.S.A. 19:37-1.
The case is properly read not as one in which there was a division on the principle involved — the majority and dissent both agreeing that the public question could not be put to the voters unless the municipality had the power to seek legislation — but rather on the question of whether the power to lobby existed. See Santoro v. South Plainfield, 57 N.J.Super. 498, 501-02, 155 A.2d 23 (App.Div.1959). Both Botkin and Santoro suggest that in some circumstances involving clearly identified local issues, the municipality, if it has the power to lobby the legislature, may use the public question method to determine whether or not to exercise that power.
We note the obvious and critical difference with this case. There were no general statewide issues involved in Botkin, and there was no potential to use the local public question method as a device to obtain a non-binding referendum on a statewide issue. Instead, there was a very limited local issue, of interest apparently to but one municipality in the state, the only action available to the municipality was lobbying, considerable resources had already been expended in connection with that purpose, and the necessity for lobbying was undeniable. Given *274those circumstances the dissent’s position in Botkin is understandable, indeed, it persuaded three members of this Court of its correctness. The case is an example of what we noted in Szaferman, supra, 117 N.J. at 104, 563 A.2d 1132: “Extended to its logical limits, this principle may generate close questions concerning the statute’s application in specific instances.”
Botkin, Rowson, and Gamrin, when read with Szaferman, demonstrate the liberality accorded the local public question statute when truly local interests are involved and the restraints that apply when it is attempted to be used to encompass state interests. Cf. In re Certain Petitions for Binding Referendum, 154 N.J.Super. 482, 484, 381 A.2d 1217 (App.Div.1977) (general liberal construction of referendum laws yields to contrary presumed legislative intent).
Y.
Various other contentions were raised below, e.g., that venue in Essex County was improper, that municipalities with similar public questions were indispensable parties whose nonjoinder was fatal, that joinder of the counties was improper, that in any event plaintiffs had no standing to bring the lawsuit, and that public questions without an interpretive statement were defective. We see no need to rule on these issues. The matter is of great public interest and should be disposed of on the merits.
VI.
We reaffirm the principle of Szaferman that the local public question law may not be used to seek voter sentiment on issues beyond the competence or jurisdiction of the local unit, here, as in Szaferman, legislative issues of statewide importance. While the statute should generally be liberally construed, it may not be used beyond its intended purpose. We express no opinion on the propriety of the counties’ attempt to place the issue on the ballot or on the merits of the laws involved.
*275The judgment of the Appellate Division allowing the resolutions of Bergen and Ocean Counties to be placed on the ballot is reversed; otherwise that judgment is affirmed. No public question substantially similar to those dealt with in this opinion may legally be placed on the ballot.4
For affirmance in part, reversal in part — Chief Justice WILENTZ, C.J., and Justices CLIFFORD, HANDLER, POLLOCK, O’HERN, GARIBALDI and STEIN — 7.
Opposed — None.