Abbott v. BurkeAnnotate this Case
119 N.J. 287 (1990)
575 A.2d 359
RAYMOND ARTHUR ABBOTT, A MINOR, BY HIS GUARDIAN AD LITEM, FRANCES ABBOTT; ARLENE FIGUEROA, FRANCES FIGUEROA, HECTOR FIGUEROA, ORLANDO FIGUEROA, AND VIVIAN FIGUEROA, MINORS, BY THEIR GUARDIAN AD LITEM, BLANCA FIGUEROA; MICHAEL HADLEY, A MINOR, BY HIS GUARDIAN AD LITEM, LOLA MOORE; HENRY STEVENS, JR., A MINOR, BY HIS GUARDIAN AD LITEM, HENRY STEVENS, SR.; CAROLINE JAMES AND JERMAINE JAMES, MINORS, BY THEIR GUARDIAN AD LITEM, MATTIE JAMES; DORIAN WAITERS AND KHUDAYJA WAITERS, MINORS, BY THEIR GUARDIAN AD LITEM, LYNN WAITERS; CHRISTINA KNOWLES, DANIEL KNOWLES, AND GUY KNOWLES, JR., MINORS, BY THEIR GUARDIAN AD LITEM, GUY KNOWLES, SR.; LIANA DIAZ, A MINOR, BY HER GUARDIAN AD LITEM, LUCILA DIAZ; AISHA HARGROVE AND ZAKIA HARGROVE, MINORS, BY THEIR GUARDIAN AD LITEM, PATRICIA WATSON; AND LAMAR STEPHENS AND LESLIE STEPHENS, MINORS, BY THEIR GUARDIAN AD LITEM, EDDIE STEPHENS, PLAINTIFFS-APPELLANTS, v. FRED G. BURKE, COMMISSIONER OF EDUCATION; EDWARD G. HOFGESANG, NEW JERSEY DIRECTOR OF BUDGET AND ACCOUNTING; CLIFFORD A. GOLDMAN, NEW JERSEY STATE TREASURER; AND NEW JERSEY STATE BOARD OF EDUCATION, DEFENDANTS-RESPONDENTS.
The Supreme Court of New Jersey.
Argued September 25, 1989.
Decided June 5, 1990.
*292 Marilyn J. Morheuser argued the cause for appellants (Marilyn J. Morheuser, attorney; H. Kit Ellenbogen, David C. Long, a member of the District of Columbia bar, Elaine M. Song, and Paul L. Tractenberg, of counsel and on the briefs).
Alfred E. Ramey, Jr., Deputy Attorney General, argued the cause for respondents (Peter N. Perretti, Jr., Attorney General of New Jersey, attorney; Alfred E. Ramey, Jr., Michael R. Clancy, Assistant Attorney General, and Bertram P. Goltz, Jr., Deputy Attorney General, of counsel; Alfred E. Ramey, Jr., Bertram P. Goltz, Jr., David Earle Powers, and E. Phillip Isaac, Deputy Attorneys General, on the briefs).
Melville D. Miller, Jr., argued the cause for amicus curiae Legal Services of New Jersey, Inc.
*293 Stephen Eisdorfer, Assistant Deputy Public Advocate, argued the cause for amicus curiae Public Advocate (Alfred A. Slocum, Public Advocate, attorney; Stephen Eisdorfer, Shirley Brandman, John V. Jacobi, Susan R. Oxford, and Clifford Gregory Stewart, Assistant Deputy Public Advocates, and Richard E. Shapiro, Director, Division of Public Interest, on the brief).
Jane M. Hanson and John M. Payne submitted a brief on behalf of amicus curiae American Civil Liberties Union of New Jersey.
Robert T. Pickett submitted a brief on behalf of amicus curiae New Jersey Black Issues Convention, Inc. (Robert T. Pickett, attorney; Robert T. Pickett and Charles D. Craig, of counsel and on the brief).
Cecilia M. Zalkind submitted a brief on behalf of amicus curiae Association for Children of New Jersey.
Richard A. Friedman submitted a brief on behalf of amicus curiae New Jersey Education Association (Zazzali, Zazzali, Fagella & Nowak, attorneys).
Richard M. Altman and Anne P. McHugh submitted a brief on behalf of amicus curiae The American Jewish Committee (Pellettieri, Rabstein and Altman, attorneys).
Douglas S. Eakeley submitted a brief on behalf of amicus curiae The League of Women Voters of New Jersey (Riker, Danzig, Scherer & Hyland, attorneys; Douglas S. Eakeley, Joyce A. Howell, and John J. Farmer, Jr., on the brief).
Peter A. Buchsbaum submitted a brief on behalf of amici curiae Metropolitan Ecumenical Ministry, New Jersey Council of Churches, The Citizens Institute for Involvement in Education, and The New Jersey Synod-Evangelical Lutheran Church in America (Hannoch, Weisman, attorneys; Peter A. Buchsbaum and Louise L. Stanton, on the brief).
Beverly A. Williams submitted a brief on behalf of amici curiae The New Jersey State Conference of NAACP Branches, The New Jersey Alliance of Black School Educators, and the *294 New Jersey Association of Black Educators (Epstein, Becker & Green, attorneys; M. Elaine Jacoby, of counsel).
Sidney H. Lehmann and David B. Beckett submitted a brief on behalf of amicus curiae Newark Teachers' Union, Local 481, AFT/AFL-CIO (Szaferman, Lakind, Blumstein, Watter & Blader, attorneys).
Francis J. Campbell, Acting General Counsel, submitted a brief on behalf of amicus curiae New Jersey School Boards Association.
TABLE OF CONTENTS I. Description of the Issues .................................. 300 II. The Constitutional Provision ............................... 303 III. Summary of the Issues ...................................... 316 IV. Facts and Conclusions ...................................... 322 A. The Funding Scheme ...................................... 324 B. Educational Funding Disparities ......................... 334 C. Substantive Educational Opportunity: The Administration of the Act by the Commissioner and the Board ......................................... 347 1. Municipal Overburden ................................. 355 D. The Quality of Education in the Poorer Urban Districts ............................................. 357 E. The Quality of Students' Needs in the Poorer Urban Districts ....................................... 369 F. Impact of the Level of Funding on the Quality of Education .......................................... 375 V. Findings ................................................... 384 VI. Remedy ..................................................... 385 VII. Conclusion ................................................. 391
The opinion of the Court was delivered by WILENTZ, C.J.
We again face the question of the constitutionality of our school system. We are asked in this case to rule that the *295 Public School Education Act of 1975, L. 1975, c. 212, N.J.S.A. 18A:7A-1 to -52 (the Act) violates our Constitution's thorough and efficient clause. We find that under the present system the evidence compels but one conclusion: the poorer the district and the greater its need, the less the money available, and the worse the education. That system is neither thorough nor efficient. We hold the Act unconstitutional as applied to poorer urban school districts. Education has failed there, for both the students and the State. We hold that the Act must be amended to assure funding of education in poorer urban districts at the level of property-rich districts; that such funding cannot be allowed to depend on the ability of local school districts to tax; that such funding must be guaranteed and mandated by the State; and that the level of funding must also be adequate to provide for the special educational needs of these poorer urban districts in order to redress their extreme disadvantages.
We note the convincing proofs in this record that funding alone will not achieve the constitutional mandate of an equal education in these poorer urban districts; that without educational reform, the money may accomplish nothing; and that in these districts, substantial, far-reaching change in education is absolutely essential to success. The proofs compellingly demonstrate that the traditional and prevailing educational programs in these poorer urban schools were not designed to meet and are not sufficiently addressing the pervasive array of problems that inhibit the education of poorer urban children. Unless a new approach is taken, these schools even if adequately funded will not provide a thorough and efficient education.
We reject the argument, however, that funding should not be supplied because it may be mismanaged and wasted. Money *296 can make a difference if effectively used, it can provide the students with an equal educational opportunity, a chance to succeed. They are entitled to that chance, constitutionally entitled. They have the right to the same educational opportunity that money buys for others.
On this record we find a constitutional deficiency only in the poorer urban districts, and our remedy is limited to those districts. We leave unaffected the disparity in substantive education and funding found in other districts throughout the state, although that disparity too may some day become a matter of constitutional dimension. We do so without implying in any way that such disparity is not important when considered as a matter of policy. Our decision deals not with optimum educational policy but with constitutional compliance.
At various points in this opinion, we refer to the administration of the Act by the State Board of Education (Board) and the Commissioner of Education (Commissioner). The record is replete with evidence of their dedication, industriousness, perseverance and, ultimately, their considerable accomplishments. The problems they face have bedeviled the entire nation. No one has solved them. Our constitutional conclusion in no way belittles their prodigious efforts and their many achievements.
This litigation is described in Abbott v. Burke, 100 N.J. 269, 495 A.2d 376 (1985) (Abbott I), in which we decided it should be contested before the administrative agency rather than the courts. Plaintiffs are school children from Camden, East Orange, Jersey City, and Irvington. Claiming that the Act violates the constitutional provision, they commenced suit in the Superior Court for a declaration of its unconstitutionality and other appropriate relief. The trial court, conceiving of the action as raising a dispute or controversy under the education laws of the state, ruled that the issue should be determined by the administrative agency with jurisdiction over such controversies the Department of Education (DOE) and therefore dismissed the complaint for failure to exhaust the administrative *297 remedy. On appeal the Appellate Division reversed, 195 N.J. Super. 59, 477 A.2d 1278 (1984), holding that since the case required adjudication of a constitutional issue, "beyond the power of the Commissioner to decide", id. at 74, 477 A.2d 1278, jurisdiction in the Superior Court was required. It found the "doctrine of exhaustion of administrative remedies ... inapplicable." Ibid. (citation omitted). We granted certification, 97 N.J. 669, 483 A.2d 187 (1984). Finding that the "considerations ... relevant to the exhaustion requirement [were] in near-equipoise," Abbott I, supra, 100 N.J. at 298, 495 A.2d 376, we decided on an administrative determination in order to develop a record adequate for the complex issues involved, a record informed by the presumed expertise of the Administrative Law Judge (ALJ), the Commissioner, and the Board. Noting that the Commissioner was a defendant, we indicated that the initial hearing and fact-finding should be before an ALJ. Id. at 302, 495 A.2d 376.
After extensive hearings and other proceedings spanning a period of over eight months, the ALJ found that evidence of substantial disparities in educational input (such as course offerings, teacher staffing, and per pupil expeditures) were related to disparities in school district wealth; that the plaintiffs' districts, and others, were not providing the constitutionally mandated thorough and efficient education; that the inequality of educational opportunity statewide itself constituted a denial of a thorough and efficient education; that the failure was systemic; and that the statute and its funding were unconstitutional. Recognizing limitations on his authority, the ALJ, while declining to rule on remedies, nevertheless recommended various options, including a "high foundation" program of funding. Implicit in the ALJ's view of the thorough and *298 efficient clause was a constitutional requirement of substantial equality of educational opportunity throughout the State.
The Commissioner declined to accept the ALJ's recommendations. He rejected the ALJ's factual finding of a strong relationship between property wealth and per pupil expenditures, or between either of those factors and certain indicators commonly thought of as related to educational quality staffing ratios, teacher experience and training, and the like. The Commissioner noted the inconsistencies in the relationship between per pupil expenditures and property wealth, some property-poor districts spending more than richer ones; and the lack of consistency in the relationship between per pupil expenditure and, e.g., staffing ratios. He noted that plaintiffs' most thorough analysis was limited largely to comparisons between the poorest and the richest districts, leaving, in his view, a very substantial gap in between, where the record was insufficient to prove any of the relationships claimed. He found more persuasive the characterization of the data by certain experts as demonstrating an idiosyncratic pattern rather than one of consistent relationships. He further rejected the ALJ's findings of substantially greater breadth of course offerings in the more affluent districts, noting that the evidence was largely anecdotal and, in any event, was limited to subjects not critically related to a thorough and efficient education.
The Commissioner's most basic disagreement with the ALJ was in his evaluation of the record concerning the relationship of educational expenditures to the quality of education offered, the educational opportunity offered; and the "production-function" question, whether input was related to output. The *299 Commissioner ruled that plaintiffs had failed to prove that such relationships existed, that most studies showed the relationship could not be established; he characterized the ALJ's belief that greater funding was needed to assure thorough and efficient education across the state as enthroning "naivete."
The Commissioner concluded, as a legal matter, that our Constitution did not require equal expenditures per pupil but rather required a minimum substantive level of education as defined in the Act and the rules and regulations of the Board and the Commissioner; that this Court had ruled that this statutory definition of thorough and efficient was constitutional; and that the procedure put in place by the Board and administered by the Commissioner was sufficient to assure, and indeed had already largely achieved, a thorough and efficient education throughout the state. He concluded that far from a systemic constitutional failure, the Act guaranteed a thorough and efficient education by virtue of the school districts' unlimited power to raise funds to satisfy their constitutional obligation, the Commissioner's power to require them to do so, and the Commissioner's power to take over the operation of any district that fails. The sufficiency of the educational opportunity now in place and to be achieved in the future was assured, according to the Commissioner, not through money but through the reporting, monitoring, and corrective provisions of the Act, rules, and regulations. Furthermore, the Commissioner found that if there was indeed any failure of a district to achieve the constitutional standard, the remedy was in the Act and in its *300 enforcement. Indeed, the Commissioner noted several districts that were not achieving the level of education expected and the steps being taken, under the rules and regulations, to compel improvement and compliance. The Commissioner concluded that any failure to provide the constitutional standard and he conceded none was district-specific and remediable under the existing educational funding system.
On review, the Board adopted the Commissioner's decision in almost all respects. While recognizing the severe problems faced by disadvantaged children, the Board observed that the school system cannot solve all the problems of our society, and conceded that three of the plaintiffs' districts (Jersey City, East Orange, and Camden) are not affording a thorough and efficient education today, but agreed with the Commissioner that a constitutionally sufficient system was in place, and that through its funding and administration the Act assured a thorough and efficient education. As did the Commissioner, the Board noted certain deficiencies in the present funding statute (mainly the provision of equalization aid based not on the district's present need but on the prior year's budget) and its inability to address capital construction needs. It recommended corrective legislation. The Board also ordered that rules and regulations be adopted to strengthen the reporting, monitoring, and corrective functions. The ultimate conclusion of the Board, however, was that even without new regulations, the Act as implemented was constitutional as applied throughout the entire state.
We certified plaintiffs' appeal to the Appellate Division. 117 N.J. 51, 563 A.2d 818 (1989).I. Description of the Issue
Predictably flowing from our decision in Robinson v. Cahill, 69 N.J. 449, 355 A.2d 129 (1976) (Robinson V), the issue now before us is whether the Act, declared facially constitutional, is *301 constitutional as applied. Despite that declaration, we recognized the possibility that in fact some districts might not provide a thorough and efficient education. Indeed, three of the Court's members expressed serious reservations on that question two of them dissenting.
The positions of the parties suggest a delineation of that issue: the plaintiffs contend that the Act as applied is systemically productive of such financial and educational disparities as to render it unconstitutional in toto. A potential subsidiary issue, if that contention is rejected, is whether the Act as applied is unconstitutional for specific districts or for a specific class of districts. If either contention is accepted, plaintiffs' claims concerning remedies raise further issues. The State's position raises the issue of whether the Act, through its definition of a thorough and efficient education, its statement of goals and guidelines, its requirement of subsidiary goals at the district level, and its enforcement through the combined powers of the Board, Commissioner, and the district, has in fact resulted in the constitutional standard of education throughout the state. If not, the subsidiary question is whether the failure, if any, is systemic, requiring a declaration of unconstitutionality or is district-specific, requiring corrective action under the Act in a limited number of failing districts. Another statement of the issue is whether plaintiffs' demonstration of substantial disparity in expenditures and educational input between the poorest and richest districts compels the conclusion that the Act is unconstitutional; or whether, as the State contends, the actual substantive education provided to the students in the poorest districts is thorough and efficient or subject to being readily corrected; whether, as the State contends, thorough and efficient education under the Constitution requires a minimum level of education, and not equality.
If plaintiffs are correct in their contention that the Constitution requires substantial equality in educational funding, one of several radical changes must result in addition to the change from the inequality that has always existed to the equality that *302 plaintiffs say is mandated. The State would, through some means, have to assure make certain that practically all districts spend the same amount per pupil. If that figure were to be pegged at the level of the more affluent suburban districts, it would require a massive infusion of public funds. If, on the other hand, the equality level were to be pegged somewhere near the average, and if strict limits were to be placed on any district's ability to exceed that amount in spending, a significant number of suburban districts would be compelled to substantially decrease their educational expenditures, in effect, to diminish the quality of education now provided to their students. The implications of the former a vast infusion of funds are particularly pointed in view of the fact that New Jersey is already one of the highest spending states in the nation in terms of per pupil expenditures. The implications of the latter a uniform funding standard at the average rate requiring a leveling down would be unusual in a state that has regarded home rule in the area of education, including gross disparity in expenditures between the rich and the poor, as an accepted part of the system.
The implications of the State's position are similarly unsettling. The inadequacy of poorer urban students' present education measured against their needs is glaring. Whatever the cause, these school districts are failing abysmally, dramatically, and tragically. Poorer students need a special supportive educational effort in order to give them a chance to succeed as citizens and workers. Their educational needs are often dramatically different from those of students in affluent districts. *303 They are getting the least education for the greatest need. The implications of that fact on their future and on the state's future is the central theme of plaintiffs' case.II. The Constitutional Provision
In order to pass on plaintiffs' contention, we must once again, in the context of this case, define the scope and content of the constitutional provision. That definition is critical to our determination of a remedy. While precision in such definition is desirable, certain considerations suggest caution against constitutional absolutism in this area. First, what a thorough and efficient education consists of is a continually changing concept. As the Legislature stated:Because the sufficiency of education is a growing and evolving concept, the definition of a thorough and efficient system of education and the delineation of all the factors necessary to be included therein, depend upon the economic, historical, social and cultural context in which that education is delivered. The Legislature must, nevertheless, make explicit provision for the design of State and local systems by which such education is delivered, and should, therefore, explicitly provide after 4 years from the effective date of this act for a major and comprehensive evaluation of both the State and local systems, and the sufficiency of education provided thereby.... [N.J.S.A. 18A:7A-2a(4).]
We observed in Robinson V that "[t]his statement reveals a perceptive recognition on the part of the Legislature of the constantly evolving nature of the concept being considered. It manifests an awareness that what seems sufficient today may be proved inadequate tomorrow, and even more importantly that only in the light of experience can one ever come to know whether a particular program is achieving the desired end." Robinson V, supra, 69 N.J. at 457-58, 355 A.2d 129. Second, whatever the content of a thorough and efficient education may be, the question of what must be done to achieve it is debatable, as this case well illustrates. Third, embedded in the constitutional provision itself, at least in its construction thus far by this Court, are various objectives and permissible outcomes equality, uniformity, diversity, and disparity that may require, *304 if they are to be allowed, a continued general definition of the constitutional mandate.
Finally, any definition of the constitutional obligation must operate in an area where confrontation between the branches of government is not only a distinct possibility but has been an unfortunate reality. See, e.g., Robinson v. Cahill, 69 N.J. 133, 351 A.2d 713 (1975) (Robinson IV); Robinson V, supra, 69 N.J. at 468, 355 A.2d 129; and Robinson v. Cahill, 70 N.J. 155, 358 A.2d 457 (1976) (Robinson VI). That potential confrontation concerns one of the most important functions of government education and involves substantial public funds, implicates the taxing power, and is potentially of a continuing nature. The Legislature's role in education is fundamental and primary; this Court's function is limited strictly to constitutional review. The definition of the constitutional provision by this Court, therefore, must allow the fullest scope to the exercise of the Legislature's legitimate power.
The initial construction of the thorough and efficient clause was permeated by the concept of equality. In Landis v. Ashworth (School District No. 44), 57 N.J.L. 509, 31 A. 1017 (Sup.Ct. 1895), the court construed the constitutional provision (effected by the 1875 amendment to our former Constitution of 1844) as requiring "equality within the intended range of that amendment...." Robinson v. Cahill, 62 N.J. 473, 514, 303 A.2d 273 (1973), cert. denied sub nom. Dickey v. Robinson, 414 U.S. 976, 94 S. Ct. 292, 38 L. Ed. 2d 219 (1973) (Robinson I). The qualifying language that we added, "permitting local decisions only above and beyond that mandated education," ibid., reflected our understanding of Landis as allowing local districts the option of operating high schools, at that time a level of education not generally available. As Landis noted, if absolute equality were mandated by the thorough and efficient clause, it would mean either that all districts must have high schools or that none could "[n]either of these consequences... contemplated by the amendment of 1875." Robinson I, supra, 62 N.J. at 514, 303 A.2d 273 (quoting Landis, supra, 57 *305 N.J.L. at 512, 31 A. 1017). Implicit in our discussion was that education below the level of high school "the intended range [of the] amendment" must be equal in all districts. We noted that without doubt, today high schools were very much a part of "the rights of all," Robinson I, supra, 62 N.J. at 514, 303 A.2d 273 (quoting Landis, supra, 57 N.J.L. at 512, 31 A. 1017), and that "a system of public education which did not offer high school education would hardly be thorough and efficient." Id. 62 N.J. at 515, 303 A.2d 273. That requirement of equality of educational opportunity through high school appears again in our observation, after our rejection of the claim that the constitutional provision required taxpayer equality, that "we do not doubt that an equal educational opportunity for children was precisely in mind." Id. at 513, 303 A.2d 273.
What that equality meant, while not precisely defined, was indicated in several ways. First, in deciding that the statute then in place was unconstitutional as not affording a thorough and efficient education, we relied solely on the disparity of funding, i.e., on the fact that the dollars spent on education per pupil varied from one district to another (from below $700 per pupil to over $1,500 per pupil, Robinson V, supra, 69 N.J. at 480 n. 4, 355 A.2d 129). As we stated:The trial court found the constitutional demand had not been met and did so on the basis of discrepancies in dollar input per pupil. We agree. We deal with the problem in those terms because dollar input is plainly relevant and because we have been shown no other viable criterion for measuring compliance with the constitutional mandate. The constitutional mandate could not be said to be satisfied unless we were to suppose the unlikely proposition that the lowest level of dollar performance happens to coincide with the constitutional mandate and that all efforts beyond the lowest level are attributable to local decisions to do more than the State was obliged to do. [Robinson I, supra, 62 N.J. at 515-16, 303 A.2d 273.]
Rather than on equality, our decision was based on the proposition that the Constitution required a certain level of education, that which equates with thorough and efficient; it is that level that all must attain; that is the only equality required by the Constitution. Embedded in our observation that if the lowest level of expenditures per pupil constituted a thorough and *306 efficient education, then the constitutional mandate would be met was the clear implication that no matter how many districts were spending well beyond that level, the system would be constitutional. Second, we noted that the State, while assigning the obligation to local government to afford a thorough and efficient education, had never defined "in some discernible way the educational obligation," "the content of the constitutionally mandated educational opportunity"; it was "an unstated standard." Id. at 519, 303 A.2d 273. Again, the clear import is not of a constitutional mandate governing expenditures per pupil, equal or otherwise, but a requirement of a specific substantive level of education. Equality of expenditures per pupil could not have been constitutionally mandated when we recognized the right of districts to spend more to address students' special needs (the "need for additional dollar input to equip classes of disadvantaged children for the educational opportunity") and disclaimed any intent to deprive the State of the power to "authorize local government to go further" than "the constitutionally mandated education" and "to tax to that further end." Id. at 520, 303 A.2d 273. Our only condition was that such excess "not become a device for diluting the State's mandated responsibility." Ibid.
Our decision in Robinson I was necessarily general because of the narrow record in that case, consisting primarily of dollar per pupil information and related socioeconomic data. Although general, however, our holding in Robinson I was clear and formed the basis for our holding in Robinson V: a thorough and efficient education requires a certain level of educational opportunity, a minimum level, that will equip the student to become "a citizen and ... a competitor in the labor market." Robinson I, supra, 62 N.J. at 515, 303 A.2d 273. The State's obligation to attain that minimum is absolute any district that fails must be compelled to comply. If, however, that level is reached, the constitutional mandate is fully satisfied regardless of the fact that some districts may exceed it. In other words, the Constitution does not mandate equal expenditures *307 per pupil. We implied that the level can and should be defined in terms of substantive educational content. But while disparity was explicitly permitted, there was a caveat the excess spending could not somehow be allowed to mask a failure to achieve thoroughness and efficiency in other districts.
This holding in Robinson I was reaffirmed in Robinson IV, supra, 69 N.J. 133, 351 A.2d 713. After the Legislature failed to act within the time limits set by the Court, we afforded a "contingent or provisional remedy," Robinson IV, supra, 69 N.J. at 146, 351 A.2d 713, that substantially increased equalization aid. Id. at 150, 351 A.2d 713. In the course of our opinion we referred to our statement of the constitutional command in Robinson I that the State afford "an equal educational opportunity for children", id. at 140, 351 A.2d 713 (citation omitted), and shortly thereafter acknowledged "the legitimacy of permitting any school district wishing to do so to spend more on its educational program through local effort" provided, again in the words of Robinson I, such did not become "a device for diluting the State's mandated responsibility." Id. at 141 n. 3, 351 A.2d 713. Foreshadowing what was to come, we also observed that while disparity in expenditures per student was the sole criterion in our decision in Robinson I, that was because "we [had] been shown no other viable criterion for measuring compliance with the constitutional mandate." Id. at 141, 351 A.2d 713 (citation omitted). We noted that in addition to the Robinson I record, the Court now had further material showing that "a multitude of other factors play a vital role in the educational result," and that therefore "while funding is an undeniable pragmatic consideration, it is not the overriding answer to the educational problem, whatever the constitutional solution ultimately required." Id. at 141 n. 3, 351 A.2d 713.
The Legislature acted in response to Robinson IV. In addition to defining and providing for the achievement of a thorough and efficient education through administrative measures, it provided a new funding mechanism to finance the substantive education defined in the Act as constituting "thorough and *308 efficient." It firmly placed responsibility on the State to assure achievement of the thorough and efficient level in every district. It did so, however, through a scheme that continued to allow disparity in both dollars per pupil and educational content. Indeed, while the statute was sustained as facially constitutional, the doubts and qualifications expressed by some members of the Court suggested the inevitability of the litigation now before us. We reaffirmed the concept of a constitutionally required level of education, equivalent to thorough and efficient, and the corresponding power to exceed that level; but we gave no further content to the warning that any excess spending must not dilute the constitutional obligation. We spoke in the context of a statute that guaranteed continuation of substantial disparities among school districts in educational expenditures per pupil. Despite the certainty of those disparities, we held the statute facially constitutional and awaited the day of its return when it would be attacked as applied.
The change of focus from the dollar disparity in Robinson I to substantive educational content in Robinson V is clear; it was the main theme underlying the Court's determination that the Act was constitutional. Noting at the outset that for the first time we had before us a statute that defined the constitutional obligation, provided for its implementation through both state and local administration, required that implementation to be monitored, directed the State to compel compliance where that monitoring revealed deficiencies, and provided a funding mechanism to achieve the constitutional goal, we observed that the state's school-aid provisions "must be considered, not in comparative isolation, but as part of the whole proposal formulated by the Legislature." Robinson V, supra, 69 N.J. at 463, 355 A.2d 129. Although the opinion sketched the State-aid formula, it dwelt in considerable detail on the new approach to a thorough and efficient education: its definition, its involvement of both the State and district in fleshing out the details, standards, and elements against which thorough and efficient was to be measured, its requirements of reporting and monitoring, *309 the whole range of remedies available to the Commissioner and the Board for corrective action to force districts to upgrade their educational offerings when monitoring revealed deficiencies, and the non-delegable duty of the State to do so. We described the process of district-by-district evaluation, monitoring, and corrective action intended to lead to thorough and efficient education everywhere in the state, not by a financial measuring rod but by an actual direct measurement of numerous factors that reflect the level of substantive education. The only question about financing was not whether it provided equal dollars per pupil indeed, we noted that "there may be and probably are legitimate differences between and among districts and students", id. at 464, 355 A.2d 129, but whether it was sufficient to support the entire system and its goal of achieving a thorough and efficient education throughout the state.The fiscal provisions of the Act are to be judged as adequate or inadequate depending upon whether they do or do not afford sufficient financial support for the system of public education that will emerge from the implementation of the plan set forth in the statute. We are no longer considering the needs of the public system as it existed before the 1975 Act. We assume the Legislature had this in mind when preparing the state aid clauses of this statute. [Ibid.]
We "acknowledg[ed] the diversity that will inevitably exist among these separate [districts]," id. at 459, 355 A.2d 129, and in referring to equal opportunity, we noted the Commissioner's and the Board's responsibility to assure that "throughout the State each pupil shall be offered an equal opportunity to receive an education of such excellence as will meet the constitutional standard." Id. at 459-60, 355 A.2d 129. In other words, it was not an equal opportunity, without qualification, but an equal opportunity only up to a point: to receive a thorough and efficient education.
The clear thrust of our decision was to render equal dollars per pupil relevant only if it impacts on the substantive education offered in a given district. Compliance with the constitutional mandate was to be determined on a district-by-district measurement, and if money was a factor in the district's *310 failure, the remedy was not to change the statute but to implement it by forcing the district to spend more or by supplying further state funds. Indeed, the only explicit suggestion for statutory change was our observation that the Legislature should consider addressing the question of "what kind of showing must be made by a school district asking for state assistance due to local inability to recruit needed funds." Id. at 466, 355 A.2d 129.
Most telling of our view of the constitutional mandate was that, while noting that the funding mechanism would equalize only about 64% of the districts if fully funded (368 out of 578, leaving 210 districts with ratable resources superior to the rest, id. at 465 n. 4, 355 A.2d 129), we nevertheless observed that the Act eliminated "gross disparities in per pupil expenditures and tax resources." Id. at 467, 355 A.2d 129 (emphasis supplied). Our description of the implementation of the Act, whereby the process of monitoring and evaluation results in a review "treating the school districts as separate entities," id. at 463, 355 A.2d 129, was consistent with our statement at the outset that whether the Act would be found to be constitutional after its funding and implementation would depend on a test of it "as applied in the future to any individual school district at any particular time...." Id. at 455, 355 A.2d 129. While not purporting to definitively describe the potential thrust of any future attack on the legislation as applied, we deemed the Act's structure and financing mechanism as effectively limiting that claim to its operation in a particular district.
The strength of our holding of facial systemic validity despite the certainty of continued substantial expenditure disparity from one district to the next is highlighted by the opposing strength of the two dissenting opinions. Both forcefully pointed out, as the centerpiece of their disagreement, the significant variations in financial resources and the almost certain consequent variations in educational expenditures and educational quality that are a fundamental ingredient in the guaranteed tax base formula of the Act. Our opinion, therefore, cannot be *311 viewed as momentarily putting aside the expenditure-disparity issue: it was at the heart of our decision. On the record then before us it was rejected.
Chief Justice Hughes' concurring opinion reflects the competing considerations that faced the Court. While concurring in result, the Chief Justice disagreed with our implied conclusion that the inequality of resources allowed under the Act did not threaten its constitutionality. His concurrence in the finding of constitutionality was clearly based on his evaluation of the importance of allowing the Legislature to devise a constitutional remedy. It was explicitly based on his hope that in the future the funding formula would be changed to assure substantially more equality. He expressed serious reservations about the ability of the Commissioner to achieve a thorough and efficient education by monitoring "one by one" 578 school districts, Robinson V, supra, 69 N.J. at 470, 355 A.2d 129 (Hughes, C.J., concurring); the preferred route to thoroughness and efficiency was through equality of funding and equality of expenditures per pupil. Id. at 471-75, 355 A.2d 129. Nevertheless, in deference to the Legislature's clear responsibility, the newness of the statutory response, and the lack of experience under it, he joined the majority, trusting this Court's continuing power, explicitly reserving the right to declare the Act unconstitutional if experience showed that in practice it did not afford a thorough and efficient education.
With that as background, Abbott v. Burke, testing the Act as applied, came before us first in a procedural controversy and now in full substance. As often occurs, constitutional questions shift and sometimes sharpen as relatively abstract issues take on factual substance. When we first viewed the apparent scope of the factual controversy in Abbott I, supra, 100 N.J. 269, 495 A.2d 376, as it bore on the issue of appropriate forum, we found it necessary to underline the basic holdings of the Robinson v. Cahill cases, including explicitly the power of local districts to spend beyond what was required for a thorough and efficient education, subject to the limitation that "such authorization *312 does not become a device for diluting the State's mandated responsibility," Abbott I, supra, 100 N.J. at 291, 495 A.2d 376 (quoting Robinson I, supra, 62 N.J. at 520, 303 A.2d 273). We also recognized the revision of Robinson I effected in Robinson V that the Court has "been constantly mindful that money is only one of a number of elements that must be studied in giving definition and content to the constitutional promise of a thorough and efficient education." Abbott I, supra, 100 N.J. at 292, 495 A.2d 376 (quoting Robinson V, supra, 69 N.J. at 455, 355 A.2d 129). But we gave, in view of the issues about to be projected, a new potential gloss to the constitutional obligation. For instance, in the context of plaintiffs' claim that the disparities in dollar expenditure disproportionately affected disadvantaged students, we recognized that the State not only had the power to spend in excess of the norm in view of the presumed greater needs of such students, but that it might be required to do so. Id. at 291-93, 495 A.2d 376. Our application of the constitutional standard first presented in Robinson I an "educational opportunity ... needed in the contemporary setting to equip a child for his role as a citizen and as a competitor in the labor market," Robinson I, supra, 62 N.J. at 515, 303 A.2d 273 reflects the context of the present case.[T]he thorough and efficient education issues call for proofs that, after comparing the education received by children in property-poor districts to that offered in property-rich districts, it appears that the disadvantaged children will not be able to compete in, and contribute to, the society entered by the relatively advantaged children. [Abbott I, supra, 100 N.J. at 296, 495 A.2d 376.]
Finally, after noting plaintiffs' contention that "educational deficiencies could substantially be ameliorated only by increasing [local districts'] entitlement to state aid", id. at 285, 495 A.2d 376, we observed that the continuation of the local fiscal burden was "based in part on the Legislature's finding that local funding is important to encourage local involvement in public education, N.J.S.A. 18A:7A-2a(7)...." Ibid. We described a wide array of potentially relevant issues allowing the broadest presentation on remand of the parties' contentions and *313 evidence, and gave point to that intention by couching our remand with a condition that made it clear the case would first be heard by an ALJ.The OAL's [Office of Administrative Law] obligation and capacity to designate specially qualified judges cannot be overemphasized in the present context. This litigation exemplifies the type of complex, sensitive, important, multi-issue, and cross-disciplinary matter that the Legislature contemplated in authorizing the selection of persons to serve as administrative law judges both from within and without state government. See N.J.S.A. 52:14F-6 b. Thus, remand to the administrative agency is particularly appropriate in this case. We anticipate that the OAL will conduct a thorough hearing, where the parties shall present all their evidence relevant to the constitutional claims and defenses. This will serve to consolidate all fact-findings in a single proceeding. We intend that the proceedings will promote development of a complete and informed record, which will reflect determinations of appropriate administrative issues as well as the resolution of factual matters material to the ultimate constitutional issues. [Id. at 302-03, 495 A.2d 376.]
Thus, while leaving the door open to the numerous factual and legal contentions of the parties, we reiterated the constitutional mandate as it had developed through Robinson V. But we added a new element of considerable relevance to this case. We said, in effect, that the requirement of a thorough and efficient education to provide "that educational opportunity which is needed in the contemporary setting to equip a child for his role as a citizen and as a competitor in the labor market," Robinson I, supra, 62 N.J. at 515, 303 A.2d 273, meant that poorer disadvantaged students must be given a chance to be able to compete with relatively advantaged students. The Act and its system of education have failed in that respect, and it is that failure that we address in this case.
Issues similar to that before us today have been litigated in various state courts. Shofstall v. Hollins, 110 Ariz. 88, 515 P.2d 590 (1973); Dupree v. Alma School Dist. No. 30, 279 Ark. 340, 651 S.W.2d 90 (1983); Serrano v. Priest, 5 Cal. 3d 584, 96 Cal. Rptr. 601, 487 P.2d 1241 (1971) (later history omitted); Lujan v. Colorado State Bd. of Educ., 649 P.2d 1005 (Colo. 1982); Horton v. Meskill, 172 Conn. 615, 376 A.2d 359 (1977) (later history omitted); McDaniel v. Thomas, 248 Ga. 632, 285 S.E.2d 156 (1981); Thompson v. Engelking, 96 Idaho 793, 537 P.2d 635 (1975); People of Illinois ex rel. Jones v. Adams, 40 *314 Ill. App.3d 189, 350 N.E.2d 767 (App.Ct. 1976); Rose v. The Council for Better Educ., Inc., 790 S.W.2d 186 (Ky. 1989); Hornbeck v. Somerset County Bd. of Educ., 295 Md. 597, 458 A.2d 758 (1983); Milliken v. Green, 390 Mich. 389, 212 N.W.2d 711 (1973); Helena Elementary School Dist. 1 v. State, 784 P.2d 412 (Mont. 1990) modifying 769 P.2d 684 (Mont. 1989); Board of Educ., Levittown, etc. v. Nyquist, 57 N.Y.2d 27, 453 N.Y.S.2d 643, 439 N.E.2d 359 (1982), appeal dismissed, 459 U.S. 1139, 103 S. Ct. 775, 74 L. Ed. 2d 986 (1983); Board of Educ. of City School Dist. v. Walter, 58 Ohio St.2d 368, 390 N.E.2d 813 (1979); Fair School Fin. Council of Oklahoma, Inc. v. State, 746 P.2d 1135 (Okla. 1987); Olsen v. State, 276 Or. 9, 554 P.2d 139 (1976); Danson v. Casey, 484 Pa. 415, 399 A.2d 360 (1979); Richland County v. Campbell, 294 S.C. 346, 364 S.E.2d 470 (1988); Edgewood Indep. School Dist. v. Kirby, 777 S.W.2d 391 (Tex. 1989); Seattle School Dist. No. 1 of King Cty. v. State, 90 Wash. 2d 476, 585 P.2d 71 (1978); Pauley v. Kelly, 162 W. Va. 672, 255 S.E.2d 859 (1979); Kukor v. Grover, 148 Wis.2d 469, 436 N.W.2d 568 (1989), reh'g denied, 443 N.W.2d 314 (1989); Washakie Co. School Dist. No. One v. Herschler, 606 P.2d 310 (Wyo. 1980), cert. denied sub nom. Hot Springs Co. School Dist. No. 1 v. Washakie Co. School Dist. No. 1, 449 U.S. 824, 101 S. Ct. 86, 66 L. Ed. 2d 28 (1980). Their resolution has depended on the court's interpretation of the state's constitutional education provision and/or the state's equal protection doctrine. These state constitutional claims, the underlying contentions and facts, although presenting great variety of detail, are remarkably similar to those facing us: an educational funding system that depends on a combination of state and local taxes producing disparity of expenditures in the face of inverse disparity of need. Fourteen of the states have rejected both constitutional claims; six, including New Jersey, have held the state system of financing education invalid under the state education article, while rejecting or declining to reach *315 equal protection claims; three determined that the existing system violated both claims, and one that the system violated only equal protection.
Almost invariably the remedy extended no farther than the observation that the Legislature will presumably revise the system to conform with the Court's decision, the Court frequently reserving jurisdiction in order to impose a judicial remedy if the Legislature failed to act.
Very few of the cases have a factual record that even begins to approach that before us. None has the unique attribute of this case: an educational funding system specifically designed to conform to a prior court decision, having been declared constitutional by the Court but now attacked as having failed to achieve the constitutional goal. In short, we are the only state involved in a second round on this issue. The command of our thorough and efficient clause is strong and clear, but to the extent that further interpretation is required, to the extent that questions of conformance to the constitutional command exist, and difficult questions remain open, we cannot look out of state for an answer it must be found through the interpretation of our own Constitution, with the aid of the parties and the numerous amici who have participated in this case.*316 III. Summary of the Issues
The numerous factual and legal contentions in this case and their interrelationships have led to a record of enormous length and complexity. Even our significant compression of them may be difficult to follow. We offer the following overview of the main issues, and our resolution of them, to explain the significance of the evidence.
1. The initial issue is whether, and to what extent, our State's educational system can be judged in terms of disparate dollar input in the face of proof of the actual education being delivered. The substantive content of thorough and efficient has been legislatively defined and sustained by this Court. The State claims it is being delivered. Local district goals and standards have been established, reports of the status of education and its improvement are filed, the Commissioner's agents are examining them to determine their adequacy, corrective measures are taken, education is evaluated, and, finally, the Commissioner, based on all of this, determines its sufficiency, determines whether it is thorough and efficient. In view of that kind of evidence, what is the relevance of the fact of disparity of expenditures between two districts if the Commissioner decides that both are delivering a thorough and efficient education? Is disparity still an issue when proof of an adequate substantive education is offered? Does the issue become one of determining whether the substantive education is in fact thorough and efficient regardless of the difference in the dollar input per pupil?
We find the largely circumstantial evidence of substantive education, while important and relevant, insufficient to exclude consideration of dollar input and expenditure disparity. We are not satisfied that a thorough and efficient education is being provided in certain districts; therefore we cannot disregard expenditure disparity as irrelevant to whether such an education is being delivered.
*317 2. The Commissioner and the Board claim that the legislative and administrative scheme provides a thorough and efficient education to practically every district with but few exceptions. The State's proof consists of the Act, the rules and regulations, and their implementation by the Commissioner. The system consists largely of definitions of required minimum content and administrative oversight to assure conformance with certain criteria. They include plans, goals, and standards that represent a combination of State and local definitions of educational objectives. All of this is monitored by the Commissioner's agents, who review budgets and other reports of the districts to assess compliance with the districts' goals and minimum statutory criteria. While the procedure suggests a certain educational content, there is no standard of the breadth of curriculum that must be offered, no standard of other commonly accepted educational criteria (staffing ratios, faculty experience and training, staffing of special positions and their number), and no broad-gauged standard of performance of any district, school, or pupil apart from the statewide tests mentioned later. We find that despite great effort, the procedure ultimately fails to measure the district's system against the accepted definition of a thorough and efficient education one that equips its students to fulfill their roles as citizens and competitors in the market.
Although the record contains proof of substantive education, we cannot determine whether that education equates with thorough and efficient. The record is inadequate, in absolute terms, to demonstrate thoroughness and efficiency. While the Commissioner asserts that his certification pursuant to the statute is presumptive proof of a thorough and efficient educational system, we find the record devoid of persuasive factual evidence as distinguished from a presumption that certification in practice means that students are equipped for their roles as citizens and workers. We do not mean to deprecate at all the process mandated by the statute and implemented by the Commissioner and the Board quite the contrary, it may very *318 well be effective. We mean only that on this record it has not been shown to lead to a thorough and efficient education. There is no evidence that certification, the process, or the educational content now present affords New Jersey's students the opportunities that will equip them in their roles as citizens and competitors in the labor market. We recognize, of course, that the burden of proof on this issue is plaintiffs', not the Commissioner's.
3. The companion issue to whether the State has proved that a thorough and efficient education exists is whether plaintiffs have met their burden of proof that it does not. Here the issue is the significance of plaintiffs' proofs on two levels: their proof of the failure of education in poorer urban districts, and the comparison between this education and that of the richer suburban districts. As we have noted about the Commissioner's evidence, so we affirm as to plaintiffs': they have not proved what is necessary for a thorough and efficient education. Neither their statistical nor other comparisons informs us of what a district needs in order to have thorough and efficient education; we are left without a standard against which to measure a given district's efforts. When arrayed against this absence of proof, the Commissioner's proof of the process of the DOE in implementing acceptable standards of education becomes more persuasive. Despite this lack of affirmative proof, have plaintiffs nevertheless proven that whatever that standard may be, their districts clearly fall below it? Does the combination of student need, disproportionately present in poorer urban districts, inferior course offerings, dilapidated facilities, testing failures, and dropout rates leave the issue in doubt? And what does the comparison with affluent suburban districts mean if the Constitution indeed requires that poor children be able to compete with the rich? On this issue we confront questions of substantive education: what is the importance of disparity in expenditures per pupil, course breadth, and staffing; are these indicators related to the quality of education that defines the opportunity? The Commissioner *319 contends that neither money nor those other indicators is proof of educational quality, that the "determinants" of the quality of education are found elsewhere, in management, in community relations, in parental interest, in staff attention, and in numerous other characteristics not clearly related to funding characteristics that transcend the relationship between the money spent and the education obtained. The "effective schools" concept and studies testing it are asserted as proving the validity of this view of the educational process a radical view in that it explicitly denies the conventional wisdom that the more spent the better the education; the more teachers the better the education; the more experienced the teachers the better the education. On this view of the educational process, expenditure disparity becomes irrelevant. The central issue is here joined, including whether a thorough and efficient education requires that children with greater needs are entitled to greater resources. Whatever else the evidence shows, it is clear that the reverse is the fact: in New Jersey today, as we assume in the United States, the greater the students' needs, the less their education. And this raises yet another issue: to what extent does the requirement of thorough and efficient education impose on the schools the responsibility to account for and attempt to remedy the problems students bring with them to the schools, intractable problems, problems never dreamed of in the past as being within the schools' responsibility, problems created not by the schools but by society?
On this issue we conclude that a significant number of poorer urban districts do not provide a thorough and efficient education for their students; that the measurement of the constitutional requirement must account for the needs of the students; that in most poorer urban districts, the education needed to equip the students for their roles as citizens and workers exceeds that needed by students in more affluent districts; that the education provided depends to a significant extent on the money spent for it, and on what that money can buy in quality and quantity and the ability to innovate. We do not dispute *320 the Commissioner's proof about "effective schools" or "determinants" except to the extent they are used to prove that disparity of expenditures does not count, and that the conventional indicators do not count.
4. Although plaintiffs claim that the Act and the entire system of education is unconstitutional based on their proof of disparity between poorer and richer districts, a vast gulf exists between the richer and the poorer districts, a gulf that includes all of the districts in the middle concerning which the proofs are limited. We know little about the quality of the education they offer, other than their expenditures per pupil and their certification. Neither the State nor the plaintiffs has attempted to prove that their education is or is not thorough and efficient.
As a matter of law, in the absence of further evidence, we cannot conclude that the education in such districts is not thorough and efficient; furthermore, on this record and given the litigated context plaintiffs attacking the entire system and the sufficiency of education in specific districts, the State defending by proof of its process we must accord deference to the administrative scheme and the actions of the Board and the Commissioner to the extent of ruling that plaintiffs have failed to prove that a thorough and efficient education does not exist in those districts between the richer and poorer ones. We do not foreclose the possibility that at some other time a different record might lead to the conclusion of total systemic failure that sweeps in even such districts, but not on this record. In the absence of proof, we believe the separation of powers requires us to defer to the Board's, the Commissioner's, and the legislative judgment concerning such districts. Given the consequences, a conclusion of constitutional deficiency cannot hang by a thread, it must rest on granite. The result is that we conclude, for the overwhelming number of districts in this State, that there has been no showing that the constitutional mandate has not been satisfied.
*321 5. Has the State proven that given more time, it will be able to achieve a thorough and efficient education in all districts under the present system? The Board and the Commissioner claim that thorough and efficient exists now, but that in any event the Act and procedures assure its arrival in the near future. Assuming one agreed, it would be hard for this Court to justify the radical interference with the legislative power that is involved in the constitutional determination of insufficiency. The constitutional command does not require relief every time the slightest deviation from a thorough and efficient education is found, or any time that deviation, though proven, is likely to be corrected soon. Were we confident that a thorough and efficient education were likely to be achieved in the near future under the present system, we would not dream of intervening. We do believe the Commissioner is making progress; we do believe that the Board is making progress; we do believe that the Legislature would, if necessary, change the Act to provide for further progress.
Nevertheless, we reject the State's claim that in these poorer urban districts a thorough and efficient education has been or will be achieved. The extent of failure is so deep, its causes so embedded in the present system, as to persuade us that there is no likelihood of achieving a decent education tomorrow, in the reasonable future, or ever. The State's argument is strong on paper: districts can raise all the money they want, districts must raise all the money they need to provide a thorough and efficient education, and if that fails, the State must pay the way, and the Commissioner must monitor all of this and correct any deficiencies. But for ten years and more there has been no thorough and efficient education in these districts. The factors that lead to this failure are described later, but the simplest is that these districts are just too poor to raise the money they theoretically are empowered to. We can keep the present system and its promise for the future for other districts without sacrificing these poorer urban districts to perpetual failure. *322 They can, and as we view it, constitutionally they must, be treated differently. Judicial deference can go just so far.
6. Can the Act and its system of education be declared unconstitutional in its entirety as plaintiffs demand even though the failure of a thorough and efficient education has been proven only for a relatively small number of districts? Certainly, if the failure is not remedied in accordance with the decree issued by this Court, such a declaration may be the Court's only recourse. In Robinson I there was no showing that any district failed to achieve a thorough and efficient education. Rather, it was the absence of substantive educational proofs a measure to assess whether the constitutional command was being met that led the Court to declare the statute unconstitutional in its entirety. While we have the trappings of the substantive education being provided, the words of the statute, rules and regulations defining the content of a thorough and efficient education, do we have any evidence that tells us that the students of this state are being equipped for their roles as citizens and workers? There is a limit to this Court's powers in this area: educational sufficiency ultimately must be a responsibility of the Legislature; we do not sit as the Commissioner's superior in these matters, as the Board's overseer in its efforts. All we are entitled to ask for under the Constitution is proof of an effort to achieve the goal and a reasonable success, and when the State has established a structure of the kind that we have seen in the record before us, when it has spent more than most any other state in the nation on education, we cannot say that it has failed to satisfy its own Constitution. Not on this record on a statewide basis. We can say only that as to certain students, it has failed, as they have failed, and that this failure must be remedied.IV. Facts and Conclusions
We note in this section of the opinion some of our factual and legal conclusions. We describe at the outset (Section A) the *323 Act's funding scheme indicating the extent to which it theoretically equalizes districts' ability to raise money for education, and the extent to which it relies on the local tax base. In Section B we note the evidence of the vast disparity in educational funding that continues to exist, corresponding to the districts' property wealth, despite the Act's equalizing provisions. The basic position of the State is then analyzed (Section C), namely, that despite these funding disparities the Act assures and has achieved a constitutionally adequate education in all districts through the operation of its provisions defining a thorough and efficient education and granting the Commissioner the power to monitor and correct districts' performance, including the power to require a district to increase taxes if its educational budget is insufficient. In that section the evidence concerning the impact of "municipal overburden" on that power of the Commissioner to force districts to increase taxes is noted, and the conclusion reached that it renders that power impotent. We also deal in that section with the Commissioner's power to certify districts as providing a thorough and efficient education, and with the significance of such certification. In response to the Commissioner's claim that a constitutionally adequate education is being delivered, we then describe (Section D) the record evidence of the quality of education actually provided in poorer urban districts and contrast it with that in richer suburban districts. The critical educational needs of the students of these poorer urban districts are then set forth (Section E), including the evidence of their performance on State-mandated tests, along with a description of the special innovative approaches to education thought to be required to meet these needs. Having described the evidence of a failing education in poorer urban districts, the contrast with that afforded in affluent districts, and the disparity in funding between the two, we deal next with the Commissioner's contention that the difference in education between the two is not caused by the amount of expenditures per pupil, but by mismanagement in some of the poorer districts (Section F). Our *324 conclusion that money does make a difference leads to a determination of the invalidity of the minimum aid portions of the Act.A. The Funding Scheme
The Act's funding scheme is described in considerable detail in Robinson V, supra, 69 N.J. at 478-90, 355 A.2d 129. It is based on a limited equalizing of the taxing power of school districts. It enables all school districts to raise funds as if their tax base were at least 134% of the average school district tax base ("tax base" meaning in this connection the district's equalized property valuation per pupil). N.J.S.A. 18A:7A-3, -18a. The school district sets the tax rate as if the real property of the district equaled this guaranteed tax base (GTB). The local revenues generated by the tax from the district's actual tax base are then supplemented by state aid, called "equalization aid," in an amount that, when added to these local revenues, equals what that tax rate would have produced if applied to the GTB.
In the school budget year 1985-86, the GTB was $250,927, with equalization aid totaling $1.34 billion. During that year it had the effect of equalizing the taxing power of 371 districts out of a state total of 578 districts, roughly two-thirds of the state's school districts. The equalization aid for the prior budget year (1984-85) aimed at giving effect to the GTB ($223,100) was $1.24 billion. Had the Act been fully funded, it would have been considerably more. Districts whose school tax base in fact exceeds the GTB do not receive any equalization aid at all.
*325 In its funding aspect, the Act does not require or assure any particular level of educational expenditure in any school district. It is indifferent to whether the district spends $1,500 per pupil or $15,000. As far as equalization aid is concerned, its only effect is to pay a portion of the school budget determined by the district. A district can decide to raise $5 million or $2 million; under the Act that is a matter solely for the district to decide. Both state equalization aid and the local tax rate would be affected by that decision. The potential legislative appropriation for equalization aid, therefore, is dependent not only on the level of the GTB but on local budget decisions. There is no limit theoretically on the district's ability to tax and spend.
The Act gives poorer districts poor in terms of property valuation per pupil taxing power to raise more money for school purposes than what a school district with the average property valuation and no equalization aid could raise. Putting aside for the moment the impact of "municipal overburden" (a condition in many poorer districts where the cost of local government police, firefighters, other municipal employees, road maintenance, garbage collection, etc. is so high that the municipality and the school district are reluctant to increase taxes for any purpose, including education), equalization aid attempts to obliterate the enormous disparity between rich and poor for school tax purposes; it creates, instead of rich and poor districts, two different classes: those districts with a guaranteed tax base almost two-thirds of the districts in the state and those with a tax base in excess of the guaranteed tax base of $223,100, running from $223,667 to $7.8 million and clustering at $300,000 (1984-85 figures). To understand better its equalizing impact, the GTB and equalization program should be compared with the situation that would exist if there were no equalization aid whatsoever (the districts' real property *326 taxing power ranging from $22,322 per pupil to $7.8 million per pupil, with 58% of the districts below average, 42% above many enormously poor and a fair number very rich); and with the situation that existed prior to the Act when equalization aid leveled only 157 districts as compared with the 368 not equalized. Robinson V, supra, 69 N.J. at 465 n. 4, 355 A.2d 129.
Three limitations on equalization aid should be noted. Most important of the three, the amount of equalization aid a district receives in a budget year is based not on its budget for that year, the current year, but on the budget of the prior year.
Just how severe the impact of that limit can be is realized when a district with equalization aid at a level, for instance, of 80% of the district's budget is analyzed. Assume that district has a $4.2 million budget for this year, representing a spending increase of $200,000 (last year's budget was $4 million). For $4 million of that $4.2 million budget, 80% aid will be forthcoming the district will have to raise only $800,000 locally to have a total of $4 million. But to get the extra $200,000, the school district will have to raise that entire amount on its own tax base, and the impact on its tax rate will be five times as much as it would have been (because, on the assumption given, the State would have paid 80% of that $200,000 but for this "prior year" equalization aid rule). This failure to provide current year funding affects districts' willingness to add to or enrich their programs in view of what could be a substantial tax impact. Both the Commissioner and the Board, in their decisions in this case, support legislative change to a current year funding system.
The second limitation is the budget cap law, applicable to all districts, restricting annual increases in school district budgets to a certain percentage over the prior year, but allowing low *327 spending districts to increase their budgets more rapidly than higher spending districts. N.J.S.A. 18A:7A-25. This restriction affects equalization aid by limiting the total budget on which such aid is based. The budget cap law is not as important to poorer districts, who do not ordinarily budget to cap. However, the statute allows the Commissioner to waive the cap limitation.
The third limitation cuts off equalization aid to the extent that the district's budget, in terms of expenditures per pupil, exceeds that dollar per pupil amount that is the sixty-fifth percentile of all school districts' budgets. In other words, if the district with a lower tax base per pupil than the guaranteed tax base nevertheless spends more than the statewide average expenditure per pupil, it will receive equalization aid even for the excess expenditure up to the point where the district's per pupil expenditure equals the sixty-fifth percentile of all districts, i.e., equalization aid will stop as the school district's budget per pupil approaches that of the highest-spending districts of the State. Again, this restriction is of little importance to the poorer districts.
For those districts with property valuation above the guaranteed tax base, the Act provides "minimum aid." Keyed to property wealth and the size of the district's budget, this *328 element of the funding scheme, while not as substantial as in the past, remains an expenditure that is counter to equalization (aiding the richer districts instead of the poorer districts). In 1984-85, for instance, minimum aid totaled $93 million (as compared to $1.2 billion of equalization aid in that same year); immediately prior to the Act, however, minimum aid equaled $290 million as compared to equalization aid of $431 million for the same year. In 1989-90, $163 million in minimum aid was distributed; equalization aid totaled $1.9 billion.
The Act also provides for "categorical aid," amounts given by the State to districts regardless of their wealth or the total size of their budget, based on the recognition that certain essential programs cost more than others. Categorical aid supports special education for students with particular handicaps, compensatory or remedial education, and bilingual education. The aid formula consists of calculating the statewide average expenditure per pupil (measured by the district Net Current Expense Budget (NCEB)) and multiplying it by a statutory fraction assigned to the category, N.J.S.A. 18A:7A-20, the product being the amount a district will get in additional aid for each pupil in that category. For instance, in 1984-85 the additional cost factor for compensatory education was .18, and bilingual education .23 (the additional cost factors can be changed each year by the Governor unless the Legislature objects, N.J.S.A. 18A:7A-21). In 1984-85 the statewide average expenditure per pupil measured by NCEB was $3,329, and therefore a district with 100 students taking compensatory education courses would have received $59,922 in compensatory education categorical aid.
*329 In 1984-1985, total categorical aid was $315.4 million ($206.1 million for special education; $88.1 million for compensatory education; $21.2 million for bilingual education). Categorical aid is distributed as a flat grant per pupil rather than in accordance with the GTB equalization aid formula.
Finally, there is transportation aid and pension aid, the latter consisting of the State's contribution to the Teachers' Pension and Annuity Fund (TPAF), both of which are non-equalizing. The transportation aid is a percentage of the total transportation costs of a district and is distributed in a way that bears no relationship to the wealth of the district. Payments by the State to the TPAF, however, are affected by the districts' wealth, since the richer districts tend to have more and better paid teachers per pupil than the poorer districts. The contribution to the TPAF is a substantial amount ($535.8 million in 1984-85) and is counter-equalizing. There are, however, as was noted in Robinson IV, supra, 69 N.J. at 149-50, 351 A.2d 713, administrative and other problems that would arise were this aid terminated on its present basis and instead distributed pursuant to the equalization aid formula.
Of considerable importance, in the State's view of this case, is federal aid. For many years, since 1965, the Elementary and Secondary Schools Act (now titled the Strengthening and Improvement of Elementary and Secondary Schools Act), 20 U.S.C.A. § 2701 et seq. (Supp. 1989), has provided needed extra *330 funding for programs for disadvantaged students. Federal aid is fairly narrowly targeted not only as to district but as to use. Its impact on poorer districts is indicated by the numbers: in 1984-85 Camden received $394 per pupil; East Orange, $166 per pupil; Jersey City, $471 per pupil; Irvington, $320 per pupil; Newark, $808 per pupil; Trenton, $480 per pupil; and Paterson, $244 per pupil. Although substantial, federal aid amounts only to approximately 5% of the total of all districts' current expense budgets (nationwide, federal aid is about 8% of districts' budgets). The State, in its arguments concerning both the adequacy of school budgets in poorer districts and the asserted disparity in expenditures per pupil between such districts and richer districts, often uses expenditure figures that include federal aid. When federal aid is included, the level of funding per pupil in poorer districts is increased and the disparity of such funding compared to richer districts decreased, sometimes dramatically.
The question of the significance of this aid and its appropriate function in school finance analysis was very much at issue in this case. While obviously such aid must be acknowledged, we have determined it should not affect our conclusions concerning either constitutionality or remedy, both as a matter of constitutional interpretation and federal law. Briefly, we view the State's constitutional obligation to provide a thorough and efficient education as not adequately satisfied if dependent on federal aid, which today is subject to substantial fluctuation. Plaintiffs' witness called it a "roller coaster."*331 Furthermore, as we read the federal law, even the mere consideration of federal aid in determining the need for or amount of state aid would violate the federal aid statute. 20 U.S.C.A. § 2854 ("No State shall take into consideration payments under [20 U.S.C.A. §§ 2701 et seq.] in determining the eligibility of any local educational agency in that State for State aid, or the amount of State aid, with respect to free public education of children."). We deem it inevitable that a constitutional determination and remedy based on the assumption that federal aid will continue would have such an effect. See also 20 U.S.C.A. § 2728 (conditioning receipt of federal funds on maintenance of funding effort by the local district and the State and further providing that federal funds shall be used to supplement rather than supplant funds from State and local sources).
Put differently, federal aid, targeted solely at helping poor children, is not intended to enable a state to keep in place a funding scheme that disproportionately penalizes them. If New Jersey's funding scheme were equal and fair, which it is not, federal aid would continue, and together with state and local expenditures it would provide an even greater opportunity to educate disadvantaged children. The State's position in advancing such an argument in order to save this failed system gives us concern. If New Jersey's educational funding system is not thorough and efficient on its own, we question whether the State should want to preserve it by relying on federal aid, at the inevitable cost of depriving poorer children of their full measure of help.
Finally, to the extent that the constitutional obligation is measured by the regular education provided by the district (the NCEB), federal aid is irrelevant. Federal aid is targeted for specific uses, e.g., compensatory education, special education, and bilingual education. Like state categorical aid, it is not *332 intended to diminish disparities in educational expenditures per pupil.
We do not mean to suggest that a state will always be foreclosed from considering federal aid in determining the constitutional adequacy of its educational system, or that the remedies designed to redress constitutional failure may not someday be affected in some way by such aid. The federal law cannot mean that the State must forever be totally blind to its existence. The implications of these federal restrictions and their potential conflict with the State's obligation to devise a funding system in the light of all relevant factors, including federal aid, may someday have to be resolved. For present purposes, however, given today's level of such aid, its fluctuating nature, and its purpose, it is not a critical factor in determining the underlying constitutional issue.
State aid for capital expenses, whether capital outlay or payment of debt incurred to fund such outlay, is also provided to districts below the guaranteed tax base pursuant to the equalization formula. If, for instance, a district spends $500,000 for capital items, including expenditures to fund debt previously incurred for capital items, the State will reimburse the district under the equalization formula used for current expenses. The needs for capital construction throughout the state are so vast that all parties, including the State, agree that the present system is not sufficient to meet that need. The State contends, however, that a thorough and efficient education is being delivered throughout the state, unaffected by this unmet need, and that maintenance and minor capital improvements can be adequately funded through equalization aid.
In our discussion of the factual issues there are some terms frequently used that should be explained. "Richer suburban districts" means those suburban districts whose indicators of wealth per capita income, property values, and socioeconomic status are the highest. "Poorer urban districts" includes urban *333 areas with the highest poverty indicators and the lowest socioeconomic status; its precise scope is set forth infra at 338-342, 575 A.2d at 384-386. "Net Current Expense Budget" (NCEB) is the school's expenditure budget for the year less all federal aid, categorical aid (and any other state aid except equalization aid), transportation expenditures, monies unused in prior years ("appropriated free balances"), as well as miscellaneous revenues. Essentially, it is the sum of local school district property tax revenues and state equalization aid for current expenses. The ALJ noted that "on a statewide basis, NCEB is about 80% of the current expense budget and is sometimes thought to represent the district's cost for its regular education program." In that sense, the "regular education program" is intended to cover the education offered in the district except those programs supported by categorical and federal aid. The "NCEB per pupil" is simply the NCEB divided by the number of resident pupils of the school district. "Current expenditures per pupil" include all expenditures other than those required for debt service and capital outlay, or expenditures supported by federal aid. Because "current expenditures" includes categorical aid, it is often expressed in terms of "current expenditures per weighted pupil," which are the current expenditures divided by the number of resident pupils augmented by the weighting factors. The weighting factors, or additional cost factors, for each educational need category are set forth in the statute. N.J.S.A. 18A:7A-20. For instance, a pupil taking a compensatory education course would count for more than one pupil, in fact, in the year 1984-85 would count for 1.18 pupils. The "day school expense" is the total expenditures of the district for the year in question, including all federal aid and state categorical aid, in addition to the items included in the NCEB. "Day school expenditures per pupil" is calculated by dividing the day school expenditures by the number of resident pupils, and although it includes expenditures for categorical programs, it does not weight pupils enrolled in those programs.*334 B. Educational Funding Disparities
Plaintiffs intended to prove that the funding and spending disparities referred to in Robinson I are worse now than they were before adoption of the Act. They have done so. Prior to the Act, in 1971-72, the spread between the lowest and highest spending districts was $700 per pupil to $1,500 per pupil, a difference of $800. Robinson V, supra, 69 N.J. at 480 n. 4, 355 A.2d 129. This difference of $800 per pupil reflects the range between the absolute highest spending districts and the absolute lowest. Plaintiffs' expert calculated the range of expenditure disparity from 1975-76 through 1984-85 from the fifth to the ninety-fifth percentiles, i.e., she excluded from her analysis those districts, each containing five percent of the pupils in the state, with expenditures at the highest and lowest extremes. Expenditures were measured by current expenditures per weighted pupil, including some federal funds (excluding, however, federal aid under Chapter 1, 20 U.S.C.A. § 2701 et seq., which is the major source of federal aid). In 1975-76, the year immediately prior to the Act, districts at the fifth percentile spent $1,076, while districts at the ninety-fifth spent $1,974, a difference of $898. For a few years after the Act was funded, the disparity diminished, but thereafter it generally increased to the point where in 1984-85 districts at the fifth percentile spent $2,687, while districts at the ninety-fifth percentile spent $4,755, a disparity of $2,068 per pupil. The disparity in expenditures is significant even when adjusted for inflation in 1975-dollars, the disparity grew from $898 per pupil in 1975-76 to $1,135 per pupil in 1984-85.
The impact of this disparity, solely in dollar terms, is that on the average, in 1984-85, a group of richer districts with 189,484 students spend 40% more per pupil than a group of poorer districts with 355,612 students; one provides an education *335 worth $4,029 per pupil, the other, $2,861. For instance, a 3,000 pupil district in a poorer area has a budget of $8.6 million, while a relatively wealthy suburban district with 3,000 pupils has a budget of $12.1 million. The disparity is not as pronounced in the middle, which covers 339 districts and 575,181 students.
That disadvantage in expenditures per pupil is clearly related to all of the other aspects of poverty that define poorer urban districts and their youth. Although the statistical relationships are the subject of considerable dispute, we conclude that generally they show that the poorer the district measured by equalized valuation per pupil, or by other indicators of poverty the less the per pupil expenditure; the poorer and more urban the district, the heavier its municipal property tax, the greater the school tax burden; whatever the measure of disadvantage, need, and poverty the greater it is, the less there is to spend.
Statisticians advise that the relationships suggested in this record do not prove causal connection, that the poorer the district the lower its expenditures per pupil does not mean that low property valuations per student cause low expenditures per pupil. It is only when the relationship is strong and when other statistical techniques are applied so as to exclude other causes that a fair conclusion may be reached, one that inevitably depends on the experience and logic that suggests a causal relationship in the first place. Here the strength and consistency of the relationship between dollars per pupil and these other *336 factors convince us that there is such a causal relationship, although that conclusion is not essential to our determination.
We note the sharp dispute between the statisticians for plaintiffs and those for the State despite their total agreement on the underlying facts. The experts used by the Board and Commissioner, some outside consultants, some DOE officials, all of admitted qualifications, deny the strength, consistency, and even the existence of these relationships. Without in any way impugning plaintiffs' experts' overall qualifications as education finance analysts, the State's witnesses argue that plaintiffs' statistical analysis is rudimentary, consisting largely of comparisons between districts at the high and low ends of the spectrum. They claim that more sophisticated methods should be used to test the asserted relationship against the entire distribution. They claim, for instance, that such analysis demonstrates that property wealth accounts for only 21% of the differences in per pupil expenditures between districts, a figure that excludes federal aid. Plaintiffs, however, claim that this figure ignores differences in the effort needed to raise the funds the school tax rate and that if this factor is considered, the figure is 47%. Most of all, however, the State's expert witnesses' fire is aimed at the undeniable aspects of inconsistency in the relationships: some low property value districts spend more than the State average per pupil; large groups of districts spend more than other large groups with higher property values per pupil; and in the vast number of districts whose indicators of property wealth or poverty fall between the extremes, not only are there numerous district-by-district inconsistencies with the asserted property wealth/expenditure per pupil relationship, but even when averages of the property wealth of groups of districts are examined, the corresponding increase in per pupil expenditures is less than substantial.
For most districts, plaintiffs have failed to prove substantively that a thorough and efficient education does not *337 exist. Furthermore, their proofs of expenditure disparity in those districts are not sufficient to overcome the deference owed to the conclusions of the Commissioner and the Board. Given the State's attempt to achieve thorough and efficient education, we doubt that any showing of expenditure disparity in the absence of some independent evidence of substantive failure would warrant a conclusion that a thorough and efficient education does not exist. But expenditure disparity does play a role, an important one, in our conclusion that the constitutional level has not been achieved in the poorer urban districts. This disparity has multiple relevance: to the extent educational quality is deemed related to dollar expenditures, it tends to prove inadequate quality of education in the poorer districts, unless we were to assume that the substantial differential in expenditures is attributable to an education in the richer districts far beyond anything that thorough and efficient demands; it indicates even more strongly the probability that the poorer districts' students will be unable to compete in the society entered by the richer districts' students; and by its consistency over the years, it suggests that the system as it now operates is unable to correct this.
Our reasoning at this point can be compared with Robinson I where we concluded that the expenditure disparity could not be remedied under the existing system, given its heavy reliance on the property tax.Upon the record before us, it may be doubted that the thorough and efficient system of schools required by the 1875 amendment can realistically be met by reliance upon local taxation. The discordant correlations between the educational needs of the school districts and their respective tax bases suggest any such effort would likely fail.... [Robinson I, supra, 62 N.J. at 520, 303 A.2d 273.]
So too here we do not believe a thorough and efficient education in the poorer urban districts "can realistically be met" by reliance on the system now in place. While local taxation no longer has the same impact, it is still significant. More than that, however, we believe that because of the complex factors leading to a failure of thorough and efficient in the poorer *338 urban districts, including disparity of expenditures, we are no more likely ever to achieve thorough and efficient than we believed we could by relying on local taxation in Robinson I. Combined with these disparities of wealth and expenditure are the much more serious disparities of educational need, students in the poorer urban districts dramatically disadvantaged compared to their peers in the affluent suburbs. These intractable differences of wealth and need between the poorer and the richer, and the "discordant correlations" within a poorer district between its students' educational needs and its ability to spend, are more than the present funding system can overcome. The failure has gone on too long; the factors are ingrained; the remedy must be systemic. The present scheme cannot cure it.
In order to understand the statistical comparisons between districts, some explanation of terms and methodology is required. One of the key statistical elements is the listing and division of districts by socioeconomic status (SES) prepared some time ago by the DOE. Districts are arranged in ten groups, known as District Factor Groups (DFGs), and designated DFG A through DFG J, A being the group with the lowest socioeconomic status, J the highest. The DOE in 1974 devised this method of measuring socioeconomic status through a numerical ranking of every district in the state. The measurement consisted of seven factors: 1) per capita income level, 2) occupation level, 3) education level, 4) percent of residents below the poverty level, 5) density (the average number of persons per household), 6) urbanization (percent of district considered urban), and 7) unemployment (percent of those in the work force who received some unemployment compensation). Each factor is weighted in accordance with the DOE's evaluation of its relative importance as an indicator of such status, then combined by a formula that produces a numerical result. Both by its usual meaning and by examining the factors, one can divine the clear intent of the measurement as quantifying the advantages and disadvantages of a district and its students for educational purposes in terms of their social *339 and economic background and environment. The DOE divided all districts into ten groups, each having approximately fifty districts but with differing numbers of students in each group. DFG A, therefore, includes the lowest fifty-five districts in terms of socioeconomic status of all ten groups with 257,000 students in 1984-85, and DFG J, the highest fifty-two districts, with 69,576 students in 1984-85. Although the particular test for which this array was designed has been abandoned, the DOE continued to update the socioeconomic status measurement, revising it in 1984 in accordance with the 1980 Census. It has been used extensively in this case.
The possible significance of the origin of this SES comparison is worth noting. It was initiated to enable districts of a particular SES to measure their performance against others like them. The DOE wanted to assist Trenton, for instance, in comparing its students' performance with those of Newark or Jersey City. Implicit is the conclusion that it would be pointless to make the comparison with Princeton or Cherry Hill. Without disputing the possible insight gained from such limited comparisons, we cannot avoid another side of this measurement. Such comparison, limited to districts with a similarly low SES, accepts the proposition that low SES districts should not be discouraged by their students' failure to perform at the level of high SES districts, or should not expect them to. The overall performance of their students should not be evaluated by measuring it against the performance of those in the affluent suburbs.
However helpful such analysis might be to administrators in particular districts, when used as a tool to evaluate a district's performance, it inevitably reinforces the belief that urban districts can go just so far and no farther. Newark simply cannot equal Millburn, Camden cannot equal Cherry Hill. We are *340 aware of the fact that a student's SES is considered by many to be the most important factor in predicting tests scores and other measures of performance; and the belief of some that it is an intractable factor. The other side of that issue is that the students of low socioeconomic status are just as capable, just as bright and industrious as those of higher SES, but that their socioeconomic status has overwhelmed their native talents.
The significance of a student's socioeconomic status and the ability of effective education to overcome it have been and are the subject of intensive study and debate. Like most other critical issues in education, there is no conclusive resolution. We are left with common wisdom: children in the poorest urban districts generally perform unsatisfactorily on tests; many, however, excel despite their SES. Effective education can help all of this, but to what extent is the subject of strong and learned disagreement.
We have decided this case on the premise that the children of poorer urban districts are as capable as all others; that their deficiencies stem from their socioeconomic status; and that through effective education and changes in that socioeconomic status, they can perform as well as others. Our constitutional mandate does not allow us to consign poorer children permanently to an inferior education on the theory that they cannot afford a better one or that they would not benefit from it.
Another statistical category used in this case is the "urban district," a category adopted by the DOE but originally compiled by the Department of Community Affairs pursuant to statute. There are fifty-six such districts, each an urban municipality. In 1984, Governor Kean, wishing to improve education in poorer urban areas, initiated a series of pilot projects to be undertaken by the DOE. The Governor provided some funding for these programs. He made it clear that his concern was with the quality of education in urban areas, in particular, the poorer urban areas. The Commissioner, in his implementation of the program, initially selected the school districts of *341 forty-nine municipalities as potential sites for the pilot projects. These were the municipalities commonly known as "urban aid districts," qualified by statute and implementing regulations of the Department of Community Affairs to receive annual state aid under N.J.S.A. 52:27D-178 to -181. The purpose of that statute was to identify those municipalities in the state requiring financial aid because of factors of poverty and need. It sets forth the factors to be measured by the Department of Community Affairs to some extent overlapping those used in defining DFGs in determining which municipalities are urban aid districts and the formula for apportioning the state aid among them. For many years, ever since the inception of such aid, the Legislature has regularly appropriated substantial sums for payment to these municipalities. The Commissioner considered the characteristics that led the Legislature and the Department of Community Affairs to designate these urban areas for continued aid to significantly define their educational needs and deficiencies. He subsequently added seven districts to the list. Five of these were districts in DFGs A and B that had not been included under the Department of Community Affairs' formula.
These "urban districts" are the poorest and most needy municipalities in the state. The clear intent of the statute and its implementation was to "enable such municipalities to maintain and upgrade municipal services and to offset local property taxes," N.J.S.A. 52:27D-179 in other words, to enable them to provide municipal services without inordinate taxation. To qualify as an urban aid district, the municipality must be of a certain size or density, have a certain number of children whose *342 families are on welfare, have public housing, and a higher tax rate or lower property valuation per capita. Its share of the urban aid is measured by a formula that includes all of those factors. N.J.S.A. 52:27D-178.
Of these fifty-six urban school districts, those in DFGs A and B have the strongest characteristics of poverty and need. The factors that define both categories (DFG A or B and urban district) compel that conclusion; their official designation by the DOE confirms and reinforces it. The DOE's use of these factors for analysis of educational needs in connection with the districts' and students' performance demonstrates the direct relevance of these categories to some of the issues before us the ability of the twenty-nine poorer urban districts in DFG A and B to achieve thorough and efficient, now and in the foreseeable future; these students' need for an education beyond the norm based on their disadvantaged status; the excessive tax burden of municipal services. When these factors are combined with the evidence in the record of substantive failure of thorough and efficient education in these districts, including their failure to achieve what the DOE considers passing levels of performance on the High School Proficiency Test (HSPT), the constitutional failure becomes apparent, the need for a remedy urgent, and the focus of that need clear. That the overwhelming proportion of all minorities in the state are educated in these poorer urban districts is of further significance. These are the districts where not only the students *343 and education are failing, these are the districts where society is failing.
These twenty-nine poorer urban districts are not only in the lowest SES classification (DFGs A and B), but even their ranking within DFGs A and B is among the lowest. In the districts with the lowest SES ranking within DFG A (twenty-eight districts, urban and non-urban), there are 195,000 students of which 186,000 are from the poorer urban districts. In other words, the overwhelming majority of students in districts with the lowest SES ranking in the state are from the poorer urban districts.
Because these poorer urban districts in DFGs A and B become the sole object of the remedy we impose, we note their difference from the districts that have some similarities but are not included in the remedy. They fall into two categories the rural poor and some older suburban districts that have become heavily populated but are also relatively poor. These non-included districts do not have the characteristics that lead to classification as urban districts by the Department of Community Affairs or by the Commissioner: they are therefore not as poor or needy; many are not within DFG A or B, their socioeconomic status is not as compelling; furthermore, there is no direct substantive evidence on this record of their failure to provide a thorough and efficient education.
Plaintiffs offer another statistical device. In various analyses the pupils in all districts were grouped into seven approximately equal parts (the DFG classification grouped districts without regard to the number of children in each DFG). The ranking of the pupils into septiles is done on different bases, usually in accordance with expenditures per pupil. Septile 1, *344 with approximately one-seventh of the state's students, includes districts with the lowest level of expenditures per pupil. Septile 7 includes districts with the highest level of expenditures per pupil, again including approximately one-seventh of the state's students. Correlated with this septile ranking of students by expenditures per pupil is a corresponding variable, for instance, equalized valuation per pupil, the combination tending to show the relationship of the two variables. Other factors are also measured, like the concentration of minorities or the school tax rate of the district. In other instances, the septiles are arranged by equalized valuation per pupil, and similar comparisons are made. The information contained in the septile analysis is comparable to that found in the DFG comparisons, as are the conclusions arising therefrom. The septiles describe a relationship between specific factors; the DFGs between a complex generalized characteristic, socioeconomic status, and specific factors. Each tells a different story, but they add up essentially to the same conclusion: the poorer the district and the greater its need, the less the money available, and the worse the education.
First of all, between the highest and lowest DFGs the disparity and the apparent relationship are vivid. In terms of NCEB, while the state average is $3,329 per pupil (in 1984-85), the average expenditure per pupil in DFG A is $2,909; in DFG J, $4,154; in DFGs A and B, $2,861, and in DFGs I and J, $4,029. Thus, the average total NCEB of a 10,000 pupil district in DFG A or B is $11.7 million lower than it would be in a DFG I or J district of the same size. As for the relationship in terms of property wealth, while the state average equalized valuation per pupil in 1984-85 was $190,401, the average in DFG A was $78,222, in DFG J, $360,101; the average in DFGs A and B was $94,998, and in DFGs I and J, $302,593. The septile analysis produces substantially the same results: in Septile 1, arranged by NCEB per pupil and containing about 160,000 students, the range of expenditures (measured by NCEB) in 1984-85 runs from $932 to $2,634 per pupil; and Septile 7, from $4,055 to *345 $10,103 per pupil; Septile 1 (again, arranged by NCEB) has an average equalized valuation per pupil of $88,907, Septile 7, $360,359. The twenty-eight poorer urban districts (excluding Atlantic City, which has a tax base in excess of the GTB amount) have an average per pupil expenditure (measured by NCEB) in 1984-85 of $2,880 and an average equalized valuation per pupil in 1984-85 of $63,066; again, compared to DFGs I and J the comparable figures are $4,029 and $360,101, and compared to Septile 7 $4,432 and $360,359. No sophisticated analysis can destroy the conclusion: the richer districts spend more than the poorer, and their ability to do so is strongly correlated to their wealth.
The State does not challenge these disparities, although it claims that no relationship to equalized valuation per pupil has been proved. That claim is based largely on an analysis of the entire spectrum of districts as well as on the fact that when all total expenditures are considered not just equalization aid, but categorical and federal aid as well, the disparities and the relationships diminish substantially. If one accepts the premise of the State's analysis, its conclusions follow to some degree. As noted above, however, we have concluded that federal aid should not be considered in determining the extent to which funding disparities and their relationships bear on the question of whether New Jersey is providing its students their constitutionally mandated education.
The State further claims that the most intensive study conducted by plaintiffs, including the study of funding disparities, is limited to comparisons between plaintiffs' districts (Camden, East Orange, Jersey City and Irvington) and three others Newark, Paterson, and Trenton, all contrasted with certain richer suburban districts (Paramus, Princeton, Millburn, Scotch Plains-Fanwood, Livingston, South Brunswick, Cherry Hill, Moorestown, Montclair, and South Orange-Maplewood). We agree that plaintiffs' most complete study of the characteristics of districts, including substantive education, was made in the form just mentioned. But when it came to funding disparities *346 and their relationships, plaintiffs' comparisons were across the board.
These responses by the State are substantial when weighed against plaintiffs' claims for relief, including what amounts to a complete overhaul of our system of financing education. We have rejected those claims primarily because of the lack of proof, in terms both of funding and substantive education, as to the vast majority of New Jersey districts. When considered against plaintiffs' claim of lack of a thorough and efficient education in poorer urban districts, however, the State's contentions of weakness in the plaintiffs' proofs are unpersuasive. While it is true that plaintiffs' proofs were the fullest as to the comparison districts mentioned above, they were still adequate across a sufficient number of poorer urban districts. Disparity of funding, its relationship to poverty, the critical needs educational and otherwise of its pupils, the practical inability to raise further funds through taxation (municipal overburden), the likelihood of the permanence of these factors, the level of substantive education actually being given, the failure rate of its students, their dropout rate, were all sufficiently shown, and dramatically contrasted with the situation of students in richer *347 districts. And even when federal and categorical aid are included, the disparities in this range, not just the extremes of the comparison districts, but the disparities between the poorer urban districts and the richer suburban districts are still significant: in 1984-85, the average day school expenditures per pupil in DFGs A and B was $3,751, while in DFGs I and J, the average was $4,493. When the comparison is made, as we conclude it should be, without federal aid, and if categorical aid is excluded on the theory that it is intended to benefit neither richer nor poorer districts, but rather, like transportation aid, intended to address special needs in all districts it is dramatic ($2,861 compared to $4,029 per pupil). This disparity relates to what is generally considered the figure (NCEB) that best indicates the "regular education" afforded a student.
We note the obvious: the State may be correct in its contention that measured against the entire spectrum of districts in the state, the disparities and their relationships are not sufficient to condemn the system, because they exist significantly only when the extremes are compared. That gives little comfort to those students confined to the poorest districts, and they number in the hundreds of thousands (in 1984-85 there were 280,081 resident students in the twenty-nine poorer urban districts, about 25% of the entire public school population). Their deprivation is real, of constitutional magnitude, and not blunted in the least by the State's statistical analysis.C. Substantive Educational Opportunity: The Administration of the Act by the Commissioner and the Board
In Robinson I the record was primarily limited to the funding scheme, its impact, and demographic data; it lacked significant evidence of substantive educational content. We ruled that without such evidence, our constitutional determination must *348 necessarily be based solely on the disparity of expenditures per pupil. Repeating the quotation from Robinson I:The trial court found the constitutional demand had not been met and did so on the basis of discrepancies in dollar input per pupil. We agree. We deal with the problem in those terms because dollar input is plainly relevant and because we have been shown no other viable criterion for measuring compliance with the constitutional mandate. [62 N.J. at 515-16, 303 A.2d 273 (emphasis supplied).]
The State claims there is now such a "viable criterion" for measuring thorough and efficient; indeed, that is the heart of the State's case when it so strenuously opposes the significance of plaintiffs' disparity measurements. It claims simply that a thorough and efficient education in fact exists regardless of the disparity of expenditures.
The proof of substantive educational content has several sources: the Act itself, the rules and regulations of the Board and the Commissioner, the Commissioner's implementing actions, and the evidence of the education that actually takes place in the district through curriculum plans, course offerings, studies, reports, evaluations, and observations. These proofs differentiate this case from Robinson I.
The Act and the rules of the Board together constitute a thoughtful, well-integrated definition of thorough and efficient, described both abstractly and in concrete terms. The standards and goals, when considered in conjunction with the mandated procedures, constitute a potentially practical, workable mechanism for achieving a thorough and efficient education throughout the state.
The Act defines thorough and efficient education on a statewide basis as including: (a) establishment of educational goals at both the state and local levels; (b) encouragement of public involvement in goal-setting; (c) instruction intended to produce the attainment of reasonable levels of proficiency in the basic communications and computational skills; (d) a breadth of program offerings designed to develop the individual talents and abilities of pupils; (e) programs and supportive services for all pupils especially those who are educationally disadvantaged or *349 who have special educational needs; (f) adequately equipped, sanitary, and secure physical facilities and adequate materials and supplies; (g) qualified instructional and other personnel; (h) efficient administrative procedures; (i) an adequate State program of research and development; and (j) evaluation and monitoring programs at both the state and local levels. N.J.S.A. 18A:7A-5. Local agencies have the responsibility to establish goals and objectives consistent with legislative guidelines, and to define standards of performance necessary to indicate achievement of the goals and objectives. N.J.S.A. 18A:7A-2b(3)(a)-(b). While necessarily general, the Act has fairly established the standard against which an educational opportunity can be measured to see if it conforms to the constitutional command. We found the standard constitutional in Robinson V, and reaffirm that conclusion now. The plaintiffs do not attack the statutory definition and its formulation of goals and standards, or the statutory provisions for its implementation. Their complaint, rather, is that without adequate funding, it simply cannot be achieved; that the administration of the Act in any event does not assure that it will be achieved; that in fact it has not been achieved.
The Act, having defined the elements, goals, and standards of thorough and efficient statewide, and the concomitant structure of goal setting and planning at the local level, then goes on to provide a mechanism for its effectuation. The Commissioner and the Board are charged with the responsibility for achieving the constitutional mandate. Both the Board and Commissioner have done their work with energy and dedication, and have achieved considerable success in view of the limited time they have had to work under the Act, and the limited resources available to them.
Pursuant to statute and regulations the Commissioner requires each district to file an annual report covering budgets, curriculum, and other matters concerning the district's operation and education offerings. N.J.S.A. 18A:7A-11; N.J.A.C. 6:8-3.2. Those reports are reviewed by county superintendents *350 for facial compliance with budget requirements, and, once every five years in the case of certified districts, by special teams that visit the district. This oversight represents a radical change from a system in which districts were almost totally independent of any State involvement, except for a few courses mandated by statute. The transformation of the system to one with an increasingly significant State role has not been easy.
Since 1983-84, the DOE has based this monitoring on ten elements of a thorough and efficient education and a number of indicators, originally fifty-one, now forty-three, of acceptable performance in these elements.N.J.A.C. 6:8-4.3. Certification of a district is based on an acceptable rating of all forty-three indicators. N.J.A.C. 6:8-4.5. If the county superintendent finds that the district has met this standard, a recommendation for certification and a summary report of the findings is submitted to the Commissioner, who then recommends that the Board certify the district as providing a thorough and efficient education.N.J.S.A. 18A:7A-14; N.J.A.C. 6:8-4.5. Among the forty-three prescribed indicators is the requirement that 75% of the ninth grade students of each school pass all three sections of the HSPT. Moreover, the failure of a certified school district to meet the State HSPT standard is cause for rescinding its certification. N.J.A.C. 6:8-4.6.
If the Board does not certify a district, it becomes subject first to a "Level II review process" (N.J.S.A. 18A:7A-14(a), (b); *351 N.J.A.C. 6:8-5.1), and failing that, to "Level III corrective action" by the State.N.J.S.A. 18A:7A-14(b), (c), (d); N.J.A.C. 6:8-5.2. Under Level II review, a district is required to develop a self-improvement plan, which is submitted to the district board of education and the county superintendent of schools. When the superintendent, after reviewing the plan with the assistant commissioner, Division of County and Regional Services, approves it, progress is monitored and interim reviews are conducted at least once every three months. The reviews may consist of reviewing the chief school administrator's reports, of on-site visits, or both. N.J.A.C. 6:8-5.1(3)(b)(e).
If a district fails to achieve certification as a result of its improvement plan activities, the county superintendent initiates Level III corrective action. N.J.S.A. 18A:7A-14(c); N.J.A.C. 6:8-5.2. An external review team is appointed from qualified staff of other districts supplemented by the DOE's compliance unit. The team assesses the reasons for the district's failure and recommends implementation of a corrective action plan or, if conditions warrant, of a comprehensive compliance investigation. If the former is undertaken, the district has a year to implement the new plan, or at least demonstrate reasonable progress toward correcting its deficiencies, with the county superintendent monitoring progress on a monthly basis. N.J.A.C. 6:8-5.2(d)(e). Pursuant to such corrective action, the Commissioner may reallocate funds within the district's budget, or order a budget increase. N.J.S.A. 18A:17-14(d), -15; see also Robinson V, supra, 69 N.J. at 461-63, 355 A.2d 129 (discussing this power).
If the district fails to demonstrate adequate progress or the external review team's report indicates that conditions preclude successful implementation of a corrective plan, the Commissioner must order a comprehensive compliance investigation (N.J. *352 S.A. 18A:7A-14(d)), including full examination by a team from the DOE's compliance unit and a comprehensive audit by a private firm hired for that purpose. N.J.A.C. 6:8-5.3. After a plenary hearing before an ALJ, the Commissioner may issue an administrative order requiring compliance with a departmental correction plan. N.J.S.A. 18A:7A-14(e). If even these steps fail, he must recommend to the Board that it create a State-operated school district. N.J.S.A. 18A:7A-15.
The Commissioner has exercised those powers on occasion in Trenton, for instance, the Commissioner, pursuant to an administrative order, appointed a monitor with the power to direct the district's personnel to take all steps necessary to operate a thorough and efficient program, to review all personnel staffing recommendations, and to oversee all budgetary matters. In re Trenton Bd. of Educ., 176 N.J. Super. 553, 559-60, 424 A.2d 435 (App.Div. 1980), aff'd, 86 N.J. 327, 431 A.2d 808 (1981). In Jersey City, the Board initiated a full take-over, appointing a state district superintendent to supplant the local district board of education and direct all operations of the district.
Notwithstanding the apparent forcefulness of these interventions, the ALJ concluded that their administration by the Board and Commissioner was seriously deficient. Most importantly, he concluded that, apart from the HSPT and similar standardized tests, the Level I and Level II process amounted to little more than a comparison of the district's goals and plans with the district's own statement of its accomplishments in achieving those goals and plans all without any independent assessment by the Commissioner of the adequacy of the goals or the quality of the education provided. Neither the Commissioner nor his agents compared the education offered by the district *353 with that offered by others. There was neither an absolute standard nor a standard implied through comparisons.
We conclude that although the monitoring function may have been designed to measure and achieve a thorough and efficient education, in practice it has not accomplished that goal. In part because resource issues were avoided, it operated largely as a self-improvement system. Beyond a few statemandated courses, the local board could approve any curriculum it chose or, presumably, could afford. The Commissioner evaluated neither its adequacy to the children's needs, nor its relationship to a thorough and efficient education. Nor did he evaluate the quality of any offering. When monitors visited classes, they did so to verify that the teacher was using a local board-approved lesson plan.
We cannot accept the Commissioner's and the Board's contention that certification is prima facie proof of a thorough and efficient education. It is not that we ascribe no importance to it, it is simply not enough. However, we need not decide whether any significance attaches to the lack of certification because regardless of certification or its lack, we find the constitutional mandate has not been satisfied in poorer urban districts.
Indeed, the Board itself has acknowledged that the design of this scheme is deficient, and in this very litigation has proposed and adopted regulations to strengthen its monitoring and assessment functions. State Board of Education, Docket No. 12-89, at 54-55 (March 23, 1989). It has also requested that the *354 Legislature shift its aid formula to reflect current, rather than prior, year funding; appropriate more money for school facilities; and ensure that minimum aid to richer districts is redistributed if there is a shortfall in appropriations. Id. at 55-57.
For the overwhelming number of districts, all we know about the substantive education is that the Act commands it, reports are produced, monitoring takes place, and the Commissioner has certified most districts as providing a thorough and efficient education; we have no evidence that they are not. While we know of expenditure disparity, we have no reason to believe the actual funding level in most districts is below that needed to achieve a thorough and efficient education. On this record plaintiffs have not shown that a thorough and efficient education does not indeed exist right now in the overwhelming number of districts in this state. Under those circumstances our State Constitution does not require that this entire system be scrapped and replaced by another at enormous expense based solely on expenditure disparities. We do not imply anything about the importance of such disparities when considered as a matter of policy. As noted at the outset of this opinion, our decision does not deal with optimum educational policy, but with constitutional compliance.
We are unable to conclude that most districts are failing to deliver the educational opportunity required by our State Constitution. There are various elements to that conclusion. In a narrow procedural sense, it reflects our belief that the burden is on the plaintiffs to show that a thorough and efficient education is not being delivered; it represents our conclusion that when the State, through the Legislature and its administrative agency, has conscientiously attempted to achieve a level of education and the State's agency says it has been achieved, some deference must be accorded to that determination. It further embodies our conclusion that all of these factors are sufficient evidence to preclude a finding of constitutional deprivation based solely on expenditure disparity. Most of all it reflects our firmly held belief that before this Court concludes *355 that there is a constitutional failure despite the legally-authorized certification of the Commissioner to the contrary, the Board's conclusion to the contrary, and the Legislature's efforts to achieve it, the proofs must be compelling. Before this Court voids the statute, overrules the Board and the Commissioner, and orders the Legislature to provide a new system, the constitutional failure must be clear. As to most districts of the state, it is not. But as to some the poorer urban districts it is glaringly clear.1. Municipal Overburden
"Municipal overburden" is the excessive tax levy some municipalities must impose to meet governmental needs other than education. It is a common characteristic in poorer urban districts, a product of their relatively low property values against which the local tax is assessed and their high level of governmental need. The governmental need includes the entire range of goods and services made available to citizens: police and fire protection, road maintenance, social services, water, sewer, garbage disposal, and similar services. Although the condition is not precisely defined, it is usually thought of as a tax rate well above the average.
The underlying causes of municipal overburden are many and complex. Its consequences in this case, however, are clear and simple. The poorer urban school districts, sharing the same tax base with the municipality, suffer from severe municipal overburden; they are extremely reluctant to increase taxes for school purposes. Not only is their local tax levy well above average, so is their school tax rate. The oppressiveness of the tax burden on their citizens by itself would be sufficient to give them pause before raising taxes. Additionally, the rates in some cases are so high that further taxation may actually decrease tax revenues by diminishing total property values, either directly because of the tax-value relationship, or indirectly by causing business and industry to relocate to another municipality.
*356 Plaintiffs argue that municipal overburden restricts these poorer urban districts' ability to raise more money through increased school taxes to achieve a thorough and efficient education. That contention is urged in support of their claim that the present system, relying on local levies, must be converted to one where state revenues predominate and local taxing ability is relatively unimportant. It is an argument of constitutional dimension: it responds to the State's claim that a thorough and efficient education can be achieved under the Act through the unlimited power of school districts to increase education revenues by raising taxes, and it supports plaintiffs' preferred constitutional remedy.
The Commissioner's position is that municipal overburden is irrelevant to the constitutional issue. He maintains that the school board is obliged as a matter of statutory and constitutional duty to raise as much money as is needed for a thorough and efficient education no matter what the tax consequences, no matter how much the tax increase, no matter how high the present tax, and no matter how high the present local municipal tax. If tax inequity exists, that may be a concern for the Legislature but it must not be for the school district. The Commissioner's views are set forth strongly and unambiguously. He portrays school boards, in even the poorest urban districts, that refuse to substantially increase taxes that already border on the confiscatory, as shirking their clear duty and, in some districts, as yielding to political pressure. His position rests on his conception of the Act as one with funding provisions that make it certain that enough money for a thorough and efficient education will be raised. That conclusion, he says, is irresistible given the districts' unlimited power to tax, their duty to do so, and the power of the Board and the Commissioner to force them to if needed.
We respectfully disagree. The social and economic pressures on municipalities, school districts, public officials, and citizens of these disaster areas many poorer urban districts are so severe that tax increases in any substantial amount are almost *357 unthinkable. To pejoratively label that uniform resistance to increased taxes as "political" fails to recognize the desperate situation that some school boards may face.
The record is clear on this issue. The former Commissioner recognized that the refusal of such districts to raise taxes was strongly based on legitimate concerns. Similarly, the present Commissioner has (with one possible exception) never compelled a tax increase during the period of his eight year tenure.
We do not pass on the legal contentions of the Commissioner in this respect. Our conclusion concerning municipal overburden is that it effectively prevents districts from raising substantially more money for education. It is a factual conclusion. It is based on the record history of the failure of any effort, whether through the actions of a school district or the Board or Commissioner, legal or otherwise, through the more than ten year period that the Act has been in effect, to achieve substantially increased local funding through school tax increases. That factual finding is one of the bases for our conclusion that the funding mechanism of the Act will never achieve a thorough and efficient education because it relies so heavily on a local property base already over-taxed to exhaustion.D. The Quality of Education in the Poorer Urban Districts
The primary basis for our decision is the constitutional failure of education in poorer urban districts. The record demonstrates beyond debate that a thorough and efficient education does not exist there. Our conclusion that the constitutional mandate has not been satisfied is based both on the absolute level of education in those districts and the comparison with education in affluent suburban districts.
Plaintiffs' proofs of the significantly inferior quality of education in poorer urban districts are persuasive. While exceptions exist, at the extremes and its strength is limited to the extremes the comparison between the education offered to *358 students in poorer urban districts with that offered in the richer districts is impressive. The characteristics of a substantive education are most difficult to prove; short of intensive examination of education in progress, at the school, in the classroom, proofs are necessarily circumstantial. The State did not insist that the only true measure of substantive education was on-the-scene observation, but rather that even accepting plaintiffs' evidence at face value, it fell short. The State's objections were of various kinds: it noted that the comparisons were largely limited to the extremes, the richest against the poorest, the very best against the very worst; evidence was lacking in most cases that would warrant a reliable comparative conclusion. Furthermore, the State claims that the adequacy of the education that was being afforded in poorer districts was not acknowledged. The State's basic objection is that the various circumstantial measures, such as course offerings, experience and education of the staff, and pupil/staff ratio cannot be considered reliable indicators of the quality of education.
In this connection we note the ALJ's description of the issue as whether a thorough and efficient education is measured by equal finances and programs, as contended by plaintiffs, or by student achievement, as contended by the Commissioner and the Board. While we have continued to adhere to our prior decisions that equalization is constitutionally required only up to a certain level that necessary to achieve a thorough and efficient education and that districts may exceed that level, we do not foreclose the possibility that changing circumstances, including future development of education in this state, may lead to an interpretation of the constitutional obligation as requiring such equality of funding. Concerning the poorer urban districts, however, we have found a constitutional failure of education no matter what test is applied to determine thorough and efficient.
The State's position that thorough and efficient should be measured against the Act and the goals adopted by the Board rather than by "input," i.e., course offerings, teacher/pupil *359 ratios, and expenditures per pupil, suggests the inherent difficulties in determining the scope of the constitutionally mandated education. For most of the goals adopted by the Board, there are no tests to determine their accomplishment. With only standardized tests that measure but a few of them, we are left without any feasible method of applying the State's measure of thorough and efficient.
We note initially that there was little direct proof of substantive education, such as course offerings, for the overwhelming number of districts in the state. On this record, most of the state's districts except the poorer urban ones may be offering an education in full compliance with the constitutional mandate.
However, the level of education offered to students in some of the poorer urban districts is tragically inadequate. Many opportunities offered to students in richer suburban districts are denied to them. For instance, exposure to computers is necessary to acquire skills to compete in the workplace. In South Orange/Maplewood school district, kindergarteners are introduced to computers; children learn word processing in elementary school; middle school students are offered beginning computer programming; and high school students are offered advanced courses in several programming languages or project-oriented independent studies. Each South Orange/Maplewood school has a computer lab.
By contrast, many poorer urban districts cannot offer such variety of computer science courses. While Princeton has one computer per eight children, East Orange has one computer per forty-three children, and Camden has one computer per fifty-eight children. Camden can offer formal computer instruction to only 3.4% of its students. In many poorer urban districts, computers are purchased with federal or state categorical funds for use in remedial education programs. Paterson offers no computer education other than computer-assisted basic skills programs. Further, many of these districts do not have sufficient *360 space to accommodate computer labs. In Jersey City, computer classes are being taught in storage closets.
Science education is deficient in some poorer urban districts. Princeton has seven laboratories in its high school, each with built-in equipment. South Brunswick elementary and middle schools stress hands-on, investigative science programs. However, many poorer urban districts offer science classes in labs built in the 1920's and 1930's, where sinks do not work, equipment such as microscopes is not available, supplies for chemistry or biology classes are insufficient, and hands-on investigative techniques cannot be taught. In Jersey City and Irvington, middle school science classes are taught without provision for laboratory experience. In East Orange middle schools, teachers wheel a science cart into a three-foot-by-six-foot science area for instruction. The area contains a sink, but no water, gas, or electrical lines.
The disparity in foreign-language programs is dramatic. Montclair's students begin instruction in French or Spanish at the pre-school level. In Princeton's middle school, fifth grade students must take a half-year of French and a half-year of Spanish. Most sixth graders continue with one of these languages. Many begin a second language in the ninth grade, where four-year programs in German, Italian, Russian, and Latin are offered. French and Spanish are offered on two tracks, one for students who began instruction in middle school and the other for those who begin in the ninth grade. Advanced placement courses are available. In contrast, many of the poorer urban schools do not offer upper level foreign language courses, and only begin instruction in high school. Jersey City starts its foreign language program in the ninth grade; Paterson begins it at the tenth grade. Most Jersey City high schools offer only two languages; both of Paterson's high schools offer only Spanish and French, although the two Paterson high schools share one German teacher and one Latin teacher. *361 Music programs are vastly superior in some richer suburban districts. South Brunswick offers music classes starting in kindergarten; Montclair begins with pre-schoolers. Millburn and South Brunswick offer their middle school students a music curriculum that includes courses such as guitar, electronic-piano laboratory, and music composition on synthesizers. Princeton offers several performing groups, including bands, choruses, and small ensembles. However, Camden and Paterson do not offer a music course until the fourth grade; only introductory level music courses are offered in high school. In 1981, Camden eliminated all its elementary school music teachers and provided "helpers" to assist in teaching music. Many poorer urban school districts have inadequate space for instrumental music lessons, bands, and choruses. In one elementary school in Jersey City, instrumental music lessons are provided in the back of the lunchroom. At lunchtime, the class moves to an area in the school's basement.
Art programs in some poorer urban districts suffer compared to programs in richer suburban districts. In Montclair, the art program begins at the pre-school level; there is an art teacher in every elementary school; every school has at least one art room; and the district has purchased a variety of art equipment, such as a kiln for ceramic artwork. In contrast, art programs in some poorer urban districts are sparse. There are no art classrooms in East Orange elementary schools, and art teachers, who must travel from class to class, are limited in the forms of art they can teach. Jersey City has an excellent art program for gifted children; however, the regular art program can now accommodate only 30% of the district's students.
In South Brunswick school district, the industrial-arts program includes an automotive shop, a woodworking shop, a metal shop, a graphics shop, and a greenhouse for a horticultural course. The vocational education program has a computer drafting laboratory and a graphics laboratory with a darkroom. In Camden, state-of-the-art equipment is not purchased; the old equipment in the classrooms is not maintained or repaired. *362 There have even been problems heating the industrial-arts wing of the school.
Physical education programs in some poorer urban districts are deficient. While many richer suburban school districts have flourishing gymnastics, swimming, basketball, baseball, soccer, lacrosse, field hockey, tennis, and golf teams, with fields, courts, pools, lockers, showers, and gymnasiums, some poorer urban districts cannot offer students such activities. In East Orange High School there are no such sports facilities; the track team practices in the second floor hallway. All of Irvington's elementary schools have no outdoor play space; some of the playgrounds had been converted to faculty parking lots. In a middle school in Paterson, fifth- and sixth-graders play basketball in a room with such a low ceiling that the net is placed at the level appropriate for third-graders.
Many of these poorer urban districts are burdened with teaching basic skills to an overwhelming number of students. They are essentially "basic skills districts." In 1985, 53% of Camden's children received remedial education; in East Orange, 41%; in Irvington, 30%. By contrast, only 4% of the students in Millburn school district received remedial education.
A thorough and efficient education also requires adequate physical facilities. N.J.S.A. 18A:7-5f. We held in Robinson I that "[t]he State's obligation includes ... capital expenditures without which the required educational opportunity could not be provided." 62 N.J. at 520, 303 A.2d 273. The Legislature's appropriations for renovation of deteriorating school buildings and construction of new facilities, although substantial, do not approach the estimated $3 billion needed for a complete upgrade of the school facilities in this state.
Many poorer urban districts operate schools that, due to their age and lack of maintenance, are crumbling. These facilities do not provide an environment in which children can learn; indeed, the safety of children in these schools is threatened. For example, in 1986 in Paterson a gymnasium floor collapsed in *363 one school, and in another school the entire building was sinking. According to East Orange's long-range facility plan there are ten schools in immediate need of roof repair, fifteen schools with heating, ventilation or air conditioning problems; two schools that need total roof replacement; nine with electrical system problems; eight with plumbing system problems; thirteen needing structural repairs; seventeen needing patching, plastering or painting; and thirteen needing asbestos removal or containment.
In an elementary school in Paterson, the children eat lunch in a small area in the boiler room area of the basement; remedial classes are taught in a former bathroom. In one Irvington school, children attend music classes in a storage room and remedial classes in converted closets. At another school in Irvington a coal bin was converted into a classroom. In one elementary school in East Orange, there is no cafeteria, and the children eat lunch in shifts in the first floor corridor. In one school in Jersey City, built in 1900, the library is a converted cloakroom; the nurse's office has no bathroom or waiting room; the lighting is inadequate; the bathrooms have no hot water (only the custodial office and nurse's office have hot water); there is water damage inside the building because of cracks in the facade; and the heating system is inadequate.
In contrast, most schools in richer suburban districts are newer, cleaner, and safer. They provide an environment conducive to learning. They have sufficient space to accommodate the childrens' needs now and in the future. While it is possible that the richest of educations can be conferred in the rudest of surroundings, the record in this case demonstrates that deficient facilities are conducive to a deficient education.
Thorough and efficient means more than teaching the skills needed to compete in the labor market, as critically important as that may be. It means being able to fulfill one's role as a citizen, a role that encompasses far more than merely registering to vote. It means the ability to participate fully in society, *364 in the life of one's community, the ability to appreciate music, art, and literature, and the ability to share all of that with friends. As plaintiffs point out in so many ways, and tellingly, if these courses are not integral to a thorough and efficient education, why do the richer districts invariably offer them? The disparity is dramatic. Alongside these basic-skills districts are school systems offering the broadest range of courses, instruction in numerous languages, sophisticated mathematics, arts, and sciences at a high level, fully equipped laboratories, hands-on computer experience, everything parents seriously concerned for their children's future would want, and everything a child needs. In these richer districts, most of which have some disadvantaged students, one will also find the kind of special attention and educational help so badly needed in poorer urban districts that offer only basic-skills training. If absolute equality were the constitutional mandate, and "basic skills" sufficient to achieve that mandate, there would be little short of a revolution in the suburban districts when parents learned that basic skills is what their children were entitled to, limited to, and no more.
The State contends that the education currently offered in these poorer urban districts is tailored to the students' present need, that these students simply cannot now benefit from the kind of vastly superior course offerings found in the richer districts. No one claims here, however, that students unable to attain a level of reading, writing, or expression even approaching the expectations of their grade, pupils who, according to plaintiffs, are two years behind others on the first day they enter school, would be able to take full advantage of the richness of course offerings found in the wealthier suburbs. The State's conclusion is that basic skills are what they need first, intensive training in basic skills. We note, however, that these poorer districts offer curricula denuded not only of advanced academic courses but of virtually every subject that ties *365 a child, particularly a child with academic problems, to school of art, music, drama, athletics, even, to a very substantial degree, of science and social studies. The result violates not only our sense of what constitutes a thorough and efficient education, but the statute as well, which requires "[a] breadth of program offerings designed to develop the individual talents and abilities of students." N.J.S.A. 18A:7-5d. However articulated, such a requirement must encompass more than "instruction ... in the basic communications and computational skills," which the statute cites as another major element in education. N.J.S.A. 18A:7-5c.
In saying this we disparage neither these districts' decision to focus on remedial training, nor the State testing requirements that may have prompted this focus. But constitutionally, these districts should not be limited to such choices. However desperately a child may need remediation in basic skills, he or she *366 also needs at least a modicum of variety and a chance to excel.
Equally, if not more important, the State's argument ignores the substantial number of children in these districts, from the average to the gifted, who can benefit from more advanced academic offerings. Since little else is available in these districts, they too are limited to basic skills.
The level of substantive education is proven by plaintiffs through other indicators. Plaintiffs have selected what are sometimes regarded as strong indicators of educational quality, and have measured them among districts. Teacher ratios (the number of teachers per 1,000 pupils), the average experience of instructional staff, their average level of education, all have been documented in chart after chart.
*367 As to each one of these indicators, the poorer urban districts suffer by comparison to the rich. Indeed, although the incremental showing is far from dramatic, teacher ratios, experience, and education consistently improve as the districts' property wealth, per pupil expenditure, socioeconomic status or other similar factor improves. For instance, when districts are ranked by socioeconomic status (SES), the percentage of teachers with advanced degrees rises from 29% in the lower SES districts to 52% in the higher; teachers' average experience rises from 12 years in the lower SES districts to 15 years in the higher; and the ratio of teachers to pupils rises from 61 to 68 teachers per 1,000 students. There are exceptions, numerous when considered alone, but not significant enough to rebut the truth of the generality. Here we deal only with disparity we do not find that one instructor per fifteen students, twenty students, or thirty students is necessary for a thorough and efficient education. Nor do the experts even agree on the significance of the quantity of staff, the experience of staff, or the staffs' educational background. We are satisfied, however, that these indicators support the conclusion that the absolute quality of education in the poorer urban districts is deficient.
The relationship between poverty and these indicators of the quality of education has been shown. In view of that relationship, other conclusions concerning educational quality emerge. The costs of changing teacher ratios, increasing average teacher experience, teachers' educational background and, of course, increasing average staff salary, are staggering. For instance, a large urban district with 8,000 pupils and a staff ratio of one *368 teacher per thirty pupils would have to budget an extra $980,500 per year to bring that ratio down to one per twenty-five (assuming the teacher salary was the statutory minimum, $18,500, N.J.S.A. 18A:29-5). Improvement of each of the other factors has a similarly high cost attached to it. We return to plaintiffs' insistent and persuasive question: if these factors are not related to the quality of education, why are the richer districts willing to spend so much for them?
In summary, although the evidence of substantive educational level is insufficient to prove systemic lack of thorough and efficient, it is more than sufficient to prove the constitutional deficiency in a limited number of districts. Furthermore, the characteristics of those districts are such as to warrant the inference that education in similar districts the twenty-nine poorer urban districts is similarly deficient, an inference buttressed by independent evidence of the level of education in those other districts instructors' education and experience, teacher ratios, and per pupil expenditures. While we have no direct evidence that the breadth of the education offered in those wealthier suburban districts whose course offerings and other educational characteristics were placed on the record extends to other districts, again it is a fair inference that many wealthy districts have a similarly high level of education. Disparity exists, therefore, between education in these poorer urban districts and that in the affluent suburban districts; it is severe and forms an independent basis for our finding of a lack of a thorough and efficient education in these poorer urban districts these students simply cannot possibly enter the same market or the same society as their peers educated in wealthier districts.*369 E. The Quality of Students' Needs in the Poorer Urban Districts
This record shows that the educational needs of students in poorer urban districts vastly exceed those of others, especially those from richer districts. The difference is monumental, no matter how it is measured. Those needs go beyond educational needs, they include food, clothing and shelter, and extend to lack of close family and community ties and support, and lack of helpful role models. They include the needs that arise from a life led in an environment of violence, poverty, and despair. Urban youth are often isolated from the mainstream of society. Education forms only a small part of their home life, sometimes no part of their school life, and the dropout is almost the norm. There are exceptions, fortunately, but substantial numbers of urban students fit this pattern. The goal is to motivate them, to wipe out their disadvantages as much as a school district can, and to give them an educational opportunity that will enable them to use their innate ability.
In 1985-86, every district in DFG A and all but two districts in DFG B failed to meet the State standard for the High School Proficiency Test (HSPT). Moreover, every poorer urban district in DFGs A and B failed to meet this standard. These tests do not purport to measure or define a thorough and efficient education. They are designed to show mastery of basic skills and, while upgraded from the past, essentially measure the minimum level of learning needed to go on to more difficult subjects. The Legislature regards these tests as a prerequisite to, not an equivalent of, a thorough and efficient education. The failure rate of these poorer urban districts on even this minimal test, the depth of that failure, testifies eloquently not just about their inadequate performance, but about their need. The shocking contrast to the performance of students from richer suburban districts completes the picture.
The Commissioner has determined that satisfactory performance on this test requires a passing score by 75% of the *370 district's students in grade nine of each school in each of the three categories. In 1985-86, of more than 14,000 ninth graders in school districts in DFG A who took the HSPT, only 54% passed the reading test, 42% passed the math test, and 43% passed writing. The poorer urban districts in DFG A did even worse. In Newark, for example, only 41% of ninth graders who took the test passed reading, 31% passed math, and 39% passed writing. In Camden, 36% passed reading, 28% passed math, and 44% passed writing. By contrast, in school districts in DFG J, of 5,400 ninth graders tested, 97% passed reading, 93% passed math and 95% passed writing. Statewide, 83% of students tested passed reading, 72% passed math, and 77% passed writing.
The dropout rate in these poorer urban districts is further testimony both to their failure and to the students' needs. The "unofficial" dropout rate (1984-85) for some urban high schools can be as high as 47%. A district cannot deliver a thorough *371 and efficient education to a dropout. For a multitude of tragic reasons, these students lack the most basic requirement for achieving a thorough and efficient education the will to learn. That characteristic is assumed, accepted, a given, in richer suburban districts.
The record evidence of the quality of education in poorer urban districts and the desperate needs of their students clearly indicates that a significantly different approach to education is required if these districts and their students are to succeed. Furthermore, there is a wealth of material outside of the record to the same effect. The nation has come to recognize the *372 education of the urban poor as a most difficult and important problem. While opinions concerning the methods, approaches, and techniques differ concerning their effectiveness, their advantages and disadvantages, there is solid agreement on the basic proposition that conventional education is totally inadequate to address the special problems of the urban poor. Something quite different is needed, something that deals not only with reading, writing, and arithmetic, but with the environment that shapes these students' lives and determines their educational needs.
Obviously, we are no more able to identify what these disadvantaged students need in concrete educational terms than are the experts. What they don't need is more disadvantage, in the form of a school district that does not even approach the funding level that supports advantaged students. They need more, and the law entitles them to more.
Many students in poorer urban districts do not have books at home. These students obviously need adequate libraries and media centers. Educators note that an adequate guidance program could give children in poorer urban districts special assistance and individual attention; that counseling services help children overcome problems associated with unwanted pregnancies, drugs, crime, or unsupportive families; that both crisis counselors and career counselors from elementary school through high school may be needed to assist students to overcome obstacles and receive a worthwhile education.
Alternative education programs for students identified as potential dropouts are suggested as necessary to motivate a substantial number of students in poorer urban districts. Several richer suburban districts provide individualized tutoring and vocational education to students in need of alternative education. However, in Jersey City, several alternative education *373 programs were eliminated, and only a program for students in legal trouble is provided; Paterson conducted a ten-student pilot program in 1985-86, then cut it the next year because of lack of funds, although the district had identified 200 students in need. Such programs may be essential in many poorer urban districts.
Other methods have been suggested for these poorer urban districts. For instance, an intensive pre-school and all-day kindergarten enrichment program to reverse the educational disadvantage these children start out with; recruitment of parents to join parent participation programs and become involved with the schools and their schoolchildren. It seems agreed that local boards of education, administrators, and teachers organizations all must join in this partnership for the benefit of these children if education in poorer urban districts is to succeed.
In Robinson I we observed that the State may "recognize ... a need for additional dollar input to equip classes of disadvantaged [students] for the educational opportunity." 62 N.J. at 520, 303 A.2d 273 (footnote omitted). This reference to students' special needs was given added content in Abbott I where we observed that "in some cases for disadvantaged students to receive a thorough and efficient education, the students will require above-average access to education resources" and that this bears "on the amount of money that a school district must be able to provide for its children." Abbott I, supra, 100 N.J. at 292, 495 A.2d 376; see Robinson V, supra, 69 N.J. at 550, *374 355 A.2d 129 (Pashman, J., dissenting). We identified the constitutional issue as being whether, "after comparing the education received by children in property-poor districts to that offered in property-rich districts, it appears that the disadvantaged children will not be able to compete in, and contribute to, the society entered by the relatively advantaged children." Abbott I, supra, 100 N.J. at 296, 495 A.2d 376.
It is clear to us that in order to achieve the constitutional standard for the student from these poorer urban districts the ability to function in that society entered by their relatively advantaged peers the totality of the districts' educational offering must contain elements over and above those found in the affluent suburban district. If the educational fare of the seriously disadvantaged student is the same as the "regular education" given to the advantaged student, those serious disadvantages will not be addressed, and students in the poorer urban districts will simply not be able to compete. A thorough and efficient education requires such level of education as will enable all students to function as citizens and workers in the same society, and that necessarily means that in poorer urban districts something more must be added to the regular education in order to achieve the command of the Constitution. Such added help is in theory afforded now through categorical aid, consisting of additional funds to address special needs, aid for such things as compensatory education, bilingual education, education for students who are developmentally disabled, or visually handicapped. The problem, however, is that this categorical aid is added to a budget that is already significantly less than the comparable budgets of richer districts. When added to that regular budget of the poorer urban district, it fails to bring even equality of expenditure dollars between districts, and certainly does not provide the help needed to address these students' disadvantages.
We realize our remedy here may fail to achieve the constitutional object, that no amount of money may be able to erase the impact of the socioeconomic factors that define and *375 cause these pupils' disadvantages. We realize that perhaps nothing short of substantial social and economic change affecting housing, employment, child care, taxation, welfare will make the difference for these students; and that this kind of change is far beyond the power or responsibility of school districts. We have concluded, however, that even if not a cure, money will help, and that these students are constitutionally entitled to that help.
If the claim is that additional funding will not enable the poorer urban districts to satisfy the thorough and efficient test, the constitutional answer is that they are entitled to pass or fail with at least the same amount of money as their competitors.
If the claim is that these students simply cannot make it, the constitutional answer is, give them a chance. The Constitution does not tell them that since more money will not help, we will give them less; that because their needs cannot be fully met, they will not be met at all. It does not tell them they will get the minimum, because that is all they can benefit from. Like other states, we undoubtedly have some "uneducable" students, but in New Jersey there is no such thing as an uneducable district, not under our Constitution.
All of the money that supports education is public money, local money no less than state money. It is authorized and controlled, in terms of source, amount, distribution, and use, by the State. The students of Newark and Trenton are no less citizens than their friends in Millburn and Princeton. They are entitled to be treated equally, to begin at the same starting line. Today the disadvantaged are doubly mistreated: first, by the accident of their environment and, second, by the disadvantage added by an inadequate education. The State has compounded the wrong and must right it.F. Impact of the Level of Funding on the Quality of Education
The State's claim that the statistical evidence fails to prove a significant relationship between education expenditures and *376 property wealth is joined with a more fundamental objection: money is not a critical factor in the quality of education in the first place. The position is not quite that extreme, of course obviously, a certain minimum amount is needed to operate an effective school system, as the Commissioner notes. But beyond that minimum amount, not clearly defined in the record, the difference, the excess, is characterized as an unreliable indicator of the quality of education, and what the money buys better staff ratios, more experienced teachers, equipment, more varied course offerings is not determinative either. The conclusion is that since all districts in the state have much more than whatever the minimum amount may be, the excess and the differences in the excess, are irrelevant to the quality of education.
The impact of this position is that if a thorough and efficient education is not being achieved, money is neither the cause nor the remedy. Furthermore, wide disparity in educational funding, between the poorer urban and richer suburban districts, does not establish failure of thorough and efficient, nor, by the State's hypothesis, does disparity of funding establish any consequent disparity in substantive education.
The State's position here is almost the reverse of one of the premises of Robinson I, that "dollar input [was] plainly relevant", 62 N.J. at 515, 303 A.2d 273, and that in the absence of other proof it was the sole determinant of the quality of education. The Commissioner attacks that proposition head-on.
We deal here with questions of educational theory debated over the years, and now debated by experts of the very highest order. These issues have come to the fore not just in this case but throughout the nation, with states struggling as we are here in New Jersey to find an answer to urban problems, especially urban education. Studies of the most sophisticated design, pilot projects, reams after reams of case histories of schools, districts, and students, learned treatises, books, television programs all directed at the same question: what produces *377 good education in urban schools? The only thing universally agreed on is that those schools are failing. After that, controversy abounds. Our observation, simplistic perhaps, that "money is only one of a number of elements [involved in education]," Robinson V, supra, 69 N.J. at 455, 355 A.2d 129, still represents most of what one can profitably glean from this controversy for the purposes of this litigation. More concretely, while we are unable to conclude from this record that the State is clearly wrong, we would not strip all notions of equal and adequate funding from the constitutional obligation unless we were convinced that the State was clearly right.
The results of all of this research, while promising and constructive, are inconclusive, at least on the underlying issue before us. It shows beyond doubt that money alone has not worked. It shows promising success in many different approaches emphasizing techniques, relationships, social forces, motivation, approaches often quite different from conventional instruction. But it does not show that money makes no difference. What it strongly suggests is that money can be used more effectively than it is being used today. The inconclusiveness of the research is conceded, at least to the extent of admitting that although money is not the main determinant of the quality of education no one is quite convinced what is, nor totally confident about what works. The underlying subject matter is most complex, and the experts note the need for studies of much greater duration than theirs. Viewing it from that perspective, despite their sophisticated designs, the studies are characterized by the ALJ as "relatively primitive."
The Commissioner has adopted as his preferred approach to education that known as "effective schools." The *378 methods and underlying concepts incorporated in the "effective schools" approach seem logical and practical. It is asserted to have the additional advantage of being inexpensive. The evidence that it works is encouraging. The Commissioner is trying to implement such methods throughout the state, in particular in Jersey City. They are, as noted later, specifically targeted to the urban areas. There is no suggestion that they can either be put in place or have their effect on students in a short time.
We agree with the ALJ that no matter how promising "effective schools" may be, the proofs fall short of any showing that at present expenditure levels they will lead to a thorough and efficient education in poorer urban districts. Furthermore, we *379 find nothing in the record, nor did the Commissioner attempt to prove, that even with implementation of "effective schools," educational disparity will not remain. We are willing, for the purposes of this case, to assume that if successful, effective schools at present funding levels would diminish the gross disparity in substantive education and pupil performance now found between poorer urban and richer suburban districts; by that, however, we mean that it would diminish, not eliminate, gross disparities. While applicable to education in all districts, the underlying theory seems to be that these would be particularly helpful techniques in poorer urban districts. The determinants, the "effective schools" approach, address many of the factors that pervade and define the low socioeconomic status that characterizes poorer urban districts: lack of community and parental involvement, consequent lack of motivation, and low expectations of student success. In short, "effective schools" attempts to redress the disparity that life and its environment bring to most of the students in poorer urban districts, the underlying disparity that is thought by many to be the major factor in the failure of education in these areas.
The Commissioner has concluded that the "effective schools" program will significantly increase the possibility that students in poorer urban districts will be able to realize their potential. If he is correct, they will, as a result, be better equipped as citizens and workers, but they may still be considerably disadvantaged as they try to enter a society and work force occupied by pupils from the richer districts. Not even this amount of amelioration is certain, and just putting it in place and testing it will take years, perhaps the better part of the school life remaining for some of those disadvantaged students now attending school.
We realize that the constitutional guarantee of a thorough and efficient education does not mean the rejection of every method of education, including "effective schools," that does not "guarantee" it. Constitutional implementation in this area must be practical and not necessarily risk free. Under the *380 circumstances of this case, however, given the plight of these districts and their students, the Court requires more than optimism. These children are entitled to the Commissioner's best thinking; they are entitled to "effective schools" if that is what he concludes is required; but they are also, at the same time, entitled to a fair chance in the form of a greater equality of funding. They have already waited too long for a remedy, one that will give them the same level of opportunity, the same chance, as their colleagues who are lucky enough to be born in a richer suburban district.
The decisions regularly made by school districts, the Commissioner, and the Board are based on the premise that what money buys affects the quality of education. Local school boards regularly try to increase their budgets for the explicit purpose of improving the education provided in the district; the Commissioner regularly decides school board budget appeals invariably casting his decision in terms of how much more money is needed for educational improvement; the entire state aid program itself is based on the assumption that money makes a difference in the quality of education; the numerous special programs and pilot projects initiated by the Board and the Commissioner, often accompanied by grant funds, demonstrate the perceived need for money in effecting innovative educational improvement; all of the evidence of high quality education in the richer suburbs on this record attests to the role of money in producing it; the Commissioner's statutory power to require local districts to increase their budgets assumes the relationship; the State's contention that local districts are obliged to increase budgets as much as is needed to achieve a thorough and efficient education also assumes it; and each one of these decisions, actions, or provisions accepts the proposition that what money buys improved staff ratios, higher teacher salaries, expanded course offerings, more equipment makes a difference. This "conventional wisdom" is not just for laypersons, it is the fundamental premise of decision-making by those in charge of education in the districts and in the state.
*381 We therefore adhere to the conventional wisdom that money is one of the many factors that counts. Staff ratios, breadth of course offerings, teacher experience and qualifications, and availability of equipment make a real difference in educational opportunity. We do not mean that money guarantees a thorough and efficient education, nor that, given the approach recommended by the Commissioner, a lower spending district with an effective schools program will not do better than a higher spending district without it. All we mean is that if "effective schools" is a desirable approach, it should be superimposed on a structure that starts out equal. There is nothing in this modern school of thought that suggests it will not work because it is applied to an urban district that has adequate and equal funding.
One aspect of the State's claims that the deficiencies in education are not related either to expenditures per pupil or to property wealth is that they are related to mismanagement in certain districts. The State's claim is that there has been incompetence, politics, and worse in the operation of some urban districts.
While mismanagement has undoubtedly occurred, we agree with the ALJ that it has not been a significant factor in the general failure to achieve a thorough and efficient education in poorer urban districts. It is not just Jersey City that has failed; students in all of the poorer urban districts simply do not receive the quality of education they need to equip them as citizens and competitors in the market, especially when compared to the education given in the affluent suburbs. No amount of administrative skill will redress this deficiency and disparity and its cause is not mismanagement.
These persistent and growing disparities force us again to face the question of the validity of the minimum aid provisions of the Act. Although not declared unconstitutional, those provisions were noted in Robinson I, IV, and V as productive of expenditure disparity. On one occasion the minimum aid funds *382 were ordered to be added to equalization funds and distributed as such. Robinson IV, supra, 69 N.J. at 149-50, 351 A.2d 713. That order never took effect. Robinson V, supra, 69 N.J. at 467-68, 355 A.2d 129.
The minimum aid formula in the Act is counter-equalizing. It is distributed only to districts whose tax base exceeds the Act's guaranteed tax base, in other words, only to relatively richer districts. Its sole function is to enable richer districts to spend even more, thereby increasing the disparity of educational funding between richer and poorer. While substantially less than the amount of such aid under prior law ($48 million in the Act's second year as opposed to $290 million prior to the Act), the amount nevertheless is substantial and the distribution formula of the Act is worse, for it provides no such aid to poor districts. In 1984-85, minimum aid totaled $92.7 million and in 1989-90 it equaled $162.7 million. Under prior law all districts shared in minimum aid.
Minimum aid, for instance, in 1984-85 went to such districts as Englewood Cliffs, which received $135 per pupil although it had an equalized valuation per pupil of $1.24 million, and Saddle River Borough, which received $177 per pupil, with an equalized valuation per pupil of $1.23 million. During that same year Camden, East Orange, Newark, and Trenton, for instance, received none. While not declaring minimum aid unconstitutional, we previously noted its deficiencies in Robinson V, supra, 69 N.J. at 467, 355 A.2d 129, as did the ALJ, the Commissioner, and the Board in this case.
Since we have ruled today that disparity alone does not render the Act unconstitutional, it might be thought that portions of the Act contributing to disparity are therefore not vulnerable to constitutional attack. Furthermore, since our remedy allows expenditure disparity to continue to exist so long as a thorough and efficient education is brought to the poorer urban districts, this particular disparity-producing cause would seem of no constitutional significance.
*383 Disparity of funding is relevant to our constitutional conclusion. That conclusion is based not only on our finding of a substantive lack in the quality of education in these poorer urban districts but also on the significant disparity of spending between them and the richer districts. That disparity strongly supports and is a necessary element of our conclusion that the education provided these students from poorer urban districts will not enable them to compete with their suburban colleagues or to function effectively as citizens in the same society. Given the history of the role of disparate funding and the denial of a thorough and efficient education, and the difficulty experienced by the Legislature in providing full funding in accordance with the Act, continuation of minimum aid in its present form threatens the Legislature's effectuation of the remedy provided herein, the attainment of its constitutional goal, and the future maintenance of a thorough and efficient education both in poorer urban districts and elsewhere.
We therefore hold such minimum aid provisions of the present Act unconstitutional, effective commencing with the school year 1991-92. If, however, the Legislature enacts a new funding system and provides for a phase-in of the new system along with a phase-out of the old, the Act's minimum aid may be eliminated in accordance with that timetable.
In effect, we hold that under the present funding scheme state aid that is counter-equalizing, that increases funding disparities, and that has no arguable educational or administrative justification, is unconstitutional. Categorical aid, although not as equalizing as equalization aid, is not counter-equalizing: it goes to all districts, and in fact more of it goes to the poorer districts. Furthermore, it has clear educational justification: it helps meet the cost of educating students with special needs, who reside in all districts, richer and poorer.
Transportation aid similarly goes to all districts and is distributed in a way that bears no relationship to the wealth of the district. It has obvious educational justification. State aid *384 to fund the Teachers' Pension and Annuity Fund (TPAF), however, distributed to all districts, is in effect counter-equalizing (since richer districts receive a larger share of TPAF contributions because they tend to have proportionately more and higher-paid teachers, the base on which TPAF aid is calculated). The administrative considerations that justify the present system were thought sufficient reason for its continuance in Robinson IV, supra, 69 N.J. at 150, 351 A.2d 713 (redistribution of the pension contribution aid would be inadvisable). We will at this time abide by that judgment without foreclosing the possibility that such aid may be constitutionally infirm.
We stress that it is state aid only that we are discussing here. The fundamental inequality of local funding through the property tax and the funding disparities it produces are asserted to have a justification in its assurance of local control and its encouragement of citizens' participation in their local school system. Without in any way commenting on that assertion, there is no such justification when state aid is concerned. Minimum aid may facilitate the compromises needed to secure passage of important legislation of this kind. We express no judgment on that process or the undoubted difficulties that accompany it. Despite this important role, however, minimum aid in the present funding scheme has no policy justification. Given its actual and potential adverse impact on today's remedy and on the future achievement of a thorough and efficient education, we declare it unconstitutional.V. Findings
From this record we find that certain poorer urban districts do not provide a thorough and efficient education to their students. The Constitution is being violated. These students in poorer urban districts have not been able to participate fully as citizens and workers in our society. They have not been able *385 to achieve any level of equality in that society with their peers from the affluent suburban districts. We find the constitutional failure clear, severe, extensive, and of long duration. We cannot find on this record, however, that there is any constitutional violation in the other districts.
We find that in order to provide a thorough and efficient education in these poorer urban districts, the State must assure that their educational expenditures per pupil are substantially equivalent to those of the more affluent suburban districts, and that, in addition, their special disadvantages must be addressed.
We find that the constitutional deficiency is a product of the Act as applied to these poorer urban districts; that the Board and the Commissioner cannot, even at full funding, achieve a thorough and efficient education in these districts under the present Act.
We find that the changes in the Act proposed by the Board and the Commissioner, and the new regulations adopted, will not achieve a thorough and efficient education in the foreseeable future in these poorer urban districts.
We find that the minimum aid provision of the Act is unconstitutional.VI. Remedy
The Act must be amended, or new legislation passed, so as to assure that poorer urban districts' educational funding is substantially equal to that of property-rich districts. "Assure" means that such funding cannot depend on the budgeting and taxing decisions of local school boards. Funding must be certain, every year. The level of funding must also be adequate to provide for the special educational needs of these poorer urban districts and address their extreme disadvantages.
We leave it to the Legislature, the Board, and the Commissioner to determine which districts are "poorer urban districts." *386 It appears to us that twenty-eight of the twenty-nine school districts designated by the Commissioner as "urban districts" located in DFGs A and B should qualify. (We omit Atlantic City since its tax base for 1989-90 is far in excess of the statutory guaranteed tax base.) Perhaps more should qualify, perhaps fewer. The assured funding per pupil should be substantially equivalent to that spent in those districts providing the kind of education these students need, funding that approximates the average net current expense budget of school districts in DFGs I and J. In addition, provision will be made, presumably similar to categorical aid, for the special educational needs of these districts in order to redress their disadvantages. Such provision will necessarily depend upon the legislative judgment, informed by the Board and Commissioner.
The total additional cost of such a system in the 1989-90 school year would have been approximately $440 million: this dollar figure represents the difference between the average NCEB per pupil, in 1989-90, for students in the twenty-eight poorer urban districts, and the average NCEB per pupil for students in DFGs I and J.
The funding mechanism is for the Legislature to decide. However, it cannot depend on how much a poorer urban school district is willing to tax.
This judicially imposed remedy draws a sharp line and leaves districts of some similarity, like urban districts in DFGs B and C, on different sides of that line. We do not claim that this is the ideal solution, but given the fundamental limits on judicial power, on this record we cannot justify a sliding scale that attempts to tailor the remedy to the varying conditions of the many districts. The record convinces us of a failure of a thorough and efficient education only in the poorer urban *387 districts. We have no right to extend the remedy any further, nor to legislatively smooth out the remedy because of considerations of fairness unrelated to the constitutional command. Moreover, we note that there appear to be significant differences in the level of need suffered by urban districts in DFGs A and B, and those in C.
We realize there will undoubtedly be concern on the part of those districts that share similar characteristics but do not fit within our definition and that therefore will not receive the aid provided for. We suggest that in most cases such districts will also prove to have advantages over those we are targeting. Thus, where such districts find that their expenditures per pupil are lower than those of the poorer urban districts, they will likely find that their socioeconomic status is significantly higher. Nevertheless, given the limitation of judicial power, we recognize that the kind of equity that can be done in this area by the Legislature cannot be accomplished by judicial order.
The Legislature may devise any remedy, including one that completely revamps the present system, in terms of funding, organization, and management, so long as it achieves a thorough and efficient education as defined herein for poorer urban districts. It may phase in that new system and phase out the old. It may choose, for instance, to equalize expenditures per pupil for all districts in the state at any level that it believes will achieve a thorough and efficient education, and that level need not necessarily be today's average of the affluent suburban districts. The most significant aspect of that average *388 today is not its absolute level, but its disparity with the average of the twenty-eight poorer urban districts. It may determine the division between state aid and local funding and allow school districts such leeway as is consistent with the constitutional obligation, or it may mandate the local share; again, however, funding in poorer urban districts cannot depend on the budgeting and taxing decisions of local school boards. We assume the design of any new funding plan will consider the problem of municipal overburden in these poorer urban districts.
Whatever the legislative remedy, however, it must assure that these poorer urban districts have a budget per pupil that is approximately equal to the average of the richer suburban districts, whatever that average may be, and be sufficient to address their special needs.
We find the evidence of the importance of competent management to the quality of education substantial. While the State has pointed to mismanagement as one of the causes of the failure of education in the poorer urban districts, it has not complained of the lack of statutory power to redress it. If there is any such lack and if it impairs the constitutional obligation, that matter would be judicially cognizable. Our power to require a thorough and efficient education is not limited to a money remedy. Cf. Southern Burlington County N.A.A.C.P. v. Mount Laurel Township (Mount Laurel II), 92 N.J. 158, 456 A.2d 390 (1983) (power to require municipalities to take affirmative action to satisfy their affordable housing obligations).
We have not attempted to address disparity of spending as such. To the extent that the State allows the richer suburban districts to continue to increase that disparity, it will, by our remedy, be required to increase the funding of the poorer urban districts. We limit our remedy at this point to increasing funding where we find a deficiency. We do not require equalized funding statewide. We are satisfied, however, that whatever *389 degree of statewide equality the Legislature may wish to achieve, or may find it feasible to achieve, it cannot constitutionally do so for these poorer urban districts simply by raising the guaranteed tax base under the present formula. These districts, even assuming the most generous GTB increase, will not be willing or able to fund what is required for a thorough and efficient education. Their need to conserve their tax dollars, their need not to increase their total tax rate, will inevitably persuade them not to spend more but to tax less.
We recognize that the factors that determine our decision can change. The only constant is the definition of a thorough and efficient education one that will equip all of the students of this state to perform their roles as citizens and competitors in the same society.
The increased funding ordered here for poorer urban districts may be more than they can efficiently absorb immediately. We are also aware of the fact that the increased funding may constitute a heavy burden for the State to adjust to. We therefore rule that while the new funding mechanism must be in place legislatively so as to take effect in the school year 1991-92, it need not be fully implemented immediately, but may be phased in.
We have been urged by plaintiffs to set forth the consequences of a legislative failure to act in conformance with our opinion. Their justification for the request is their belief that the political problems created in this area are so severe that unless the Legislature understands the cost of failure to conform, it may be institutionally unable to take the steps required. Despite the experience over these many years, we continue to believe that the Legislature will conform. We reject plaintiffs' suggestion on this score.
We decline to rule on plaintiffs' state equal protection claim. The core of their argument is that wealth-based disparity is causing educational disparity. They contend, in effect, that what they consider the fundamental right of education is affected *390 by the property wealth of the school district, that the system in reality consists of a classification of students that determines their level of education by a characteristic not only irrational but suspect, the property wealth of the districts they live in, and that there is no compelling State interest to justify the classification. We referred in Robinson I to the monumental governmental upheaval that would result if the equal protection doctrine were held applicable to the financing of education and similarly applied to all governmental services. 62 N.J. at 492-501, 303 A.2d 273. We need not deal with those implications, for the remedy afforded in this opinion, although not based on equal protection, substantially mitigates plaintiffs' equal protection claim.
As for plaintiffs' claim that our system of funding education violates the Law Against Discrimination, N.J.S.A. 10:5-1 et seq., the ALJ, the Commissioner, and the Board found that plaintiffs had failed to prove racial discrimination. The Commissioner further found that discriminatory impact had not been proven, and both he and the Board concluded that the Law Against Discrimination may not be applicable to a claim of racial discrimination based on the effect of the school funding scheme. The claim under this law was not included in the original pleadings or in the prehearing order setting forth the issues. It was apparently never explicitly advanced as a claim until the hearing had concluded. Furthermore, it formed no part of our remand in Abbott I. It appears that the issue was not fully litigated. Under these circumstances, it would be unfair to the State were we to rule on it now. We therefore dismiss this part of plaintiffs' claim without prejudice.
Finally, we do not pass on plaintiffs' claim for relief in the form of a capital improvement timetable for all facilities in all school districts "to be conformed to contemporary educational standards." As we noted in Robinson I, supra, 62 N.J. at 520, 303 A.2d 273, "the State's obligation includes as well the capital expenditures without which the required educational opportunity *391 could not be provided." That obligation has not been fulfilled; the lack is so great, the State concedes, that under the present funding plan sufficient capital investment cannot be provided. We find the record insufficient to fashion a remedy concerning capital construction. As implied in Robinson I, it is a matter within the judicial power to correct, as is the current financing scheme embedded in the Act. Its correction, however, is a massive undertaking, one that this Court could not consider until it knew precisely what the deficiencies are, how much their correction would cost, and how best to bring about such correction. All we have before us are, in addition to some specific figures concerning several districts, general agreement on the desperate condition of school facilities, gross estimates of the cost of correction, and concurrence on the urgent need. It is obviously a matter best suited for legislative treatment, but if squarely presented to us with an adequate record of need and legislative failure, we would be obliged under the Constitution to consider the matter.VII. Conclusion
This case has a special context that brings the constitutional obligation into sharp focus as it applies to the urban poor. While we necessarily deal with our system of education statewide, the issue put to us by the plaintiffs is the education of those children who live in poverty. Their cities have deteriorated and their lives are often bleak. They live in a culture where schools, studying, and homework are secondary. Their test scores, their dropout rate, their attendance at college, all indicate a severe failure of education. While education is largely absent from their lives, we get some idea of what is present from the crime rate, disease rate, drug addiction rate, teenage pregnancy rate, and the unemployment rate.
*392 Without an effective education they are likely to remain enveloped in this environment. Their overall needs are not limited to education, but that need is overwhelming.
Clearly, we are failing to solve this problem. It is the problem of bringing this important and increasingly isolated class into the life of America, for this is not just a New Jersey problem. There is progress, and there are some successes in education, but the central truth is that the poor remain plunged in poverty and severe educational deprivation. The devastation of the urban poor is more significant in New Jersey than in most states both because of our demographics and the structure of our society. Our large black and hispanic population is more concentrated in poor urban areas and will remain isolated from the rest of society unless this educational deficiency in poorer urban districts is addressed.
While the constitutional measure of the educational deficiency is its impact on the lives of these students, we are also aware of its potential impact on the entire state and its economy not only on its social and cultural fabric, but on its material well-being, on its jobs, industry, and business. Economists and business leaders say that our state's economic well-being is dependent on more skilled workers, technically proficient workers, literate and well-educated citizens. And they point to the urban poor as an integral part of our future economic strength. In short, they urge the state to go about the business of substantially improving the education of the very subjects of this litigation, the students in poorer urban districts. So it is not just that their future depends on the State, the state's future depends on them. That part of the constitutional standard requiring an education that will enable the urban poor to compete in the marketplace, to take their fair share of leadership and professional positions, assumes a new significance.
We note a further impact on the continuing constitutional failure. Soon, one-third of our citizens will be black or hispanic, and many of them will be undereducated. This substantial *393 segment of our population is isolated in a separate culture, in a society they see as rich and poor, for to the urban poor, all other classes are rich. There is despair, and sometimes bitterness and hostility.
The fact is that a large part of our society is disintegrating, so large a part that it cannot help but affect the rest. Everyone's future is at stake, and not just the poor's. Certainly the urban poor need more than education, but it is hard to believe that their isolation and society's division can be reversed without it. That part of the constitutional standard, then, requiring an education sufficient to enable students to assume their proper roles as citizens takes on a new significance.
Measured by any accepted standard, New Jersey has been generous in the amount of money spent for education. We currently spend more dollars per student for education than almost any other state. Given that fact, this Court could not conclude that the State has failed to provide for a thorough and efficient education in all school districts. To so conclude would mean that our State Constitution has invented a standard so different from, and substantially higher than, the rest of the country that even though we spend almost the most, constitutionally that is not enough. The dilemma is that while we spend so much, there is absolutely no question that we are failing to provide the students in the poorer urban districts with the kind of an education that anyone could call thorough and efficient.
There is another perspective. Our citizens and our government are obviously dedicated to education and generous towards our children, otherwise we would not spend that much. There are other reasons for this level of spending, however. The need is great and the money is there. We are the second richest state in the nation. Therefore, while the relatively *394 high level of our present expenditures must give us pause, it must also be viewed in the light of our needs and our wealth.
After all the analyses are completed, we are still left with these students and their lives. They are not being educated. Our Constitution says they must be.
Included in our perspective are the stories of success. They show that the urban poor are capable, that given sufficient attention in an adequately financed system using the best knowledge and techniques available, a thorough and efficient education is achievable.
This record proves what all suspect: that if the children of poorer districts went to school today in richer ones, educationally they would be a lot better off. Everything in this record confirms what we know: they need that advantage much more than the other children. And what everyone knows is that as children the only reason they do not get that advantage is that they were born in a poor district. For while we have underlined the impact of the constitutional deficiency on our state, its impact on these children is far more important. They face, through no fault of their own, a life of poverty and isolation that most of us cannot begin to understand or appreciate.
We reverse the Board's decision. The Act is unconstitutional as applied to poorer urban districts.APPENDIX SCHOOL DISTRICTS IN DISTRICT FACTOR GROUPS A AND B ([*] indicates "urban districts" included in Supreme Court remedy)
District Factor Group A
[*]Asbury Park City (Monmouth County) Atlantic City (Atlantic County) Bass River Township (Burlington County) Beverly City (Burlington County)
*395 Bradley Beach Borough (Monmouth County) [*]Bridgeton City (Cumberland County) Buena Regional (Atlantic County) [*]Camden City (Camden County) Chesilhurst Borough (Camden County) Commercial Township (Cumberland County) Cumberland County Regional (Cumberland County) Deerfield Township (Cumberland County) Dennis Township (Cape May County) Downe Township (Cumberland County) Eagleswood Township (Ocean County) East Newark Borough (Hudson County) [*]East Orange City (Essex County) Egg Harbor City (Atlantic County) [*]Elizabeth City (Union County) Fairfield Township (Cumberland County) [*]Gloucester City (Camden County) [*]Hoboken City (Hudson County) [*]Jersey City (Hudson County) [*]Keansburg Borough (Monmouth County) Lakehurst Borough (Ocean County) Lawrence Township (Cumberland County) Lower Alloways Creek (Salem County) Lower Cape May Regional (Cape May County) Lower Township (Cape May County) Maurice River Township (Cumberland County) Middle Township (Cape May County) National Park Borough (Gloucester County) [*]Newark City (Essex County) North Wildwood City (Cape May County) [*]Passaic City (Passaic County) [*]Paterson City (Passaic County) Paulsboro Borough (Gloucester County) Pemberton Borough (Burlington County) Penns Grove-Carneys Point Consolidated (Salem County) [*]Perth Amboy City (Middlesex County) Pinelands Regional (Ocean County) [*]Pleasantville City (Atlantic County) Salem City (Salem County) Seaside Heights Borough (Ocean County) Swedesboro-Woolwich (Gloucester County) [*]Trenton City (Mercer County)
*396 Tuckerton Borough (Ocean County) Union Beach Borough (Monmouth County) [*]Union City (Hudson County) Washington Township (Burlington County) West Cape May Borough (Cape May County) [*]West New York Town (Hudson County) Weymouth Township (Atlantic County) Wildwood City (Cape May County) Woodbine Borough (Cape May County) District Factor Group B Alloway Township (Salem County) Bellmawr Borough (Camden County) Berkeley Township (Ocean County) Berlin Township (Camden County) Brooklawn Borough (Camden County) [*]Burlington City (Burlington County) Central Regional (Ocean County) Clayton Borough (Gloucester County) Clementon Borough (Camden County) Egg Harbor Township (Atlantic County) Elk Township (Gloucester County) Folsom Borough (Atlantic County) Franklin Township (Gloucester County) [*]Garfield City (Bergen County) Greater Egg Harbor Regional (Atlantic County) Greenwich Township (Cumberland County) Hainesport Township (Burlington County) Hammonton Town (Atlantic County) [*]Harrison Town (Hudson County) [*]Irvington Town (Essex County) Keyport Borough (Monmouth County) Lacey Township (Ocean County) Lawnside Borough (Camden County) Little Egg Harbor Township (Ocean County) [*]Long Branch City (Monmouth County) Manchester Township (Ocean County) Mannington Township (Salem County) [*]Millville City (Cumberland County) Monroe Township (Gloucester County) Moonachie Borough (Bergen County) Mount Ephraim Borough (Camden County)
*397 Mullica Township (Atlantic County) [*]New Brunswick City (Middlesex County) Ocean Gate Borough (Ocean County) Ocean Township (Ocean County) Oldmans Township (Salem County) [*]Orange City (Essex County) Oxford Township (Warren County) [*]Pemberton Township (Burlington County) [*]Phillipsburg Town (Warren County) Pine Hill Borough (Camden County) Pittsgrove Township (Salem County) Plumsted Township (Ocean County) Quinton Township (Salem County) South Harrison Township (Gloucester County) Southern Gloucester County Regional (Gloucester County) Southern Regional (Ocean County) Stafford Township (Ocean County) Sussex-Wantage Regional (Sussex County) Upper Deerfield Township (Cumberland County) Upper Pittsgrove Township (Salem County) [*]Vineland City (Cumberland County) Westville Borough (Gloucester County) Wildwood Crest Borough (Cape May County) Woodlynne Borough (Camdem County)
For reversal Chief Justice WILENTZ, and Justices CLIFFORD, HANDLER, POLLOCK, O'HERN, GARIBALDI and STEIN 7.
For affirmance None.NOTES
 N.J.Const. of 1947 art. VIII, § 4, para. 1 ("The Legislature shall provide for the maintenance and support of a thorough and efficient system of free public schools for the instruction of all the children in the State between the ages of five and eighteen years.").
 "The commissioner shall have jurisdiction to hear and determine ... all controversies and disputes arising under the school laws ... or under the rules of the state board or of the commissioner." N.J.S.A. 18A:6-9.
 "Production-function" is an economic term describing the relationship between quantitative measures of output and various inputs. For example, in the educational setting, production-function analysis would examine whether teachers with higher salaries, or advanced degrees, or more experience (input) can increase student achievement (output), as measured by test scores, or other objective indicia. For skeptical reviews of this research, see Murnane and Pauly, "Lessons from Comparing Educational and Economic Indicators" Phi Delta Kappan (March 1988); Robinson and Brotheroe, "Cost of Education: An Investment in America's Future" in Status and Trends in School Finance, An Educational Research Service Report (1987). For an earlier discussion, including larger issues such as the evolution of government funding programs and the relationship between school performance, income distribution, and societal productivity, see Carnoy, "Economics and Education", in 2 Encyclopedia of Educational Research (H. Mitzel, J. Best, and W. Rabinowitz 5th ed. 1983) at 519-32.
 In the 1988-89 and 1989-90 school years, New Jersey ranked second only to Alaska in this respect. Education Daily, at 7 (Feb. 2, 1990); Public Elementary and Secondary State Aggregate Nonfiscal Data, by State, for School Year 1988-89; and School Revenues and Current Expenditures for Fiscal Year 1988 (Table 7), National Center for Education Statistics, U.S. Department of Education, Office of Educational Research and Improvement, E.D. Tabs (March 1990). As the ALJ noted, in prior years, the state had usually ranked third or fourth in the nation.
 Arizona, Colorado, Georgia, Idaho, Illinois, Maryland, Michigan, New York, Ohio, Oklahoma, Oregon, Pennsylvania, South Carolina, and Wisconsin.
 Kentucky, Montana, New Jersey, Texas, Washington, and West Virginia.
 Respectively, Arkansas, Connecticut, and Wyoming (both claims); California (equal protection only).
 Amicus briefs were submitted by: the American Civil Liberties Union of New Jersey; American Jewish Congress; Association for Children of New Jersey; League of Women Voters of New Jersey; Legal Services of New Jersey; Metropolitan Ecumenical Ministry, et al.; Newark Teachers' Union, Local 481, AFT/AFL-CIO; New Jersey Black Issues Convention, Inc.; New Jersey Education Association; New Jersey School Boards Association; N.J. State Conference of NAACP Branches, et al.; Public Advocate.
 The Act has been fully funded at the 134% guaranteed tax base level on only two occasions, first in 1977-78 (the Act's second year) and in 1978-79. For all other years funding has varied from 129% to 134% (the statutory guaranteed tax base level) thereby diminishing the equalizing effect of the Act. For the school year 1990-91, due to a severe fiscal crisis affecting the entire state, the school districts apparently are scheduled to receive considerably less aid than in the past.
 There is, however, as indicated infra at 297, 575 A.2d at 364, a limit, subject to waiver, on the amount a school district can increase its budget each year, and a limit on the amount of equalization aid a school district can receive.
 The budget putting aside categorical aid and some other kinds of aid of some districts, such as Camden, is supported by as much as 90%. If such a district's budget is $5 million and the State pays $4.5 million, the local school tax rate is affected: instead of a $20.00 tax rate, it may be $2.00.
 Minimum aid is calculated by taking 10% of the following quantity: 1.000 minus the quotient of the school district's equalized property valuation per pupil divided by the minimum aid guaranteed property valuation. The result is a fraction less than 1.0 and usually expressed as a percentage. Guaranteed valuation equals 11.5 times the state average equalized valuation per pupil. The figures used in the following example were provided by the DOE. In the Mahwah school district for 1985-86, its equalized property valuation per pupil was $414,420. To calculate minimum aid, multiply 10% by 1.000 minus the quotient of $414,420 divided by $2,147,257 (11.5 times $186,718, the state average equalized valuation per pupil, which results in the minimum aid guaranteed property valuation for 1985-86) which yields .0807. Thus, the share for Mahwah is 8.07%. Mahwah's 8.07% is multiplied by $6,541,832, the "maximum support budget" (the statewide sixty-fifth percentile of NCEB budgets ($3,764) times the number of resident pupils (1,738) in the particular district) to get the minimum aid entitlement. For 1985-86 it was $527,926. In 1984-85, minimum aid amounted to $504,654. Minimum aid to Mahwah has increased steadily over the years (1986-87 $531,456; 1987-88 $570,285; 1988-89 $586,959; 1989-90 $640,746). In 1985-86, 34.4% of New Jersey's school districts, or 207 districts, were above the guaranteed tax base and received minimum aid.
 Some insight into the strength of the commitment of those attacking the Act and the funding mechanism of the past is gained by noting the contingent remedies sought here and in Robinson v. Cahill. In both cases the plaintiffs requested that all state aid except transportation aid should be dispensed in accordance with the equalization formula. In other words, all of the funds now distributed through minimum aid, categorical aid, and TPAF should be transferred to and distributed in accordance with the equalization aid formula. That remedy was dealt with to some extent in Robinson I, supra, 62 N.J. at 520-21, 303 A.2d 273; Robinson v. Cahill, 63 N.J. 196, 197-98, 306 A.2d 65 (1973) (Robinson II); and Robinson IV, supra, 69 N.J. at 147-50, 351 A.2d 713, and was provisionally granted (although never put into effect) in Robinson IV, supra, 69 N.J. at 149-50, 351 A.2d 713.
 The fluctuating role of federal aid to public education is well documented. See, e.g., Carnoy, "Economics and Education", supra at 297, n. 3, 575 A.2d at 364 n. 3, where the authors trace federal appropriations from the generous contributions as part of the federal war against poverty in the 1960s, to the retrenchment in the 1970s based on rising inflation and declining enrollment, to the efforts in the late 1970s to limit tax and expenditure levels, and finally to the Reagan Administration's proposal to reduce federal funding for education by 25% in 1981. More recently, President Bush proposed cutting an estimated $400 million from the federal education budget. See Time, October 9, 1989, at 60.
 Unless otherwise indicated, all dollar figures are for 1984-85; today, these dollar figures would be substantially higher. The poorer districts in this comparison are those classified by the DOE as those in District Factor Groups (DFGs) A and B. The richer districts are those in DFGs I and J. The middle districts are those in DFGs C through H. The NCEB per pupil averages have been calculated to indicate the pupil-average rather than the district-average, as have all averages unless otherwise noted. The NCEB per pupil figures are in the record. At our request, the DOE calculated DFG averages. DFGs are discussed infra at 338-340, 575 A.2d at 384-386.
 The State's position on the significance of SES and education is that DFG comparisons were designed solely for the purpose of curriculum planning and evaluation.
 DOE officials testified that the seven additional districts were found through a process of inspection and a survey done in the early 1970s asking school superintendents to designate their districts as one of eight community types ranging from urban to rural. If a district had been rated urban or urban-suburban and was also in DFG A or B, it was added to the DOE's list. The record does not disclose by what criteria the remaining districts, Howell Township and West Orange Township, were added.
 Our use of terms should be clarified. "Urban districts" are those fifty-six districts classified as such by the DOE. "Poorer urban districts" are those twenty-nine urban districts within DFGs A and B. The attached Appendix lists all school districts in DFGs A and B. An asterisk appears before twenty-eight of the twenty-nine districts designated by the Commissioner as "urban districts" (we have excluded Atlantic City because of its high property wealth). See infra at 384-391, 575 A.2d at 408-411 for application of the remedy to these districts.
 In 1984-85, 71% of all minorities in the state were educated in these poorer urban districts. In 1986-87, Camden's school enrollment was 95% minority; East Orange, 99% minority; Jersey City, 85%; Trenton, 88%; Newark, 91%; Paterson, 90%; and Irvington, 94%.
 The septiles contain all districts, and, unlike another of plaintiffs' statistical analyses, do not exclude the fifth and ninety-fifth percentiles.
 The State claims that plaintiffs have not proven that these seven urban districts (the plaintiffs' districts plus Newark, Trenton, and Paterson) are typical of poorer urban districts or that the ten comparison districts are typical of richer suburban districts. While differences undoubtedly exist, we have absolutely no doubt, based on the record, that the twenty-eight urban districts that are the subject of our remedy share the educational characteristics and problems of the seven urban districts mentioned above; and that enough of the richer suburban districts in DFGs I and J share the educational characteristics and advantages of the ten comparison districts to warrant classifying all of them together for the purpose of determining a remedy. While we agree with the Commissioner and the Board that the proofs are lacking concerning districts in between, they are convincing beyond doubt concerning poorer urban districts and richer suburban districts; the seven poorer urban districts and the ten richer suburban districts are sufficiently typical of their class, and the disparity between the two classes almost compels, by itself, the conclusion that the students in the poorer urban districts are not receiving a thorough and efficient education.
 Although the measure of day school expenditures per pupil includes federal and state categorical aid, the figure is not adjusted by weighting pupils who have special needs and on whose account the districts receive categorical aid.
 The ten elements are, 1) district educational planning, 2) community relations, 3) comprehensive curriculum and instruction, 4) pupil attendance, 5) facilities, 6) staff, 7) mandated programs, 8) mandated basic skills tests, 9) equal educational opportunity and affirmative action, 10) financial matters.
 Although the DOE's original regulations prescribed fifty-one indicators, only forty were mandatory for certification. Under current regulations all forty-three indicators are mandatory. N.J.A.C. 6:8-4.5.
 The Commissioner noted that 98% of all school districts are certified or will soon become certified; and that of the fifty-six districts classified as "urban", most are certified or will soon become certified.
 As of March 30, 1989, seventeen districts had been placed in Level II or III monitoring. Of these seventeen, thirteen are urban districts in DFGs A or B. Two of the remaining four districts are also in these DFGs.
 The mandated basic skills test element prescribed by N.J.A.C. 6:8-4.3(8) also requires that 75% of the pupils in grade three and 75% of the pupils in grade six in each school in the district score at or above the minimum level of proficiency for commercially published or district criterion referenced tests.
 We do not intend to detract in the least from the efforts of the Commissioner. The ALJ noted that "the regulatory challenge imposed by Chapter 212 was so immense that monitoring had to evolve as the Department became familiar with specific problems." He noted further that the process addressed significant issues and was improving. We add that the Commissioner's efforts occur in a context where experts have failed to agree on what constitutes a reliable indicator of good education, and that the challenge itself is very new. In short, we intend no more by our judgment than that certification as it has thus far evolved does not assure a thorough and efficient education.
 We note in this connection plaintiffs' suggestion that the school day and school year be expanded to accommodate such activities, and to their hope that new and enhanced programs, beginning in preschool, may eventually shrink the time needed for remedial skills. Plaintiffs argue for full-day kindergarten and pre-kindergarten programs and intensive counseling to help poor children adjust to school, for lower student-teacher ratios in the early grades and later, for parental training and community schools in which parents, often themselves teenage dropouts, may continue their own education and learn how to help their children succeed in school, for money to improve facilities and allow teachers to learn new instructional strategies. They point out the special needs of these children, their developmental lags, their often poor health and nutrition, the difficulties they have seeing themselves as students and workers.
For their part, the State's witnesses often agreed with plaintiffs on the value of specific programs. Thus, Dr. Bloom, Assistant Commissioner of the Division of General Academic Education, was enthusiastic about early intervention: "if you start kids in pre-school, take them through summer ... you will have them achieving at levels no one ever predicted.... It's proven." Nevertheless, current funding from all sources, state and federal, is inadequate to provide quality day care and full-day kindergarten for more than a fraction of the children in need. Similarly, parental involvement is a key component of "effective schools," defendants' preferred approach to improving urban education, but they provide no discussion of the cost in staff time and other resources of fostering such relationships.
 Beyond this, recent scholarly discussion has focused heavily on the need for individualized instruction tailored to children's different needs and development patterns, experimental learning ranging from scientific experiment to poetry writing that responds to and develops children's curiosity, and interventions that address the hostile attitudes minority children may bring to a system often perceived as "white" and alien. Much of this research suggests that schools may fail to build on the knowledge and skills poor and minority children bring to school, strengths often different from those of white, middle class children of the same age. For a summary of recent work on the needs of disadvantaged children, see Better Schooling for the Children of Poverty: Alternatives to Conventional Wisdom, U.S. Department of Education, Office of Planning, Budget and Evaluation, prepared under contract by SRI International, Menlo Park, California, and Policy Studies Associates, Washington, D.C., Volume 1: Summary (by M.S. Knapp and B.J. Turnbull); Volume II: Commissioned Papers and Literature Review (M.S. Knapp and P.M. Shields eds.) (1990). For a more general collection of scholarly articles on the crisis in public education, see Daedulus, Journal of the American Academy of Arts and Sciences, "Literacy in America" (Spring 1990).
As judges rather than educators or social scientists, we are in no position to assess the value of these approaches or the place they should be given relative to more traditional book and workbook exercises. Nevertheless, it seems clear to us that experimentation is needed to reverse the staggering failure of our poorer urban districts, and that experimentation itself requires money.
 Notwithstanding our use of these statistics, we note that the importance of teacher/pupil ratios, at least where the classes exceed three to five students per teacher, has been the subject of considerable debate. See, e.g., Tomlinson, "Class Size and Public Policy: Politics and Panaceas", Educational Policy 3 261 (1989). Nevertheless, plaintiffs' experts testified that class size is important, particularly in the lower grades, and that research indicates that disadvantaged children benefit more from smaller classes than other children.
 These averages include federally funded teachers. If such teachers were excluded from the comparison, the disparity would be greater. The districts compared here are those in DFGs A and B, and those in I and J. See supra at 338-342, 575 A.2d at 384-386 for discussion of DFGs.
 These twenty-nine districts are defined by the Department of Education as "urban districts" within DFGs A and B. See supra at 338-342, 575 A.2d at 384-386 for discussion of these classifications.
 We note the alleged improvement in the HSPT scores in 1989 and the controversy surrounding them. Sunday Star-Ledger, "`No-shows' skew rise in scores on skills test," December 10, 1989, at 1, col. 3. The factors leading to our conclusion herein would not be affected even were we to credit these scores as reported. Since they formed no part of the record before us, were never subject to adversarial inquiry, and have been severely questioned concerning their significance and reliability, it would not be appropriate to consider them in connection with this decision. The State has not requested us to do so.
 In 1977-78 and 1978-79, the official dropout rate for twenty-three urban high schools was 11%. DOE selected these schools for the study because of their failure to meet the DOE established basic-skills test cutoffs. However, the official figures greatly understate the severity of the dropout problem because those figures represent only students who formally withdraw from school. In other words, the official dropout numbers do not account for those students who just stop attending school. A 1980 study prepared for the DOE used a different approach in calculating dropouts for the selected twenty-three urban high schools. Updated in 1985, the study reflects the following "unofficial" dropout figures. Pleasantville High School (Pleasantville) had a dropout rate of 55%; Woodrow Wilson High School (Camden) 58%; Newark Central High School (Newark) 55%; Orange High School (Orange) 50%; Dickinson High School (Jersey City) 44%; Ferris High School (Jersey City) 49%; Lincoln High School (Jersey City) 56%; Trenton Central High School (Trenton) 48%; Asbury Park High School (Asbury Park) 41%; Kennedy High School (Paterson) 50%; Eastside High School (Paterson) 46%; Passaic High School (Passaic) 45%.
 In an insightful article, Gene Maeroff, a former New York Times reporter, now a senior fellow at the Carnegie Foundation for the Advancement of Teaching, Princeton, New Jersey, sounds themes that recur in much of the literature on schools and the urban underclass:
The inner city has long been populated by poor people, but what makes the situation different today and exacerbates the isolation is the flight of middle-class blacks and the virtual abandonment of entire black neighborhoods to the poorest of the poor. There has been a bifurcation, described by Nicholas Lemann in The Atlantic, that has drained the ghettos of many of those who might be constructive role models for the young. A new kind of society is emerging in these neighborhoods, one with its own values, one that is "utterly different from that of the American mainstream."
Isolation ... strengthens the hold of the subculture, giving free play to values that neither reinforce schooling nor encourage the development of the habits of mind needed for academic success.
What is clearly needed is a fresh approach to urban education, involving smaller learning units and a different philosophy of instruction. The goal must be to create within each learning unit a sense of community and a desire on the part of students to belong to that community.... More classes must be organized around seminars, discussions, and cooperative learning, and students must be encouraged to take greater responsibility for their own education. Ideas and concepts, not facts and statistics, should form the core of this experience.
[Maeroff, "Withered Hopes, Stillborn Dreams: The Dismal Panorama of Urban Schools", Phi Delta Kappan, 633, 634-35 (May 1988).]
 In February 1990, the National Governors' Association adopted a set of "National Education Goals." The formulation was an outgrowth of the Governors' conference on education convened by President Bush five months earlier. Under Goal 1, readiness for school, the governors suggested that "[a]ll eligible children should have access to ... a successful pre-school program with strong parental involvement" and that "[o]ur first priority must be to provide at least one year of preschool for all disadvantaged children." National Education Goals, statement issued by National Governors' Association on February 25, 1990, at 7.
 The "effective schools" approach derives from research begun in the 1970s to identify common characteristics of unusually successful schools. Almost all of this research focused on elementary schools in poor urban areas and used standardized tests in basic skills to measure performance. While there are now several models explaining "effective schools," the one defendants primarily rely on contains six correlates. Assistant Commissioner Bloom explained these correlates as: (1) leadership is executed by the principal and faculty; (2) the schools must have a mission of basic skills achievement; (3) the assessment of achievement is ongoing; (4) staff have raised expectations of student success; (5) the schools have a safe, orderly environment; and (6) parents are involved with the schools.
Plaintiffs point out that this approach has had mixed results where it has been applied and that extra money has usually been supplied to implement it. They note that retraining teachers and principals requires money, that eliciting effective participation by parents involves staff time, and that it is easier to maintain a safe and orderly environment if fewer students are assigned to each teacher. Moreover, teaching oriented programs sometimes thought complementary to effective schools, such as Achievement Directed Leadership (ADL), a protocol developed by Research for Better Schools in Philadelphia, and Mastery Learning, a teaching method originally developed by Dr. Benjamin Bloom at the University of Chicago, are expensive.
Beyond this, plaintiffs point out that the major impetus for both the research and application of effective schools methodology has been limited to basic skills acquisition and that even there the approach has narrowed, not eliminated, the gap between poor and wealthy children. Indeed, many programs have not even aimed at more. The Connecticut Effective Schools Program, for example, which defendants cite as a successful low cost program, used the 30th percentile of the California Achievement Test, 20 percentiles below the national average, as its standard for proficiency. Even so, when in 1985 Connecticut's Department of Education evaluated ten schools that had participated in the program for at least three years, it found that only five showed significant achievement improvement.
 The 1989-90 data were not in the record. They were produced by the DOE at our request; based on those data we have attempted to calculate the additional cost.
 As an indication of these differences, in FY 1986, the State allocated $37 million in "urban aid" under N.J.S.A. 52:27D-178 to urban districts in DFGs A and B, but only $1.8 million to urban districts in DFG C. We suggest that if the Legislature was willing to draw the line defining "urban aid districts" where it did and to allocate aid strictly within that line, and if the Commissioner was willing to draw a similar line in defining "urban districts," there should be no inequity perceived in our use of the same criteria in conjunction with this demarcation of our remedy and its further refinement by drawing the line at DFGs A and B.
 Survey of Current Business, 70 Bureau of Economic Analysis of Dept. of Commerce, No. 4, at 157 (April 1990).