delivered the opinion of the Court.
For the seventh time since the late-1980s, we are called upon to assess the constitutionality of the Texas school finance system, a recondite scheme for which the word “Byzantine” seems generous.
In this round, more than half of the State’s 1,000-plus school districts have brought the most far-reaching funding-challenge in Texas history. We are presented with a court reporter’s record exceeding 200,000 pages and a trial court judgment accompanied by 1,508 findings of fact and 118 conclusions of law. Dozens of briefs, many filed by new parties raising new claims, frame the intricate arguments now before us. The depth and breadth of Texans’ attention is understandable — and also commendable: Good education is good policy.
But our judicial responsibility is not to second-guess or micromanage Texas education policy or to issue edicts from on high increasing financial inputs in hopes of increasing educational outputs. There doubtless exist innovative reform measures to make Texas schools more accountable and efficient, both quantitatively and qualitatively. Judicial review, 'however, does not licence second-guessing the political branches’ policy choices, or substituting the wisdom of nine judges for that of 181 lawmakers. Our role is much more limited, as is our holding: Despite the imperfections of the current school funding regime, it meets minimum constitutional requirements.
Imperfection, however, does not mean imperfectible. Texas’s more than five million school children deserve better than *834serial litigation over an increasingly Dae-dalean “system.” They deserve transformational, top-to-bottom reforms that amount to more than Band-Aid on top of Band-Aid. They deserve a revamped, nonsclerotic system fit for the 21st century.
I. Factual and Procedural Background
A. Overview of the Texas Public School System
1. The Public Education System
The Legislature has enacted numerous statutes articulating its goals for Texas schools. Among those statutes, section 4.001(a) of the Education Code1 provides:
The mission of the public education system of this state is to ensure that all Texas children have access to a quality education that enables them to achieve their potential and fully participate now and in the future in the social, economic, and educational opportunities of our state and nation. That mission is grounded on the conviction that a general diffusion of knowledge is essential for the welfare of this state and for the preservation of the liberties and rights of citizens. It is further grounded on the conviction that a successful public education system is directly related to a strong, dedicated, and supportive family and that parental involvement in the school is essential for the maximum educational achievement of a child.
Subtitle F of the Education Code sets out provisions for curriculum, programs, and services. Section 28.001 of the subtitle2 states:
It is the intent of the legislature that the essential knowledge and skills developed by the State Board of Education under this subchapter shall require all students to demonstrate the knowledge and skills necessary to read, write, compute, problem solve, think critically, apply technology, and communicate across all subject areas. The essential knowledge and skills shall also prepare and enable all students to continue to learn in postsec-ondary educational, training, or employment settings.
In Neely v. West Orange-Cove Consolidated Independent School District (WOC II), we noted that the Texas school system has “four integrated components: a state curriculum, a standardized test to measure how well the curriculum is being taught, accreditation standards to hold schools accountable for their performance, and sanctions and remedial measures for students, schools, and districts to ensure that accreditation standards are met.”3
As part of that system, the Education Code requires school districts and open-enrollment charter schools to offer both a foundation curriculum and an enrichment curriculum. A foundation curriculum consists of English language arts, mathematics, science, and social studies.4 An enrichment curriculum consists of languages other than English to the extent possible, health, physical education, fine arts, career and technology education, technology applications, religious literature, and personal financial literacy.5 Alongside these requirements, Texas law establishes bilin*835gual education programs with the goal of “providing a full opportunity for all students to become competent in speaking, reading, writing, and comprehending the English language.”6
Establishing the required curriculum is a significant undertaking. The State Board of Education (SBOE)7 is required to “identify the essential knowledge and skills of each subject of the required curriculum that all students should be able to demonstrate.” 8 To that end, as we discussed in WOC II, the SBOE adopted the Texas Essential Knowledge and Skills (TEKS) curriculum.9
In 2006, the Legislature required the SBOE to incorporate college readiness standards into the TEKS curriculum.10 The Legislature has defined college readiness as “the level of preparation a student must attain in English language arts and mathematics courses to enroll and succeed, without remediation, in an entry-level general education course for credit in that same content area for a baccalaureate degree or associate degree program.”11 Adoption of these standards has been a laborious process that continues today, involving “the direct input of educators, parents, business and industry representatives, and employers,”12 followed by public hearings and comments.
As we explained in WOC II, “[t]o correspond to curriculum changes, the Legislature required the development of a new state standardized test-the Texas Assessment of Knowledge and Skills (‘TAKS’) test — to replace the Texas Assessment of Academic Skills (‘TAAS’) test.”13 The TAKS test was harder than the TAAS test and covered more subjects.14 Since WOC II, the incorporation of college-readiness standards into the TEKS curriculum led to a new test, the State of Texas Assessments of Academic Readiness (STAAR) test.15 The STAAR testing program is directed at postsecondary readiness. The record is clear that the STAAR test is significantly more difficult than the TAKS test.
Beginning with the 2011-12 school year, the STAAR test was gradually introduced, beginning with Grade 9. The 2014-15 school year was the first in which all high school students took the test. High school students currently must pass five STAAR end-of-course exams in order to graduate: Algebra I, Biology, English I, English II, and United States History.16
The Commissioner of Education (Commissioner)17 sets three performance levels for the STAAR testing program: Level I (unsatisfactory), Level II (satisfactory), and Level III (advanced). Level II was chosen as the college readiness standard. As we noted in WOC II, passing or “cut” scores on the TAKS test, were gradually increased “to give teachers and students *836time to adjust to the more difficult test.”18 Similarly, Level II “phase-in” scores were set for STAAR testing. Standards on the STAAR test are currently set to increase, over three stages, to the final Level II standard in the 2021-22 school year.
In addition to the STAAR testing program, student performance is measured by National Assessment of Educational Progress (NAEP) tests. These tests are administered every two years in math and reading to students in Grades 4 and 8.
Under the system’s accountability regime, the Commissioner adopts rules for evaluating the performance of school districts, campuses, and charter schools.19 Performance ratings are based in part on the school’s or district’s performance on the STAAR tests, dropout rates, and graduation rates.20 Ratings of “Met Standard” (or “Met Alternative Standard” for alternative education campuses and districts) and “Improvement Required” are assigned. The TEA plans to raise performance standards over time.
Each school district must maintain accreditation status. The Commissioner determines accreditation based in part on student achievement indicators (such as testing performance, dropout rates, and graduation rates) and performance under the system’s financial accountability rating system.21 The consequences of a school district losing its accreditation status are substantial. The non-aceredited district may not receive state funding or hold itself out as operating a public school.22 The Commissioner may also pursue various interventions and sanctions against a district or charter school that does not meet accreditation standards, including appointing a monitor, conservator, or management team for the district, closing the district or school, or annexing it to another district.23 The Commissioner can also invoke interventions and sanctions against individual schools within a district.24
2. The School Finance System
Since WOC II, the basic structure for funding the public schools has remained in place, with a few notable changes. Local, state, and federal funds are provided. Local property taxes include an M & O tax, for maintenance and operations, and a facilities tax, referred to as I & S, for “interest and sinking fund.” At the time WOC II was decided, the M & O tax rate was capped at $1.50 per $100 in property value.25
The Foundation School Program (FSP), set out in chapter 42 of the Education Code, is the primary source of funding. The FSP functions to guarantee that each district has “adequate resources to provide each eligible student a basic instructional program and facilities suitable to the student’s educational needs.”26 The FSP consists of two tiers.27 In 2006, the Legislature adopted a “compressed” M & O tax rate that lowered district tax rates by one third.28 The State says the tax compression was in response to WOC II ⅛ holding *837that the system imposed an unconstitutional statewide property tax. The Plaintiffs say the Legislature’s intent was to cut property taxes. Maybe both are true. In any event, a district taxing at the maximum rate of $1.50 saw its rate lowered to $1.00. The Legislature set a new cap of $1.17 for most districts.29 A district that wishes to tax at a tax rate over the “rollback tax rate” of $1.04 must hold a tax rollback election (TRE), also known as a tax ratification election, to approve the rate.30
From 2009 to 2013, for districts taxing at $1.00 or above, the State provided a Tier I basic allotment of $4,765 per student.31 This figure was then adjusted by a cost of education index (CEI) adjustment that accounts for “geographic variation” and “costs of education due to factors beyond the control of the school district,” such as district size, teacher salary variations, and the number of low-income students.32 The basic allotment is also adjusted by a “small and mid-sized district adjustment” recognizing diseconomies of scale for smaller districts, and a “sparsity adjustment” recognizing additional costs (such as additional transportation costs) associated with sparsely populated districts.33 These adjustments result in the district’s adjusted basic allotment.34 The TEA then multiplies the adjusted basic allotment by the district’s average daily attendance (ADA) to determine the district’s “regular program allotment.”35 ADA is the average daily number of students in attendance for the school year.
Tier I funding also includes “special allotments” for certain students, including economically disadvantaged, bilingual, gifted and talented, and special education students, and students requiring transportation.36 This additional funding is provided by multiplying an adjustment factor by the number of students falling into these categories. In today’s case, the adjustment factors for economically disadvantaged and bilingual students are especially important. The adjustment factor for economically disadvantaged students is .2;37 the factor for bilingual students is .l.38 The total of all special allotments plus the regular program allotment is the amount of Tier I funding.
A school district’s Tier I funding is provided by local property taxes and state funds. The district provides funding equal to the product of its compressed tax rate and its property values for the prior year.39 The State then pays the difference between the district’s portion and the total Tier I allotment.40
The system also provides for Tier II funding, which allows districts to obtain *838additional funding by adopting a tax rate between the compressed rate and $1.17.41 Tier II funding is based on the number of students in “weighted average daily attendance” or WADA. A district’s WADA is the sum of its students’ basic allotments plus some special allotments divided by the basic allotment.42 The weighting effect of WADA is to inflate the number of a district’s students in order to account for the needs of students whose education may require more funding.
Tier II funding consists of “golden pennies” and “copper pennies.” Golden pennies are the first six pennies above the district’s compressed rate, and are guaranteed to yield at least the same as Austin ISD’s tax revenue per weighted student per cent of tax effort, which was $59.97 in 2010 and has since been increased to $61.86 for fiscal year 2015, $74.28 for fiscal year 2016, and $77.53 for fiscal year 2017.43 Copper pennies are pennies for the tax rate between the golden-penny range and $1.17, and are guaranteed to yield $31.95 per weighted student.44
Recapture, the “Robin Hood” feature first analyzed in Edgewood Independent School District v. Meno (Edgewood IV),45 remains in place, although it has become more complex. The options available to wealthy districts are essentially the same as those described in Edgewood IV: Districts with property wealth exceeding the statutory limit can consolidate with another district, detach property, purchase average daily attendánce credits (the recapture option), contract for the education of nonresidents, or consolidate their tax base with another district.46 Recapture depends on the tax rate.47 For M & O tax pennies up to a district’s compressed rate, the district is subject to recapture if property wealth per weighted student exceeds certain levels, $476,500 in 2012-13 and $495,000 in 2013-14.48 There is no recapture for golden pennies, hence one reason for their name. For copper pennies, recapture applies to districts with property wealth per weighted student greater than $319,500.49
An alternative to funding “on formula” is known as “target revenue.” Target revenue, described as a “hold harmless” system, is intended to blunt the short-term effects of tax compression. Target revenue is based on certain hold harmless rules and applies if the target revenue amount exceeds the amount the district receives under the provisions described above.50 The State makes up the difference with Additional State Aid for Tax Reduction (ASATR). ASATR is presently only about 1% of total FSP funding, and is set to expire in 2017.
Facilities funding has remained largely unchanged since WOC II. With voter approval, districts may issue bonds for facility construction and renovation, and levy an I & S tax to meet debt-service require*839ments.51 Before a district may issue such bonds it must demonstrate to the Attorney General that it can cover debt service with a tax rate not to exceed 50 cents per $100 valuation.52 The FSP has two programs respecting facilities funding. First, the Instructional Facilities Allotment (IFA) guarantees a yield of $35 per student per penny of tax effort.53 Eligible districts must apply for funding, with low-wealth districts receiving priority, but the Legislature did not appropriate any funding for new IFA grants in 2011 or 2013.54 Second, the Existing Debt Allotment (EDA) guarantees $35 per student for the first 29 cents of tax 'effort needed to make bond payments.55
In 2011, the State cut education funding. According to the State, these cuts were the result of an economic downturn combined with a loss of temporary federal funding under the American Recovery and Reinvestment Act of 2009. Some Plaintiffs suggest the cuts were not the result of an economic downturn, pointing out that the State ran surpluses in 2010-11 and 2012-13 and that the Legislature left its Rainy Day Fund untouched. The underlying reasons for the cuts are debatable, but the amounts of the cuts are not. The Legislature cut FSP funding by $4 billion for the biennium. The Legislature also cut funding for special programs and grants by $1.3 billion. However, the 83rd Legislature restored some of this lost funding in 2013, adding $3.4 billion in FSP funding and another $2.2 billion to adjust for enrollment growth. It brought back $290 million in funding for special programs and grants. The FSP basic allotment was increased to $4,950 for the 2013-14 school year, and $5,040 for the 2014-15 year. The 84th Legislature has funded an additional estimated $1.5 billion to the FSP and added $118 million in new funding for pre-kindergarten (pre-k) programs.
B. The Pending Suit
In this case, various entities and individuals sued the State for constitutional violations regarding the school system. School districts comprise the largest group of plaintiffs. The school district Plaintiffs (the ISDs or ISD Plaintiffs) include the Texas Taxpayer and Student Fairness Coalition (TTSFC) Plaintiffs,56 the Calhoun County ISD (CCISD) Plaintiffs,57 the Edgewood ISD Plaintiffs,58 and the Fort Bend ISD Plaintiffs.59 The ISD Plaintiffs *840alleged the current school finance system violates the adequacy and suitability requirements of article VII, section 1 of the Texas Constitution. They also alleged the system imposes a statewide ad valorem tax in violation of article VIII, section 1-e of the Texas Constitution. With the exception of CCISD, the ISD Plaintiffs also alleged that the system violates the financial efficiency requirement of article VII, section 1. CCISD, a group of property-wealthy districts, sided with the State on this issue.60
Another group of plaintiffs, the Charter School Plaintiffs,61 consists of the Texas Charter School Association and parents of charter school students. The Charter School Plaintiffs claimed the financing system for charter schools violates the adequacy, suitability, and efficiency requirements of article VII, section 1. Their case was consolidated with the related claims filed by the ISDs.
Finally, other parties intervened, alleging the system is unconstitutional under the “qualitative efficiency” requirement of article VII, section l.62 Although the In-tervenors 63 also alleged the system imposes a statewide ad valorem tax in violation of article VIII, section 1-e, they were largely adverse to the ISD Plaintiffs. To the Intervenors, the flaw in the system is not inadequate funding but wasteful spending resulting from inefficient rules and policies.
The trial court conducted a bench trial beginning in October 2012 and ending in February 2013. It reopened the case and took additional evidence in January and February 2014 to consider the impact of 2013 legislation passed by the 83rd Legis*841lature. In August 2014, the court rendered a final judgment declaring the school system constitutionally inadequate, unsuitable, and financially inefficient under article VII, section 1, and unconstitutional as a statewide ad valorem tax prohibited by article VIII, section 1-e. The court also declared the system did not meet constitutional adequacy and suitability requirements for two subgroups of students: English language learner (ELL) and economically disadvantaged students.64 The Education Code defines an ELL student as “a student whose primary language is other than English and whose English language skills are such that the student has difficulty performing ordinary classwork in English.”65 Consistent with section 42.152 of the Education Code, the trial court defined economically disadvantaged students as those eligible for free or reduced-price lunches under the national school lunch program.66 The trial court found that as of the 2012-13 school year, over 60% of Texas’ public school students were economically disadvantaged students and 17% were ELL students.67
The trial court ruled in favor of the Charter School Plaintiffs in holding that funding for charter schools was based on average funding of school districts and was therefore constitutionally inadequate. But the court denied the Charter School Plaintiffs’ separate adequacy, suitability, and efficiency claims that as applied to charter schools specifically the system violated article VII, section l’s requirements. The trial court denied the Intervenors’ claim that the system violated article VII, section 1 on “qualitative efficiency” grounds.
The court signed extensive findings of fact (FOFs) and conclusions of law (COLs) running 364 pages.68 The trial court’s judgment and an executive summary set out in abbreviated form the court’s findings and conclusions. The court enjoined the State from “giving any force and effect to the sections of the Education Code relating to the financing of public school education (Chapters 41 and 42 of the Education Code) and from distributing any money under the current Texas school financing system until the constitutional violations are remedied.”69
The court also granted attorney fees to the ISD Plaintiffs. It denied fee requests filed by the State, Charter School Plaintiffs, and Intervenors. And it retained “continuing jurisdiction” over the case until the State has “fully and properly complied with its judgment and orders.”70
A direct appeal to this Court followed.71
II. Review of the Trial Court’s Decision
A. The Two Key Constitutional Provisions
The trial court held the school system unconstitutional under two provisions.
Article VII, section 1 states:
A general diffusion of knowledge being essential to the preservation of the liber*842ties and rights of the people, it shall be the duty of the Legislature of the State to establish and make suitable provision for the support and maintenance of an efficient system of public free schools.
Article VIII, section 1-e states:
No State ad valorem taxes shall, be levied upon any property within this State.
B. Summary of Our Prior School Finance Decisions
Our analysis depends heavily on our pri- or school finance cases, which comprise a unique body of Texas jurisprudence. Indeed, we have recognized that adherence to our prior decisions is particularly appropriate in this area of the law.72
Here is a chart illustrating the inter-branch conversation, followed below by a summary of the cases:
[[Image here]]
We begin with Edgewood 1.73 There, the Court held the school finance system unconstitutional under the “efficiency” element of article VII, section 1. The basis of this holding was wide disparities in property wealth, tax rates, and spending per student, perhaps most memorably a 700 to 1 ratio between the property wealth per student in the richest and poorest school districts.74 The Court adopted a standard of “financial efficiency” under article VII, section 1 that has been followed since:
Efficiency does not require a per capita distribution, but it also does not allow concentration of resources in property-rich districts that are taxing low when property-poor districts that are taxing high cannot generate sufficient revenues to meet even minimum standards. There must be a direct and close correlation between a district’s tax effort and the educational resources available to it; in other words, districts must have substantially equal access to similar revenues per pupil at similar levels of tax effort.75
*843At the time, the State provided funding under the Foundation School Program (FSP), which continues to this day and distributes money to districts under a complex formula, partly in an effort to equalize funding. But the system at the time did not have the recapture mechanism now in place to further equalize funding.
In Edgewood II,76 the Court addressed the school finance system following the 1990 enactment of Senate Bill 1, the Legislature’s response to Edgewood I. SB 1 left intact the FSP, but created two tiers that remain in place today. Tier 1 was designed to provide guaranteed funding for basic education to all districts that taxed at or above a minimum level. Tier 2 funding guaranteed a certain level of funding for each cent of local tax effort above the Tier 1 minimum level. However, SB 1 excluded the wealthiest 5% of school districts. The Court held this exclusion rendered the system unconstitutionally inefficient:
Most property owners must bear a heavier tax burden to provide a less expensive education for students in their districts, while property owners in a few districts bear a much lighter burden to provide more funds for their students. Thus, Senate Bill 1 fails to provide “a direct and close correlation between a district’s tax effort and the educational resources available to it.” To be efficient, a funding system that is so dependent on local ad valorem property taxes must draw revenue from all property at a substantially similar rate.77
On rehearing, the Court clarified that school districts in an efficient system are permitted to “generate and spend local taxes to enrich or supplement an efficient system.”78 In other words, once an efficient system produces a general diffusion of knowledge, districts can choose to tax and spend additional sums for enrichment or supplementation purposes.
In Edgewood III, 79 the Court addressed the system following the 1991 enactment of Senate Bill 351, a legislative attempt to cure the constitutional infirmities recognized in Edgewood II. SB 351 created 188 county education districts (CEDs), whose “sole function is to levy, collect, and distribute property taxes as directed by the Legislature.”80 The Court held the CED system violated article VIII, section 1-e by creating a statewide ad valorem tax. The CED system created such a tax because “the very purpose of the CEDs is to levy a uniform tax statewide.” 81 “If the State mandates that- a tax be levied, sets the rate, and prescribes the distribution of the proceeds, the tax is a state tax, regardless of the instrumentality which the State may choose to use.”82 Edgewood III held:
An ad valorem tax is a state tax when it is imposed directly by the State or when the State so completely controls the levy, assessment and disbursement of revenue, either directly or indirectly, that the authority employed is without meaningful discretion.83
*844In Edgewood IV,84 the Court addressed the constitutionality, of Senate Bill 7, enacted in 1993 in response to Edgewood III. SB 7 retained the two-tier system. Tier l’s purpose was to provide a basic program meeting all legal standards, while Tier 2 provided additional funding for enrichment and facilities under a guaranteed yield system. Subject to exceptions for debt repayment, SB 7 capped combined Tier 1 and Tier 2 tax rates at $1.50. A new feature was a cap on taxable property above $280,000 per student. If a district’s property wealth exceeded that amount, the district could consolidate with another district, detach property, consolidate its tax base, contract for education of nonresi- . dents, or purchase “average daily attendance credits.” The last option was the one used by almost all districts whose property wealth exceeded the cap, requiring a pay- , ment to the State that was redistributed to poorer districts, the original recapture or “Robin Hood” mechanism.85 The Court rejected all constitutional challenges. It held the system met the financial efficiency requirement of article VII, section 1. It noted the 700 to 1 ratio of wealth per student found in Edgewood I had been reduced to 28 to 1, with further reductions under Tier 2.86 Although it did not analyze “adequacy” using that term, it concluded the system’s accountability regime met “the Legislature’s constitutional obligation to provide for a general diffusion of knowledge.”87
In Edgewood IV, the Court also rejected the wealthy districts’ argument that SB 7 unconstitutionally imposed a statewide ad valorem tax. The Court held that although the statute set maximum and minimum tax rates, and provided financial incentives for -districts to tax at maximum rates, enough discretion remained in the system to distinguish it from a statewide property tax. But the Court warned:
Eventually, some districts may be forced to tax at a maximum allowable rate just to provide a general diffusion of knowledge: If a cap on tax rates were to become in effect a floor as well as a ceiling, the conclusion that the Legislature had set a statewide ad valorem tax would appear to be unavoidable because the districts would then have lost all meaningful discretion in setting the rate.88
In WOC 1,89 the Court addressed arguments that the system had indeed come to impose a statewide property tax. The trial court had dismissed a statewide property tax claim on the pleadings, concluding that the claim was not viable unless “something approaching or exceeding half the districts” in the State were required to tax at the $1.50 cap to provide an accredited education.90 We adhered to the Edgewood III test for a statewide ad valorem tax: “An ad valorem tax is a state tax when it is imposed directly by the State or when the State so completely controls the levy, assessment and disbursement of revenue, either directly or indirectly, that the authority employed is without meaningful discretion.”91 We held that because the Constitution prohibits such a tax on “any *845property within the State,” it did not matter whether such a tax was imposed on all or most districts, or even a single district. Thus, the plaintiffs had stated a claim if they alleged that they had no meaningful discretion in setting tax rates because they were forced to tax at or near the $1.50 cap to provide a general diffusion of knowledge.92
Finally, in the most recent school finance case, WOC 17,93 the Court again addressed challenges to the school finance system under SB 7. It rejected the State’s standing and justiciability arguments.94 Synthesizing our prior caselaw, WOC II explained that article VII, section 1 sets out three standards: efficiency, adequacy, and suitability.95 Efficiency consists of both a “qualitative” component and a “quantitative” component, the latter also referred to as “financial efficiency.”96 The Court held that the system met the adequacy requirement for providing a general diffusion of knowledge, and that the system was also constitutionally efficient and suitable. But we acknowledged substantial evidence “that the public education system has reached the point where continued improvement will not be possible without significant change,” and that “it remains to be seen whether the system’s predicted drift toward constitutional inadequacy will be avoided by legislative reaction to widespread calls for changes.”97 And the Court again held the system imposed a statewide property tax in violation of article VII, section 1-e.98 As in prior cases, the Court focused on whether districts had meaningful discretion in setting rates and whether the legislatively imposed $1.50 cap had become “a floor as well as a ceiling.”99
In sum, the Court has twice held, in Edgewood I and Edgewood II, that the school finance system was unconstitutional under article VII, section 1 because it was financially inefficient. It has twice held, in Edgewood III and WOC II, that the system was unconstitutional under article VIII, section 1-e because it imposed a statewide ad valorem tax. The Court has never held the system constitutionally inadequate, unsuitable, or “qualitatively” inefficient under article VII, section 1.
C. Standards of Review
Philosophical views and emotions run deep on the subject of school funding in this State, as indicated by the large number of parties and amici curiae' and the content of their briefing. We embrace this input because all Texans, judges included, desire an efficient school system that cultivates and maximizes student achievement.100 But this Court must always be mindful of its limited constitutional role in this controversy. Our role is consequential yet confined, strictly circumscribed by a deferential standard of review, as well as *846our own prior decisions in this unique area of the law.
In this direct appeal, we have no jurisdiction “over any question of fact,” 101 and must “rely entirely on the district court’s findings.”102 But in deciding the constitutional issues, “those findings have a limited role.”103 Whether the public school system is constitutional is ultimately a question of law.104 And under our settled precedent, which frowns upon judicial second-guessing of policy choices, we presume the system is constitutional.105
In. WOC II, we recognized an arbitrariness standard for challenges under article VII, section l.106 Under this “very deferential” standard,107 we must not substitute our policy preferences for the Legislature’s, but “must on the other hand examine the Legislature’s choices carefully to determine whether those choices meet the requirements of the Constitution.”108 “If the Legislature’s choices are informed by guiding rules and principles properly related to public educatiori — that is, if the choices are not arbitrary — then the system does not violate the constitutional provision.” 109 At bottom, the “crux” of this standard is “reasonableness,” and the lens through which we view these challenges maintains a default position of deference to the Legislature — that political branch responsible for establishing a constitutionally compliant system.110
D. Justiciability, Standing, and Ripeness
The State argues we should not reach the merits of the Plaintiffs’ claims but should instead dismiss the case for various jurisdictional reasons. We reject these arguments.
The State primarily invokes the political-question doctrine, arguing the claims under article VII, section 1 involve policy-making reserved to the Legislature and are nonjusticiable political questions. The State says there are no judicially manageable methods for assessing the constitutionality of the state educational system, and there is no practicable basis for assessing the “reasonableness” of such a complex system.
We rejected this argument in Edgewood I, and do so again today. We explained that the Constitution imposes standards that are not committed unconditionally to the Legislature, but are instead subject to judicial review.111 Although the “imprecise” language • of Article VII, section 1 necessarily grants the Legislature great discretion to determine what constitutes “suitable provision” for an “efficient system” to provide a “general diffusion of knowledge,” it is not inherently the Legislature’s role to define and interpret the Constitution. To the contrary, as we have *847explained throughout our school finance cases, courts have the ultimate authority to determine whether the Legislature’s interpretation of these terms is arbitrary or unreasonable and, therefore, unconstitutional. In Edgewood I, for example, we explained that the terms used in article VII, section 1 “provide a standard by which this court must, when called upon to do so, measure the constitutionality of the legislature’s actions.”112 Similarly, in WOC I, we affirmed that the “final authority to determine adherence to the Constitution resides with the Judiciary,” and although “the Legislature has the sole right to decide how to meet the standards set by the people in article VII, section 1, ... the Judiciary has the final authority to determine whether they have been met.”113 And in WOC II, we made clear that “the Constitution nowhere suggests that the Legislature is to be the final authority on whether it has discharged its constitutional obligation.”114 “If the framers had intended the Legislature’s discretion to be absolute, they need not have mandated that the public education system be efficient and suitable; they could instead have provided only that the Legislature provide whatever public education it deemed appropriate.” 115
The Legislature’s discretion, in other words, is “not without bounds.”116 Although the determinations of what constitutes a “general diffusion of knowledge” and how to provide such an education can be said to involve “political questions,” they are not nonjusticiable political questions.
On another front, the State argues that the Plaintiffs do not have standing to make claims under article VII, section 1 because the courts cannot provide the relief sought, namely legislative changes including changes in funding. The Court rejected a similar standing challenge in WOC II, holding that “being required to implement unconstitutional statutes” is sufficient to give the ISDs standing to assert constitutional violations.117 The ISDs also argue that the Intervenors lack standing. The Intervenors include individual students who have standing because they are allegedly suffering directly by enduring an education provided by an unconstitutional system.118
To be sure, the State focuses more precisely than before on the “re-dressability” component of standing, but we remain persuaded that the Plaintiffs have standing. Generally, standing under Texas law “requires a concrete injury to the -plaintiff and a real controversy between the parties that will be resolved by the court.”119 On the question of the court’s capacity to fashion relief, we have recognized that a plaintiff lacks standing if the requested injunctive relief “could not possibly remedy his situation.” 120 In today’s case as in prior decisions, the trial court’s relief was to enjoin the Legislature from funding the school system until the *848Legislature cured the constitutional infirmity.121 In fact, the court’s injunction used language identical to language used in the WOC II injunction that we affirmed.122 While the trial court could not write its own substitute legislation,123 we cannot say that for standing purposes the relief granted by the trial court could not possibly spur the Legislature to act, as it has done in the past in response to court decisions. After all, the Legislature has acted each time this Court has declared the school finance system unconstitutional.124 The Plaintiffs have standing.
Finally, the State argues that the Plaintiffs’ claims, including the state property tax claims under Article VIII, section 1-e, are not ripe because school financing has changed over time. In particular, the State argues the evidence introduced in the first phase of the trial relating to the funding cut was superseded by the 83rd Legislature’s decision to restore most of the funding in 2013. We have not previously addressed a ripeness argument of the sort the State now presents, but we reject it as well. Generally, the ripeness doctrine concerns whether there is sufficient development of the facts and issues to ensure that the court’s judgment is not based on contingent or uncertain events.125 Some of the evidence presented to the trial court might have been mooted by subsequent events, including the restoration of some of the funding. But as the Plaintiffs correctly point out, the trial court reopened the evidence and made new findings. The State argues that the legislative changes considered when the evidence was reopened were so new that relevant data on their effectiveness was unavailable. Trial court findings based on the new evidence may not be as persuasive as the earlier ones, but the changed circumstances do not render the' entire trial court proceeding unripe.
The State’s advocacy for a strict application of the ripeness doctrine would mean school finance cases could never be entertained, because the facts relating to funding, test scores, tax rates, property values, etc., are always changing to some extent. This is not to say that changed circumstances are not problematic for the Plaintiffs, for reasons discussed below. But the inevitable changes in relevant factual circumstances do not place school finance cases completely beyond the decision-making reach of the courts; again, holding otherwise would effectively overrule our longstanding recognition that the courts *849play a legitimate, constitutionally authorized role in these disputes.
E. Adequacy
Article VII, section 1 imposes a duty on the Legislature to achieve a “general diffusion of knowledge.” This requirement has come to be known as the “adequacy” requirement. The school system is constitutionally adequate if it achieves a general diffusion of knowledge.126 Deferring to the Legislature’s role in providing a general diffusion of knowledge, we have looked to sections 4.001(a) and 28.001 of the Education Code (quoted above), through which the Legislature has declared that its obligation to provide a general diffusion of knowledge is satisfied if school districts are reasonably able to provide their students with (1) access to a quality education that enables them to achieve their potential and fully participate now and in the future in the social, economic, and educational opportunities of Texas and the nation, and (2) a meaningful opportunity to acquire the essential knowledge and skills, reflected in curriculum requirements, such that upon graduation they are prepared to continue to learn in postsecondary educational, training, or employment settings.127 We did not adopt these as the Court’s standards for a constitutional education, or the final word on constitutionally required standards. Rather, we recognized this to be “the Legislature’s understanding of a general diffusion of knowledge.”128 As the policy-making branch of government, the Legislature may retain, revise, or replace these provisions, to reflect its current view of what the required curriculum, and thus a “general diffusion of knowledge,” should produce.
In our last three school finance cases, we have made clear that we presume the Legislature achieves a general diffusion of knowledge by devising a curriculum and an accountability regime to meet legislatively designed accreditation standards for schools and districts.129 Again, that presumption is not irrebutta-ble, and is subject to judicial review, but our review of the adequacy requirement under the arbitrariness standard is always “very deferential.”130
Perhaps in contrast to earlier lawsuits, the issue of adequacy looms especially large in today’s case. CCISD accurately describes the issue as “now at the heart of this case.” We conclude that the district court’s analysis of this issue was flawed, and its ultimate determination of constitutional inadequacy wrong. This error, unfortunately, bleeds over into other issues and infects much of the trial court’s analysis of them, as detailed below.
1. The Trial Court’s Reliance on Spending Levels
The trial court determined the school system was constitutionally inadequate because it agreed with three of the Plaintiffs’ experts that the system was underfunded. The court endorsed Allan Od-den’s opinion as providing a reasonable estimate of the cost of an adequate education. Odden used an “evidence-based model” to calculate the cost of an adequate educational system. He calculated the cost by considering “best practices” such as a core class size of 15 for kindergarten through third grade and 25 for other *850grades, full-day kindergarten, instructional coaches to provide professional development, tutors and summer school for struggling students, including one-on-one tutors for some students, and other “key strategies.” 131 Based on Odden’s calculations, combined with the cost of providing transportation, food services, and security as calculated by another expert, Lynn Moak, the trial court concluded that a reasonable estimate of the cost of an adequate education for the entire school system in 2010-11 was $46,766,551,937, a figure $3.66 billion above the amount spent that year and, after budget cuts, $6.16 billion annually above the amount budgeted for the following two years.132 The court also found reasonable Moak’s opinion that about $1,000 of additional funding per weighted student was necessary to correct outdated weights and adjustments and to allow schools to meet increased state standards.133
The court found additional support for Odden’s and Moak’s opinions in the testimony of Bruce Baker, who took the $3,500 figure per weighted student that was supposedly required to achieve a general dif- ' fusion of knowledge in Edgewood IV, and adjusted that figure for inflation. This yielded a figure of $6,576. The court concluded that only a small percentage of districts could raise this amount even taxing at the maximum rate of $1.17.134
Relying on these experts, the trial court measured the level of inadequate funding on a per student basis. It found that the total spending needed for 2010-11, measured in FSP M & O dollars per student in WADA, was $6,176 per Odden’s estimates, $6,562 per Moak’s estimate, and $6,576 per Baker’s estimate, all well in excess of the actual funds available, estimated at $5,662.135 The court also updated these figures for later years, finding for example that the inflation-adjusted Odden estimate grew to $6,404 for 2013-14, and $6,532 for 2014-15.136 The trial court held the system was unconstitutional because spending was below the “adequacy estimates” of these experts.137 It found that these estimates “provide a credible range that definitively establishes that the State has failed to make suitable provision of funds for an adequate education.”138
We do not agree with the trial court’s analysis for several reasons. First, we stated in WOC II that an adequacy determination should not depend on “inputs” such as funding per student; instead, the determination “is plainly result-oriented,” looking to “the results of the educational process measured in student achievement.” 139 Yet, in the trial court’s judgment, all seven of its declarations regarding adequacy are expressly based on the input of inadequate funding.140 To place *851so much reliance on expert testimony as to the specific amount of funding needed was therefore misguided.
Second, the trial court’s “fact” findings as to the specific amount of funding needed to achieve a general diffusion of knowledge are, we think, beyond the current state of science in this field. We have warned that in school finance cases where we must decide constitutional questions, the trial court’s findings play a “limited role.”141 This case demonstrates why. To determine as a matter of fact that specific funding levels are required to achieve the constitutional threshold of a general diffusion of knowledge, a court not only must find that a cost-quality relationship exists, but also must assign specific quantitative measures to that relationship. Even the general, qualitative question of the existence of a cost-quality relationship remains a highly contested issue in the social sciences. Perhaps for this reason, we have ourselves spoken only obliquely and in general terms on this issue, stating for example in Edgewood I that “[t]he amount of money spent on a student’s education has a real and meaningful impact on the educational opportunity offered that student,”142 but recognizing in WOC II, our latest school finance decision, that “more money does not guarantee better schools or more educated students.”143 We have never sanctioned a trial court’s ordering the Legislature to spend a specific amount of money on the schools to achieve constitutional adequacy, as doing so would deprive the Legislature of the broad discretion the Constitution provides for such inherently political decisions.
We are not alone in expressing uncertainty as to the correlation between more money and better education. In 1973, the United States Supreme Court rejected a federal constitutional challenge to the Texas school finance system in San Antonio Independent School District v. Rodriguez.144 The Court noted:
On even the most basic questions in this area the scholars and educational experts are divided. Indeed, one of the major sources of controversy concerns the extent to which there is a demonstrable correlation between educational expenditures and the quality of education- — -an assumed correlation underlying virtually every legal conclusion drawn by the District Court in this case.145
The Court cited numerous conflicting studies, including the Coleman Report,146 a famous study required by section 402 of the Civil Rights Act of 1964 and often cited for finding a weak correlation between public school funding and educational achievement. The Coleman Report, “widely considered the most important ed*852ucation study of the 20th century,” 147 surveyed 4,000 public schools- and over 645,-000 students,148 and runs over 700 pages. Its primary conclusion is this:
The first finding is that the schools are remarkably similar in the way they relate to the achievement of their pupils when the socioeconomic background of the students is .taken into account. It is known that socioeconomic factors bear a strong relation to academic achievement. When these factors are statistically controlled, however, it appears that differences between schools account for only a small fraction of differences in pupil achievement.149
According to one retrospective on the Report, “Coleman’s findings indisputably documented that variation between schools in their resource levels mattered little for variation among individual students, a result that remains the seminal finding in U.S. sociology of education.... Forty years on, the findings of the Coleman re•port hold up remarkably well, in some ways distressingly so.”150
We express no opinion on the 50-year-old Coleman Report, nor, frankly, are we qualified to do so. But the controversy hardly ended with' the Coleman Report. Hundreds of similar studies, reaching conflicting conclusions, have followed.151 Some amici curiae have filed Brandéis briefs citing recent studies going both ways on the issue of whether more spending means a better education.152 But the *853trial court in effect decided to end that ongoing debate and declare the school system unconstitutional because spending was below a specific price point it deemed necessary to achieve a general diffusion of knowledge.
Courts should not sit as a super-legislature. Nor should they assume the role of super-laboratory. They are not equipped to resolve intractable disagreements on fundamental questions in the social sciences. Arthur Miller may have referred to a trial as the crucible, but we doubt he saw it as the best place for reducing scientific truth when the scientific community itself has reached an impasse.
Another problem with placing a dollar figure on adequacy is the trial court’s reliance on footnote 10 of Edgewood TV.153 In FOF 622, the trial court stated that in Edgewood TV a general diffusion of knowledge required “about $3,500 per weighted student.” An expert, Dr. Baker, adjusted this figure for inflation in education costs and calculated an equivalent amount of $6,576 in 2011. Using this figure, the trial court found in FOFs 623 and 624 that fewer than a quarter of districts could raise this amount even if taxing at the maximum rate of $1.17. In FOF 624 the court recognized that the $3,500 figure was only a “rough approximation,” but nevertheless found it supportive of the other dollar figures reached by Odden and Moak. In reaching its conclusions that funding was constitutionally inadequate, the trial court expressly relied on “the amount of spending the courts have found necessary to achieve the general diffusion of knowledge in the past.”154
The problem with this analysis is that we never held in Edgewood TV that $3,500 per student was required to achieve a general diffusion of knowledge. We most assuredly did not decree that the Legislature must at a minimum fund at this level, forever and a day, with adjustment under the latest and best index for measuring inflation in the cost of education. Our reference to this figure did not even occur in a discussion of the adequacy requirement. Rather, we mentioned the figure in our discussion of financial efficiency, a separate requirement that focuses on equality of funding and not on the absolute level of funding. We assumed the trial court was correct that the $3,500 level was the minimum required for a basic education, and then held that financial efficiency standards were met, explaining that the equal access to funding required for financial efficiency only applies to “the funding necessary for a general diffusion of knowledge” and did not apply to supplemental funding for enrichment. A $600 advantage in funding per student available to the richest districts taxing at a $1.50 rate was not dispositive, because taxing at lower rates would bring in the $3,500 per stu*854dent.155 We found the. disparity acceptable given that we were only concerned with the disparities in funding and tax rates needed to provide a general diffusion of knowledge. We did not separately decide whether, under an adequacy challenge, the $3,500 figure was in fact the amount needed to provide an adequate education. That question was not before us.156 Instead, Edgewood IV says almost in passing, in an earlier discussion, that the system provided a general diffusion of knowledge because of the accountability regime the Legislature set up.157
It is important to see that, in Edgewood IV, we found the $3,500 figure relevant to support the Legislature’s discretionary choices in providing the system’s financial efficiency. We accepted that figure as the level of funding necessary to provide a general diffusion of knowledge, to identify the ceiling up to which the Constitution required substantially equal access to “efficient” funding. In today’s case, the trial court used the figure as a floor to create a constitutional minimal funding amount, and establish a violation of the adequacy requirement. We think that use is very different from the use we made of it in Edgewood IV, and is misplaced.
We do not question that a school system must spend money to accomplish a general diffusion of knowledge. Common sense says as much, as have we.158 Our financial efficiency doctrine presupposes that some good comes from equalizing access to funding. But here the trial court went much further, embarking on a quest to calculate the statewide dollar cost of an adequate education, and declaring the system unconstitutional because the Legislature had not provided funds to meet that threshold. What is not clear, given the current state of knowledge in the social sciences, is that spending a specific amount of additional money necessarily correlates to a better education as measured by objective outcomes. Nor is it clear that the specific cost of a constitutionally adequate education for the entire State can reasonably be determined by a court and therefore justifiably imposed on the Legislature as a constitutional mandate. In defense of the trial court’s calculation of the minimum dollar figure needed for a constitutionally adequate education, the Charter School Plaintiffs argue that “the State did not present an alternative legal minimum funding level.” CCISD similarly argues that the State “offered no estimate of the costs of a constitutionally adequate education at trial.” But such argument begs the question vexing us, whether the state of human knowledge makes such a calculation possible, much less constitutionally required.
Yet another problem with Odden’s testimony is that he based his dollar figures wholly on what he determined to be “best practices,” without regard for or deference to the Legislature’s chosen practices. As *855explained further immediately below, the Constitution only requires the State to provide an “adequate” education, not the “best” education available through the use of various inputs deemed “best practices” by an expert.
By focusing so heavily on the input of spending, attempting to decide a fundamental question that remains unresolved in the social sciences, relying on a misinterpretation of this Court’s jurisprudence, and relying on what the court deemed “best practices,” the trial court erred in assigning a minimum dollar figure as constitutionally necessary to achieve a general diffusion of knowledge. This error infected the entire adequacy analysis, influencing the trial court over and over, and rendering its ultimate conclusion that the school system is constitutionally inadequate hopelessly flawed.
2. The Trial Court’s Focus on Other Inputs and “Best Practices”
The trial court made many findings and conclusions that can only be characterized as findings of inadequate inputs, relating to class size, tutoring, interventions for special needs students, nurses, security guards, etc.159 The court made these findings statewide, as to some 36 individual districts, and as to ELL and economically disadvantaged students. The trial court’s heavy reliance bn what it found to be inadequate inputs was in our view flawed. Moreover, even if we sanctioned a focus on inputs generally, many of these findings do not appreciate that the constitutional standard demands not the best education, but only an educational system that is adequate to provide a general diffusion of knowledge.
We have never held that constitutional adequacy requires the State to employ what are, in the view of one expert or another, the “best practices” recognized by a segment of the expert community. Funding questions aside, the trial court strayed by repeatedly looking to “best practices” regarding pre-k programs, school size, dual language classes, class size, providing for the needs of ELL and economically disadvantaged students, and so on, and holding that a failure to implement such practices “could be considered arbitrary and unconstitutional.”160 The *856trial court noted, in FOFs 612 and 615, that Dr. Odden looked to best practices in arriving at his dollar figures for an adequate education.161 The court accepted Odden’s analysis as providing “a reasonable estimate of the cost of an adequate education in Texas,”162 and expressly found that best practices “are credible factors for determining the cost of education.” 163 But the Legislature is not constitutionally required, under an inputs approach to adequacy we have rejected, to assure that districts statewide impose specific inputs in the form of myriad best practices.
3. The State’s Failure to Calculate Education Costs
Continuing their input-based approach, the Plaintiffs argue that the educational system is inadequate because the State has failed to make its own calculations of the funds needed to meet its performance standards or to obtain a general diffusion of knowledge. They rely on section 42.007 of the Education .Code, which provides in part:
(a) The Legislative Budget Board shall adopt rules, subject to appropriate notice and opportunity for public comment, for the calculation for each year of a biennium of the qualified funding elements, in accordance with Subsection (c), necessary to achieve the state policy under Section 42.001.
(b) Before each regular session of the legislature, the board shall, as determined by the board, report the equalized funding elements to the commissioner and the legislature.
Subsection (c) then sets out various funding elements. Section 42.001 describes general state educational policies such as the policy that “each student enrolled in the public school system shall have access to programs and services that are appropriate to the student’s educational needs.”
The Plaintiffs say that section 42.007 requires the State to make its own adequacy calculations. Fort Bend ISD, for example, argues that section 42.007 “is the way *857the Legislature can evaluate, objectively, whether the State is meeting its constitutional duty to structure, operate, and fund a public school system that spreads knowledge diffusely.” The trial court agreed with the Plaintiffs, citing section 42.007 for the proposition that the State has a “constitutional responsibility” “to make a reasonable effort to determine what it will cost to adequately provide for its own standards and meet its own definition of a general diffusion of knowledge.”164
The State does not deny that the Legislative Budget Board has failed for years to comply with section 42.007. But even assuming that section 42.007 is a statutory mandate for the Legislature to calculate the level of funding needed to provide for a general diffusion of knowledge, this failure does not establish a constitutional violation of the adequacy requirement.
As explained above, the spending level is an input that generally is not decisive in determining whether the State is providing a general diffusion of knowledge. Again, we have made clear that the approach should be output driven. Given the highly controversial issue of whether more spending will necessarily raise student achievement, a debate that has raged for decades in the social sciences, we cannot say that the State has acted in an unconstitutionally arbitrary manner simply because it has not attempted to assign precise dollar values as to the costs of providing a general diffusion of knowledge. One could argue with equal persuasiveness that attempting to assign such dollar values is itself arbitrary, given the uncertainty that prevails in this realm of educational research.
To be sure, the better practice might be for the Legislature to regularly calculate the cost of a general diffusion of knowledge, or components thereof, particularly in light of. section 42.007. But complaining that the State has not come up with its own dollar figures for meeting legal mandates for public education does not render the system constitutionally inadequate, because the Plaintiffs bear the burden of proving the system does not achieve a general diffusion of knowledge.165 A failure to comply with what is arguably a creative accounting requirement does not shift the burden of proof and create a presumption that the underlying system is, as a constitutional matter, failing to provide an adequate education to Texas students.
Further, the State offered evidence that, through the Academic Excellence Indicator System, the State generates exhaustive, publicly available reports at the campus, district, region, and state levels on students, staff, programs, and expenditures.166 While additional data might be useful, the Plaintiffs did not demonstrate a system so devoid of data that the lack of information amounts to a constitutional violation.
4. Statutory Weights and Adequacy Claims Relating to Subgroups
The trial court separately held that, due to inadequate funding, the school finance system was constitutionally inadequate and unsuitable as to ELL and economical*858ly disadvantaged students.167 We must decide whether the courts should entertain Article VII, section 1 claims as to subgroups, as urged especially by the Edgewood ISD Plaintiffs. They point to troubling findings as to .economically disadvantaged and ELL students, such as performance gaps that are large and growing. They argue the State must target more funding toward bridging these gaps.
The performance of student sub-populations, especially a large group such as economically disadvantaged students who comprise more than 60% of students, .is relevant to whether the system as a whole is constitutionally adequate. But this Court has never squarely held that a separate, cognizable adequacy claim can be asserted by a student subpopulation such as economically disadvantaged or ELL students. In reviewing an adequacy claim in WOC II, the Court noted “wide gaps in performance among student groups differentiated by race, proficiency in English, and economic advantage.”168 This observation suggests that the Court could in theory find constitutional inadequacy as to specific groups. But the Court- did not proceed to decide adequacy as to any particular subgroup in WOC II and ultimately found no adequacy violation. And the Court also said that article VII, section 1 only requires “an efficient system of free public schools, considering the system as a whole, not a system with efficient components.” 169 It asked whether “the system as a whole is providing for a general diffusion of knowledge.”170 The State also points out that article VII, section 1 only requires a “general diffusion of knowledge,” not a diffusion of knowledge to particular groups.
*859We do not today foreclose completely a ruling of constitutional inadequacy as to subgroups, but conclude that the showing necessary for such a ruling would have to be truly exceptional, for several reasons. First, as noted the Constitution itself and our own decisions suggest that we are typically confined to asking whether the system as a whole is providing a general diffusion of knowledge.
Second, for practical reasons we are reluctant to start down the path of entertaining claims of subpopulations. Here we have claims on behalf of ELL and economically disadvantaged students. Similar claims could be made for countless other groups—students in certain grades or of certain ages, students of certain races, boys versus girls, students with certain mental or physical disabilities, etc. The only practical limitation on such claims, beyond creativity, might be the availability of data, but if the record in this case is any indication, a staggering amount of data exists on the educational system and its students. If the Court starts down that path, then truly perhaps “there is no end in sight.”171 Some resort to practical concerns is appropriate in constitutional deci-sionmaking, because our Constitution’s framers were practical people with practical concerns and intentions. In describing the federal Constitution,' the United States Supreme Court has said, “It must be remembered that the framers of the Constitution were not mere visionaries, toying with speculations or theories, but practical men, dealing with the facts of political life as they understood them [.] ”172 Surely the same can be said of our State’s framers and our own Constitution.
The Plaintiffs relied heavily at trial on achievement gaps themselves as the reason funding should be increased for these groups. For example, Moak testified that the funding multipliers for economically disadvantaged and ELL students should be doubled because of a persistent education gap. The trial court, too, concluded that the school system’s failure to close achievement gaps of economically disadvantaged and ELL students established a constitutionally inadequate system as to these groups.173 But differences in achievement among subgroups do not necessarily establish a failure of the school system in its allocation of resources. The Coleman Report, for example, though not the final word on this issue, concluded that factors distinguishing the students themselves accounted for vastly larger differences in achievement than differences in the resources provided by the school system.174 The Plaintiffs *860concede that economically disadvantaged students face challenges outside the schools that affect their educational achievement, and indeed offered much evidence to the trial court in support of this position. According to the Edgewood Plaintiffs, “The challenges that economically disadvantaged students face stem largely from the opportunities they have available to them where they live.” This seems inarguable. Demography is not destiny. Many of our most celebrated achievers — in every walk of life — overcame tough odds to thrive, conquering headwinds galore. But factors outside the classroom play an undeniable role in many children’s lives. The Coleman Report, as discussed above, reached the seismic conclusion a half-century ago that family-related variables, for example, matter more — far more — than per-pupil expenditures when it comes to predicting academic success. The Plaintiffs presented much data on achievement gaps of ELL and economically disadvantaged students, but did not prove that those gaps could be eliminated or significantly reduced by allocating a greater share of funding to these groups. Again, as we have recognized, “more money does not guarantee better schools or more educated students.”175
In the same vein, the Plaintiffs complain that weights assigned for funding economically disadvantaged and ELL students are inadequate. They complain of the .2 statutory multiplier for economically disadvantaged students176 and the .1 multiplier for students in bilingual and special language programs177 as inadequate, pointing out that these multipliers have not changed since 1984. They also complain that other adjustments such as the cost of education index (CEI), and the adjustments for districts with low populations or low population density, have not changed for many years.178 But the fact that a multiplier has not changed over time in itself proves little. Changes in real-world conditions over time may warrant an upward or downward revision to a multiplier. This Court, obviously, does not assume that an old decision is necessarily wrong, else we abolish the doctrine of stare decisis. One of the Plaintiffs’ lead experts, Allan Odden, relied heavily on the Tennessee STAR study conducted in the 1980s to support his opinions regarding class size, as well as studies going back to the early 1960s to support his opinions on full-day kindergarten. Yet the trial court was' willing to rely on Od-den’s assessments of best practices in arriving at the cost of an adequate education.
Moreover, multipliers may seem inadequate if considered in isolation, but they sometimes have a cumulative effect. For example, a poor immigrant student’s status may boost district funding through three separate adjustments — the multipliers for economically disadvantaged and ELL students, and the CEI adjustment that separately considers the number of low-income *861students in the district.179
In arguing for adjustments that give proportionately more to certain subgroups, the Plaintiffs offered evidence, including expert testimony, that their favored subgroups benefitted from certain programs and that the State’s multipliers are too low. For example, they offered evidence that economically disadvantaged students benefitted from access to pre.-k programs and smaller class sizes.
This argument can be viewed two ways. One understanding is that the Plaintiffs are in effect arguing that, for any level of total funding, certain groups deserve a larger piece of the pie. The Plaintiffs are hard put to justify this result as necessary to improve “the system as a whole” unless they can show that the achievement gains to the allegedly underfunded subgroup will more than offset the losses that other students will sustain if they receive less funding. That conclusion does not necessarily follow from evidence that certain disadvantaged groups are more expensive to educate or benefit from certain targeted interventions such as . pre-k programs or smaller classes, evidence that was the subject of numerous trial court findings. For a given sum of money, the achievement gains, as measured by objective outputs such as test scores, for the subgroup favored by a plaintiff may be larger, smaller, or the same as the gains achieved by spending that same sum in targeted funding on other groups. Plaintiffs’ expert Dr. Baker, for example, conceded that “[h]av-ing access to more resources always gives the opportunity to do more with it, whether you’re talking about in a high poverty community or, a high wealth community.” One might expect any student could benefit from one improvement or another made available with additional spending, given the trial court’s 1,508 fact findings that virtually everything about the current system is broken.
An alternative understanding of the Plaintiffs’ argument is that favored and disfavored subgroups both deserve more funding. But this argument merely recasts the argument for more dollars, the weaknesses of which are discussed above. The trial court concluded that adjustments for certain subgroups were insufficient, but it also held that funding for the system as a whole was insufficient. For example, the court relied on Moak’s estimate that an updated cost of education index (under section 42.102 of the Education Code) “should provide approximately $1 billion more to school districts.”180 If the court was merely relying on its view that more money was needed, this finding suffers from the same problems we have already discussed regarding the court’s efforts to tie specific dollar amounts to a greater diffusion of knowledge. Again, as we have recognized, “more money does not guarantee better schools.”181 Alternatively, if it is all a zero-sum game, the court and the Plaintiffs did not take the crucial step of showing that diverting a billion dollars from one group and giving it to another would measurably improve the system’s overall diffusion of knowledge.
Stated another way, the Plaintiffs would have a better argument if they had shown that their selected subgroups — ELL and economically disadvantaged students here — see greater improvements in test scores or some other quantifiable measure of achievement for each dollar of additional spending than other groups see, and that these subgroups therefore deserve a disproportionate share of existing or addition*862al funding. The Plaintiffs failed to make such showings. Indeed, according to one of the State’s experts, Dr. Michael Podgur-sky, such a showing is impossible because no one knows “the best relationship between inputs and outputs.” He took issue with questioning “assuming someone knows how to measure — someone knows the truly efficient way to raise student achievement and use the inputs in the most efficient way possible to produce the outputs, and I don’t think anyone knows that ... you’re asking me to do something I don’t think anyone can do.” Presumably, a very precise knowledge would be required to target specific dollar inputs for one subgroup and thereby raise their achievement, without correspondingly lowering the outputs of other groups who would be deprived of funding, and Podgur-sky appeared to think that no such knowledge exists. The trial court agreed with Podgursky’s assessment.182
The trial court nevertheless concluded that the school finance system as a whole, and as to economically disadvantaged and ELL students in particular, was unconstitutionally unsuitable due to inadequate funding weights assigned to those two subgroups, and more generally because of the system’s reliance on “outdated, arbitrary weights and allotments that do not reflect the actual cost, of education to determine funding levels- for districts.”183 The court relied heavily on the testimony of Moak, who thought the multipliers for ELL and economically disadvantaged students should be doubled, though Moak conceded to having no data in his report that marginal increases in funding for these groups would increase their performance.
The trial court concluded that economically disadvantaged and ELL students particularly benefitted from some strategies such as smaller classes in lower grades,184 but that does mean that other groups would not benefit equally from spending on other strategies that particularly benefit those other groups. If, after all, funding is the key to an adequate education, as the trial court found, then those students receiving less will suffer if funds are diverted to Plaintiffs’ selected subgroups. But article VII, section 1 requires “that the public school system be structured, operated, and funded so that it can accomplish its purpose for all Texas children.”185
If the Plaintiffs are arguing that economically disadvantaged and ELL students are entitled to a greater share of funding because performance gaps by themselves demonstrate a constitutional violation, we reject this argument. The financial efficiency doctrine requires a rough equality of access to district funding for similar tax effort. Its aim is equality of opportunity, not equality of results. We have never interpreted our Constitution, under the adequacy requirement, to mandate equality of student achievement by district or student subgroup. Such equality of results may not be possible through changes in school funding alone, given the *863respected body of educational research holding that school resources account for only a small fraction of differences in student achievement. Equality of educational achievement is a worthy goal of government, and society at large, but it is not a constitutional requirement.
For all of these reasons, we hold the Plaintiffs failed to prove that the system was so parsimonious in its allocation of resources to ELL and economically disadvantaged students as to amount to an unconstitutionally inadequate allocation. We also hold the Plaintiffs failed to prove that current funding weights are so flawed as to render the school system as a whole constitutionally inadequate.
5. Test Scores and Other Outputs
Because the adequacy standard “is plainly result-oriented,”186 the proper focus of a constitutional adequacy analysis should be on outputs that measure student performance. The trial court considered these outputs, concluding in its judgment that “[a]ll performance measures considered at trial, including STAAR tests, EOC exams, SATs, the ACTs, performance gaps, graduation rates, and dropout rates among others, demonstrated that Texas public schools are not accomplishing a general diffusion of knowledge due to inadequate funding.”187 We disagree with this legal conclusion.
We have already discussed why we disagree with the trial court’s analysis of funding. That problem aside, we examine whether the record shows that test scores and other outputs — the proper focus of the adequacy determination — demonstrate a constitutionally inadequate system. They do not.
At the outset, we must consider whether the system, as the Legislature has designed it, is adequate to meet the constitutional requirement to provide a general diffusion of knowledge. As explained above, the Legislature need only act “reasonably” in making these policy decisions. “[N]o one would dispute that a public education system limited to teaching first-grade reading would be inadequate.”188 But “few would insist that merely to be adequate, public education must teach all students multiple languages or nuclear biophysics.” 189 Somewhere between such extremes, the Legislature must determine and define the “required curriculum” that, in its view, the education system should provide. Once it has done so, our review is “very deferential,” and we must uphold the Legislature’s determination unless it is arbitrary and unreasonable.190
We have explained in some detail the system the Legislature has chosen, and we have no difficulty concluding that it does not represent an arbitrary and unreasonable effort to provide a general diffusion of knowledge. District and campus performance is reviewed based on student achievement measures including test scores, dropout rates, and graduation rates.191 Continued accreditation of districts and charter schools also depends on these measures of outputs.192
In Edgewood TV, we held that this accountability regime, by itself, satisfied “the Legislature’s constitutional obligation to *864provide for a general diffusion of knowledge.” 193 In WOC /,we again accepted that, under the Legislature’s chosen system, the provision of an accredited education satisfies the Legislature’s obligation to provide a general diffusion of knowledge.194 We presumed that the system as designed accomplishes a general diffusion of knowledge, showing “deference to the Legislature’s choices” in this area of shared constitutional authority.195 We acknowledged in Edgewood IV and later in WOC I that the Legislature’s accountability regime, as designed, may not always be sufficient to meet the constitutional requirement of providing a general diffusion of knowledge.196 Here, however, we cannot conclude that the legislatively designed system provides and requires so little as to indicate that the Legislature has arbitrarily abandoned its duty to provide a general diffusion of knqwledge.
We next consider whether the system is sufficiently operating as designed. In today’s case, the State showed that in 2013, 92.8% of school districts and charter schools and 84.2% of individual school campuses achieved the “Met Standard” rating. In 2014, 90.2% of school districts and char-, ter schools and 85.0% of individual school campuses achieved this rating.197 In 2015 the comparable figures were 94.5% for districts and charter schools and 86.5% for individual campuses.198 In WOC II, the percentage of students meeting the college-readiness standards of the Texas Higher Education Coordinating Board stood at 28% in English and 42% in math.199 Those figures stood at 65% and 66% in 2013. We agree with CCISD that, in this case, “accreditation alone simply cannot be presumed to show satisfaction of the constitutional requirements,” but we think it should be considered some support in favor of an adequacy determination.
The trial court rejected the presumption of adequacy that follows compliance with the accountability regime. It reasoned that “the accountability standards are set not to measure whether districts are achieving a general diffusion of knowledge, but rather to ensure that most districts and campuses fall on the ‘academically acceptable’ or ‘met standards’ side of the line.” 200 We question this reasoning. There is nothing unusual in calibrating a grading system to allow most participants to pass. Further, in WOC II, the trial court had made an essentially identical finding that “the requirements for an ‘academically acceptable’ rating are set to assure, not that there will be a general diffusion of knowledge, but that almost every district will meet them.”201 Yet we found the system adequate.
The trial court in today’s case also criticized what it saw as the State’s “history of *865slowly phasing in standards,” noting the phase-in of the TAKS-based accountability . system.202 In WOC II, we noted the phase-in of the TAKS regime,203 and saw nothing to criticize from a constitutional perspective. Instead, in discussing the statutory mandates found in sections 4.001(a) and 28.001 of the Education Code (quoted at the beginning of this opinion), we held that “they cannot be used to fault a public education system that is working to meet their stated goals because it has not yet succeeded in doing so.” 204 And in discussing passing scores for the new TAKS test, we saw nothing remarkable in the State’s decision that the cut scores “should be lower at first, increasing over three years, to give teachers and students time to adjust to the new and more difficult test.” 205 A phase-in period is also warranted because, as the trial court recognized, the college-readiness standards were “vertically aligned” into “the entire curriculum-from kindergarten all the way to high school.” 206 Students who had already started school after the introduction of these new standards did not have the benefit of schooling under the new vertically aligned curriculum in prior grades.
The trial court also considered performance on the STAAR tests. It found that the initial student performance was “sobering.” 207 Ninth graders took the initial round of tests in 2012. Under an initial Level II phase-in standard set by TEA, 32% of students failed the English I Reading exam. The failure rates were 45% for English I Writing, 17% for Algebra I, 13% for Biology, and 19% for World Geography.208 The trial court also noted significant gaps between economically advantaged and disadvantaged students.209
In condemning the system as constitutionally inadequate, the triál court considered the early administrations of the test. Students, teachers, and schools were undoubtedly challenged by new, harder tests they had not seen. The same was true in WOC II, where we noted, in discussing district performance ratings: “After the change to the harder test, ratings predictably slid.”210 Again, the system should not necessarily be written off because it “is working to meet” its new standards but “has not yet succeeded in doing so.”211 Over the 2004 to 2011 period, scores on the TAKS tests improved on all subjects, and for all students and various student subgroups (white, African-American, Hispanic, and economically disadvantaged). One can speculate that the early poor performance on the STAAR tests was due to funding cuts. But surely some of the difficulties were simply due to adjustments required to teach to and take the harder and unfamiliar tests. A TEA testing director so testified, explaining that performance standards are phased in to provide a meaningful period of adjustment.212 The *866system is not unconstitutional because the State has decided to demand more of its schools and teachers, and they predictably faced a transitional period of adjusting to the more difficult tests.
More recent STAAR results show improvements by some measures. The trial court noted that the cumulative failure rate on all tests had improved slightly from 2012 to 2013.213 The State offered evidence that the combined pass rates for. spring, summer, and fall of 2012 showed improvement as compared to the first test in the spring of 2012. For Algebra I, the pass rate went from 76.8% to 84.7%; for English I Reading, from 67.7% to 81.2%; for English I Writing, from 54.4% to 72.6%; for Biology, from 86.4% to 91.0%; and for World Geography, from 79.7% to 84.4%. For 2013, the results again improved, with pass rates for the class of 2015 ranging from 78% on English II (combining English reading and writing exams) to 96% in Biology. However, gaps among certain subgroups remain, some more pronounced than others. For example, the 78% pass rate on English II came in at 68.5% for economically disadvantaged and 27.1% per ELL students; the 96% pass rate on Biology came in at 93.8% for economically disadvantaged students and 79% for ELL students. Poor performance by ELL students on an advanced English exam is troubling, but it is not by itself a basis for holding the “system as a whole ”214 constitutionally inadequate.
The Class of 2015 was the first to graduate under the STAAR requirements. After Spring 2015 testing, 92% of students had passed all five end-of-course (EOC) exams required for graduation,215 though some students had to take tests more than once. Broken down by individual test and certain student subgroups, the numbers also paint a less grim picture than the scores on which the trial court focused.216 The 2015 EOC pass rates are as follows:
*867[[Image here]]
The NAEP tests, considered by the Court in WOC II,217 show mixed results. For 2013, the trial court found that NAEP scores dropped on two exams and went up on two. From 2005 to 2011, the results are described as flat except for the eighth grade math score.218 On that test, the Texas average went from 281 to 290, compared to the national average which went from 278 to 283. In 2011, Texas ranked 29th in overall scores (reading and math for grades 4 and 8). The State points out that when the numbers are broken out by subgroup, Texas in 2011 ranked fourth in the nation for African-American students, seventh for Hispanic students, and eleventh for economically disadvantaged students.
Graduation rates are another output the State uses in performance and accreditation reviews.219 According to evidence in the record, Texas had a graduation rate of 86% in 2011, tying for fourth in the nation. According' to the U.S. .Department of Education, Texas had a graduation rate of 88% in 2013, tying for second in the nation.220 For that year, the graduation rate for African-American students was 84.1%, first in the nation.221 The rate for Hispanic students was 85.1%, first in the nation. The rate fot economically disadvantaged students was 85.2%, second in the nation.222 These performances stand in contrast to the “severe dropout problem” we described in WOC II.223
On this issue and others, the parties single out, parse, emphasize, and minimize many other statistics in arguing their respective positions. We have noted a few of the differences in statistical measures found in the record, such as those relating to measuring graduation rates and revenue per student.224 The current system leaves much to be desired. But again, using WOC II as a guide, we noted many difficulties in the system at the time:
*868In the extensive record before us, there is much evidence, which the district court credited, that many schools and districts are struggling to teach an increasingly demanding curriculum to a population with a growing number of disadvantaged students, yet without ad- • ditional funding needed to meet these challenges. There are wide gaps in performance among student groups differentiated by^ race, proficiency in English, and economic advantage. Non-completion and dropout rates are high, and the loss of students who are struggling may make performance measures applied to those who continue appear better than they should. The rate of students meeting college preparedness standards is very low. There is also evidence of high attrition and turnover among teachers statewide, due to increasing demands and stagnant compensation.
Despite those difficulties, we also noted improvements in some test scores and held that the system was constitutionally adequate.225 Likewise, today, we conclude that the performance of the system as measured by the outputs we describe does not establish a violation of the adequacy requirement. While Texans may desire a public education system that produces even “better” results or better results more quickly, their remedy lies in the Legislature and thus in the privilege and duty that all Texans have to elect the legislators who will implement the policy choices they desire. The Constitution, meanwhile, requires only a system that produces a “general diffusion of knowledge,” and we cannot say that the Legislature has acted arbitrarily or unreasonably in its efforts to produce that result.
6. Conclusion Regarding Adequacy
Our decision on the adequacy requirement is largely driven by our standard of review, the legal lens through which we examine this issue. In WOC II, we said that “it remains to be seen whether the system’s predicted drift toward constitutional inadequacy will be avoided by legislative reaction to widespread calls for changes.” 226 It is safe to say that the current Texas school system leaves much to be desired. Few would argue that the State cannot do better. But our function is limited to reviewing the constitutionality of the system under an extremely deferential standard, one that places the burden of proving the system constitutionally inadequate on the party challenging it. The Plaintiffs did not meet that burden.
F. Suitability
The trial court held the current school system is constitutionally unsuitable. We have recognized a suitability requirement deriving from the text of article VII, section 1, imposing a legislative duty “to establish and make suitable provision for” an efficient public school system.
The State essentially argues the system is suitable because it is efficiently providing a general diffusion of knowledge under the adequacy and financial efficiency requirements. The Plaintiffs argue that adequacy, efficiency, and suitability are separate elements. We have made clear that the suitability requirement “is not merely redundant of’ the adequacy and efficiency requirements.227 Suitability is therefore a separate requirement under article VII, section 1.
But the elements are certainly related, as it would be the odd case where the *869Legislature is complying with its duty to efficiently provide a constitutionally adequate education but the system is nevertheless somehow unsuitable. In WOC II, in two paragraphs, we held the suitability requirement was met, noting that “[n]either the structure nor the operation of the funding system prevents it from efficiently accomplishing a general diffusion of knowledge.” 228
Our prior decisions offer some insights on suitability. We have stated generally that “if the Legislature substantially defaulted on its responsibility such that Texas school children were denied access to that education needed to participate fully in the social, economic, and educational opportunities available in Texas, the ‘suitable provision’ clause would be violated.” 229 We have also stated that the suitability requirement “refers specifically to the means chosen to achieve an adequate education through an efficient system.”230 The provision “requires that the public school system be structured, operated, and funded so that it can accomplish its purpose for all Texas children.”231 We have given as an example that the system would be unsuitable if it were providing an efficient and adequate education but there was no legal mechanism in place that “actually required” school districts “to provide an adequate education.” 232 So if the Legislature established goals for an adequate education of students and employed school districts to provide a general diffusion of knowledge, but those districts were free to ignore legislative goals, the system would be unsuitable. 233 However, reliance on school districts to carry out legislative goals for a general diffusion of knowledge and heavy reliance on local tax revenues to fund the system do not, by themselves, render the system unsuitable.234
There is a somewhat' fine distinction in our caselaw on suitability. On the one hand, we have said this requirement “refers specifically to the means chosen to achieve an adequate education.” 235 On the other hand, we have stated generally that we leave it to the Legislature “to decide how to meet the standards” of article VII, section 1, while, we have final authority on “whether they have been met.”236
This Court has never held the school system constitutionally unsuitable. Such a defect appears to be reserved for some fundamental and insurmountable structural flaw, especially where the system is succeeding in efficiently providing an adequate education for Texas students.237 For example, we held the system' met the suitability requirement in WOC //because its structure and operation were not so flawed as to make it “impossible” to efficiently provide a general diffusion of knowledge.238 The example given above— the lack of a legal mechanism to ensure an adequate education — would not apply because numerous statutory sanctions and interventions are available to remedy a *870district’s or school’s failure to meet legislatively mandated accreditation standards. School districts are required to provide instruction in State-designated knowledge and skills to receive accreditation.239 The State can withhold funding from districts and charter schools failing to meet these standards, appoint a monitor or conservator, close the district or annex it to another district, or pursue other statutory remedies.240 The fundamental structure found suitable in WOCII remains in place today.
The trial court held the school system was unsuitable because it was underfunded,241 essentially tying the suitability analysis to its flawed conclusion that the system is constitutionally inadequate because it is underfunded, a conclusion we reject above. The Plaintiffs failed to prove that the school system is unsuitable.
G. Financial Efficiency
The trial court held the current school finance system violated the “financial efficiency” requirement of article VII, section 1. On this issue, the ISD Plaintiffs part company. CCISD sides with the State in arguing that the system is financially efficient.242
Our basic framework for deciding this issue has not changed since Edgewood I, where we held that “districts must have substantially equal access to similar revenues per pupil at similar levels of tax effort.” 243 And in Edgewood II, we said, “To be efficient, a funding system that is so dependent on local ad valorem property taxes must draw revenue from all the property at a substantially similar rate.”244 Exact equality of funding among districts has never been required. 245
The State’s duty to provide equal access to funding applies only to the amounts necessary to provide for a general diffusion of knowledge.246 Once the system provides for a general diffusion of knowledge, the Legislature “may, so long as efficiency is maintained, authorize local school districts to supplement their educational resources if local property owners approve an additional local property tax.”247
Since the Edgewood II decision in 1991, we have not found a violation of the financial efficiency requirement. Since then, the system has used recapture among other mechanisms to equalize funding.
In analyzing this issue, our prior decisions considered various metrics. In making comparisons to prior decisions, we particularly focus today on ratios, because *871financial efficiency turns not on absolute tax rates or levels of funding, but the relative differences between wealthier and poorer districts that can be presented mathematically with ratios.
In Edgewood I, we found a financial efficiency violation, noting that property wealth per student varied from $14 million in the richest district to $20,000 in the poorest, a ratio of 700; the 300,000 students in the wealthiest districts had over 25% of the State’s property wealth, while the 300,000 in the poorest districts had 3% of the property wealth, a ratio of about 8; the average property wealth in the 100 wealthiest districts was more than 20 times the average property wealth in the 100 poorest districts. Spending per student ranged from $19,333 to $2,112, a ratio of about 9; an average of $2,000 more per year was spent on each of the 150,000 students in the wealthiest districts than was spent on the 150,000 students in the poorest districts. Local tax rates varied from $1.55 to 9 cents, a ratio of about 17. The 100 poorest districts had an average tax rate of 74.5 cents, while the 100 richest had an average tax rate of 47 cents, a ratio of about 1.6. The 100 richest districts spent an average of $7,233 per student, while the 100 poorest spent $2,978, a ratio of about 2.4.248
By contrast, in Edgewood IV, we held the financial efficiency requirement was satisfied.249 We noted that the tax rates required to achieve a general diffusion of knowledge were $1.31 for the poorest 15% of districts and $1.22 for the richest 15%, a ratio of about 1.07.250 We noted that under the recapture and guaranteed funding then available, wealthier districts could raise up to $28 per weighted student, and all districts were guaranteed $20.55 per weighted student, for every cent raised beyond the 86 cents required for Tier I funding, yielding a ratio of 1.36.251 We also noted that once SB 7 was fully implemented, the richest 15% of districts would have a yield of $28.74 per penny of tax effort, while the poorest 15% of districts would have $26.74,252 for a ratio of 1.07.
Similarly, in WOC II, we concluded that no violation of the financial efficiency requirement was presented.253 Looking at property wealth per student (measured in WADA), the ratio of the richest district over the poorest district had fallen to 200, compared to 700 in Edgewood I.254 We compared “chapter 41” districts — property-rich districts subject to recapture — to “chapter 42 districts” that receive the recaptured funds. FSP revenue per student (measured by WADA) was $5,457 for chapter 41 districts and $4,330 for chapter 42 districts, a ratio of 1.26. We noted that the comparable figures at the time of Edgewood IV were $3,510 for chapter 41 districts and $3,005 for chapter 42 districts,255 a ratio of about 1.17. In WOC II, looking at districts in the top 5% and bottom 5% measured by property wealth, we observed that the revenue per student in WADA was $5,895 for the richest districts and $4,217 for the poorest,256 a ratio of about 1.4.
*872In today’s ease,, giving the Financial Efficiency Plaintiffs the benefit of the doubt and assuming that all spending goes to providing a general diffusion of knowledge and must therefore be considered in the financial efficiency determination,257 the current ratios are in the range of similar ratios in' Edgewood TV and WOC II that did not present a constitutional violation. The ratios are far below those in Edge-wood I where a constitutional violation was presented.
In FOFs 1265 and 1267, the trial court found, based on data from Edgewood ISD expert Dr. Albert Cortez, that in 2011-12, measured by property wealth, the wealthiest decile of districts had average M & 0 revenue per student in WADA of $7,097, while the poorest decile had average M & 0 revenue per student in WADA of $5,654, a ratio of about 1.26. In 2012-13, the comparable figures were $6,715 and $5,617, for a ratio of about 1.19. In 2013-14, according to a different expert, Dr. Catherine Clark, comparable figures were $6,708 and $5,801, for a ratio of 1.16.258 Significantly, these ratios are declining, indicating a trend toward more equal funding. Clark also calculated M & 0 revenue per student in WADA for 2013-14, assuming that all districts taxed at the same rates of $1.00, $1.04, and the maximum rate allowed (usually $1.17).259 The ratios come in at 1.11,1.11, and 1.12.
A State expert, Dr. Lisa Dawn-Fisher, offered data that for 2013-14, the wealthiest 5% of districts had a weighted average of $7,673 in M & O dollars per student in WADA while the poorest 5% had $5,797, for a ratio of 1.32. TTSFC refers us to calculations from its expert Dr. Wayne Pierce, showing a slightly different ratio.260 *873It shows, for 2013-14, average M & 0 revenue per student in WADA for the top and bottom 5% of districts at $8,146 and $5,737, for a ratio of 1.42.
All of these ratios are below or near the ratio of 1.4 for funding of the wealthiest versus the poorest 5% of districts noted in WOC II. The WOC II ratio existed in a system found constitutionally efficient. All of the ratios in today’s case are well below the spending ratios in Edgeivood I, where we held the system unconstitutional, such as the 2.4 ratio between the richest and poorest 100 districts, or the ratio of 9 between the highest and lowest-spending district.
Data from Dawn-Fisher shows that for 2011-12, total FSP revenue per student in WADA was $7,318 for chapter 41 districts and $6,350 for chapter 42 districts, for a ratio of 1.15. The comparable ratio was 1.11 for 2012-13, 1.08 for 2013-14, and 1.07 for 2014-15. These ratios are below the ratio of 1.26 for chapter 41 versus chapter 42 districts in WOC II, and the ratio of 1.17 at the time of Edgeivood IV. And again, this data shows a trend toward greater equality of funding. Plaintiffs’ expert Pierce also testified that the revenue gaps between wealthy and poor districts were decreased by the actions of the 83rd Legislature. And as the trial court found, the Legislative Budget Board has projected a decrease in revenue gaps as well, though the court thought the reductions were insignificant.261
WOC //noted that “although 95% of public school funds are equalized through the FSP, 5% are not.” 262 The percentage of unequalized FSP M & O revenue (such as revenues from golden pennies) has decreased from 8.2% in 2006 to 3.5% in 2013.
Looking again to FOFs 1265 and 1267, the trial court found that the average M & Ó tax rate for the poorest decile-was $1.11 in 2011-12 while the average rate for the wealthiest districts was $1.00, for a ratio of 1.11. The average tax rate for the poorest decile was $1.11 in 2012-13 while the average rate for the wealthiest districts was $1.01, for a ratio of about 1.1. Dawn-Fisher offered data that in 2012-13 the weighted average M <& O tax rates for the bottom and top 15% of districts were $1.10 and $1.04, for a ratio of 1.06. Pierce’s numbers came in at $1.097 and $1.026, for a ratio of 1.07. Pierce also offered combined M & O and I & S tax rates, finding a rate of $1.256 for the poorest 15% and $1.181 for the richest, for a ratio of 1.06. In Edgewood IV, a ratio of 1.07 applied to the tax rates of the richest and poorest 15% of districts, and we upheld that system. The numbers in today’s case are certainly in the same range for constitutional purposes. All of these ratios are well below the ratio of 1.6 between the richest and poorest districts in Edgewood I, where we found a constitutional violation.
The trial court found that property values per student in WADA ranged from $22,218 in Boles ISD to $7,341,341 in Ken-*874edy ISD,263 for a ratio of about 330. This compares to ratios of 200 in WOC II, where we found no constitutional violation, and 700 in Edgewood I, where we found a violation. Suffice it to say that the ratio in today’s case remains huge. But this metric is of limited value today because of recapture, a process intended to address this particular disparity. Recapture did not exist at the time of Edgewood I, and other metrics, such as spending per weighted student, incorporate the effect of recapture.
In considering ratios, we can finally look to measures of “yield,” that is, measures of how much revenue a district can raise for each penny of tax effort. Such a measure numerically corresponds to our standard for financial efficiency in Edgewood I, requiring that “districts must have substantially equal access to similar revenues per pupil at similar levels of tax effort.” 264 In Edgewood I, we described the tax rates and revenues per student of the richest and poorest 100 districts, indicating yields of 153.89 and 39.97. In other words, the richest districts could raise $153.89 for each penny of tax effort, while the poorest could raise $39.97 for each penny,265 resulting in a ratio of 3.85, meaning that the richest districts could raise 3.85 times the amount the poorest could raise for each penny of tax effort. In Edgewood TV, we recited figures relating to two yield ratios. We noted a ratio of 1.36 as the ratio of what the wealthiest districts could raise ($28.00 per weighted student per penny of tax effort) over the yield guaranteed by SB 7 ($20.55).266 We also noted that after full implementation of SB 7 the richest 15% of districts would have an average yield per penny of tax effort of $28.74, while the poorest 15% would have an average yield of 26.74, for a ratio of 1.07.267
The parties offered yield evidence in today’s case, including calculations by Dawn-Fisher, a State expert, and Pierce, an expert for the Financial Efficiency Plaintiffs. Depending on the methods used and the precise figures being measured,268 the yields vary considerably, but overall are much closer to the constitutionally acceptable yields described in Edge-wood TV than the constitutionally unacceptable yields in Edgewood I.
Using 2013 data and 2014 legislative parameters, and considering actual tax rates, Dawn-Fisher’s assessed the richest and poorest 10% of districts, with each group consisting of 102 districts, inviting a comparison to Edgewood I ⅛ richest and poorest 100 districts. She calculated weighted average M & O yields of 66.30 for the richest 10% and 52.41 for the poorest, for a ratio of 1.265. Looking to the richest and poorest 15% of districts yielded a ratio of 1.17. These ratios fall within the ratios of 1.36 and 1.07 described in Edgewood TV, which considered the richest and poorest 15% of districts, and are far below the ratio of 3.85 in Edgewood I, which considered the richest and poorest 100 districts. Similar though not identical ratios can be derived from Pierce’s calculations, which also considered 2013 actual tax rates and *8752014 legislative parameters. Looking at the districts comprising the top and bottom 15% of students, as was done in Edge-wood, IV, produced average M & 0 yields of 72.85 and 51.94, for a ratio of 1.40. Pierce’s data showed a ratio of 1.56 for the top and bottom 15% when Pierce combined M & O and I & S average revenues. These ratios are higher than the ratios described in Edgewood IV (1.36 and 1.07) but still far below the 3.85 ratio in Edge-wood I.
The trial court made hundreds of fact findings on this issue, but many were inextricably and in our view erroneously tied to its findings that the State had to spend at a higher level — the adequacy estimates of experts Odden, Moak, and Baker — to achieve adequacy.269 For example, in the course of its financial efficiency analysis, the court looked not to actual spending levels but instead repeatedly noted that some districts were spending below the Odden or Baker numbers. The court time and again did not compare actual tax rates, but analyzed the tax rates that would be necessary to yield the revenues matching the adequacy estimates of Odden, Moak, or Baker. Much of the trial court’s thinking was that the system was inefficient because many districts, even taxing at the maximum legal rate, could not raise sufficient funds to satisfy the trial court’s chosen adequacy figures. For the reasons previously discussed, we disapprove of the court’s adequacy analysis and in particular its reliance on the adequacy estimates of these experts to arrive at minimum spending levels allegedly needed to achieve a general diffusion of knowledge. This error spills over into the financial efficiency analysis.
The trial court also made findings that property-poor districts levy higher I & S taxes but raise less revenue for facilities.270 The system for facilities funding is essentially the same as the one that existed when we decided WOC II. There we held:
The State defendants argue that disparities among districts in available facilities are not proof of inefficiency absent evidence that the districts’ needs are similar. They contend that facilities needs vary widely depending on the size and location of schools, construction expenses, and other variables. We agree that such evidence is necessary and lacking. The State defendants also argue that to prove constitutional inefficiency the intervenors must offer evidence oí-an inability to provide for a general diffusion of knowledge without additional facilities, and that they have failed to do so. Again, we agree. .Efficiency requires only substantially equal access to revenue for facilities necessary for an adequate system271
We think the proof in today’s case was similarly lacking with regard to 'the Financial Efficiency Plaintiffs’ arguments, that unequal I & S funding renders the system financially inefficient. Further, as discussed above, Dr. Dawn-Fisher calculated total FSP funding gapsr between chapter 41 and chapter 42 districts. Those comparisons included M & O and I & S spending. Her data showed ratios below those found in WOC II, as well as a trend toward more equal funding.
We recognize that the comparisons to past ratios are not matters of mathematical perfection, and that even in the same case the ratios can vary considerably from expert to expert. But using our precedents as a guide, the ratios we describe in *876today’s case indicate a system that is financially efficient for constitutional purposes. There is no single magic number or ratio that determines financial efficiency. However, as our discussion above shows, the ratios in today’s case are in the range of those from prior cases where we found the system constitutionally efficient, and well below ratios where we found the system constitutionally inefficient. We accordingly conclude that the system does not violate the financial efficiency requirement of article VII, section 1.
H. Intervenors’ Qualitative Efficiency Claims
The Intervenors argue the school system is qualitatively inefficient because it does not provide for a general diffusion of knowledge with little waste. They contend the system is structurally unsound, wasteful, and unproductive of results. They complain of constant unproductive litigation. Alleged structural inefficiencies include establishing school districts as near monopolies, a cap on the number of charter schools, and a failure of the system to determine the cost of educating a child. The Intervenors further complain that the system is too top-down and mandate-driven, teacher tenure and compensation rules are inefficient, and class-size limits add to inefficiency. Conflicting evidence was presented as to whether these rules contribute to student achievement. The Interve-nors ask the Court to “declare the system unconstitutional because its structural inefficiencies impair .its ability to produce educational results with little waste.”
The trial court did not err in rejecting the Intervenors’ claims. Article VII, section 1, of course, expressly imposes on the Legislature a duty to provide “an efficient system of public free schools.” In Edge-wood I, the Court recognized, in addition to the equal funding requirements of financial efficiency, a more general qualitative efficiency requirement in article VII, section 1, referring to a system “effective or productive of results and connot[ing] the use of resources so as to produce results with little waste.” 272 In Edgewood IV, we held the qualitative efficiency requirement was met because the Legislature met its “constitutional obligation to provide a general diffusion of knowledge statewide.”273 In WOC II, we again recognized the element of qualitative efficiency, quoting Edgewood I ⅛ reference to a system effective or productive of results and producing results with little waste.274 We then discussed the adequacy requirement as the requirement that public education achieve a general diffusion of knowledge,275 and later emphasized that “the accomplishment of ‘a general diffusion of knowledge’ is the standard by which the adequacy of the public education system is to be judged.” 276 In short, we have analyzed both qualitative efficiency and adequacy by asking whether the system was achieving a general diffusion of knowledge. Qualitative efficiency and adequacy are certainly related.
If “qualitative efficiency” and adequacy can mean different things, there is surely a great deal of overlap. We have expressly recognized “[t]he interrelated constitutional standards of efficiency and adequacy.” 277 A system that, as here, is adequate to achieve a general diffusion of *877knowledge may not be a model of efficiency, and probably is not. But as with adequacy, suitability, and the “general diffusion of knowledge,” the Constitution grants the Legislature broad discretion in determining the acceptable level of qualitative efficiency, and we cannot interfere unless it acts arbitrarily and unreasonably in making that choice. We have difficulty imagining conditions so structurally unsound as to violate a separate qualitative efficiency standard where the system is constitutional under an adequacy review. Since both standards, as well as the suitability standard, are all concerned with whether the State is providing a general diffusion of knowledge, the case would be rare where a plaintiff fails to prove constitutional inadequacy or unsuitability, but nevertheless succeeds in proving that the Legislature has acted arbitrarily and unreasonably with regard to qualitative inefficiency. We do not rule out completely such a showing, nor does the State,278 but we are convinced the Intervenors did not make one here.
The Court has never found a constitutional violation due to qualitative inefficiency. We think a plaintiff faces a steep challenge in establishing that a system, once found to be constitutionally adequate, is nevertheless constitutionally deficient under a separate qualitative efficiency requirement. The Court moreover has made clear that it focuses on results and does not micromanage the programs and methods used to achieve those results: “the Legislature has the sole right to decide how to meet the standards set by the people in article VII, section l.” 279 In deciding whether constitutional standards are met, we do not “judge the wisdom of the policy choices of the Legislature.”280
The Intervenors make intriguing suggestions for improving the efficiency of the school system. We hope the Legislature will consider these and similar suggestions. But the Intervenors did not cross the threshold of proving that such changes are constitutionally mandated. The Interve-nors’ claims are reminiscent of claims in Edgewood TV, where some plaintiffs alleged the school finance system was unsuitable and inefficient in denying student-centered funding that empowers parents to direct some of their tax dollars to schools of their choice. The Court rejected these arguments because the , Court leaves to the Legislature “the primary responsibility to decide how best to achieve an efficient system.” 281 The Intervenors essentially urge that the Texas Constitution mandates for K-12 students something akin to the Tuition Equalization Grant Program, which since 1973 has provided grant aid to financially needy students who attend private, non-profit colleges and universities in Texas.282 The TEG Program has earned bipartisan kudos for opening educational doors for a generation of low-income Texans. In the Texas higher education setting, school choice has proven to be smart policy. But the Intervenors have not proven that a TEG-like program for K-12 students is constitutionally mandated.
*878It is not for the Court “to judge the wisdom of the policy choices of the Legislature, or to impose a different policy of our choosing.” 283 Rules governing class size, for example, are policy decisions that we should leave to the Legislature. There is no scientific consensus on this issue. One expert noted “a small number of good studies that produce variable results.” As for rules governing teacher certification, we have stated that courts should not become “the arbiter of education and policy, overseeing such issues as curriculum and testing development, textbook approval, and teacher certification.”284 ■
The Intervenors argue that a structural failing of the system is its “monopolized statutory scheme” and that school districts are “near monopolies.” They specifically complain about a cap on the number of charter schools and the absence of competition. This harkens to the lone dissent in WOC II, stating that efficiency is produced “not by protectionism, not by higher taxes, and not by state control, but by freedom for competition.” 285' ’ But the Court’s prevailing view, to which we adhere today, is to recognize the Legislature’s “broad discretion to make the myriad policy decisions concerning education.” 286 The Court has never held the system unconstitutional because it is too monopolistic or because of a lack of competition. There is of course some competition from inside and outside the current system, offered by charter and private schools, and home schooling. The Intervenors did not quantify the inefficiencies about which they complain, in dollar terms or through some other quantitative measure, or the efficiencies that would be achieved with the additional competition they advocate. Some inefficiency, even a considerable amount of it, must be tolerated when governments, rarely models of efficiency, are charged with Augean tasks.
The Intervenors specifically complain that the cap on the number of open-enrollment charter schools stifles competition.287 At the time of the trial court’s judgment, the cap on the number of open-enrollment charters was 215. Currently 240 open-enrollment charters are authorized and the number is being increased to 305 over the next few years. Evidence at trial showed charter school enrollment growing at 15% per year.288 So the Legislature is increasing competition, and the pace of allowing more charter schools is well within the policymaking prerogatives of the Legislature. And as discussed further below with respect to the Charter School claims, creating charter schools imposes unique administrative challenges. We cannot say the cap is so arbitrary as to render the entire system constitutionally inefficient.
Making an argument similar to one made by the ISDs, the Intervenors contend the system is unconstitutional because the State has failed to determine the cost of educating a child. Such determinations might be the better practice, but again this argument misplaces the burden of proof. Constitutionality is presumed; it is the Intervenors’ burden to show funding is constitutionally inadequate. And as not-, ed above, calculating such costs may be a largely ineffectual exercise given that the social sciences may not have advanced to the point of allowing for such calculations in a meaningful way. For these and the *879further reasons explained above, the system is not unconstitutional because the State has failed to comply fully with section 42.007 of the Education Code or otherwise calculate the costs of educating a child.
We accordingly affirm the trial court’s rejection of the Intervenors’ claims, except we remand their claim for attorney fees along with all other claims for such fees, as discussed below.
I. Charter School Plaintiffs’ Claims
The Charter School Plaintiffs complain that charter schools receive at least $1,000 less per weighted student than other schools largely because they receive no funding for facilities. They also complain that they are not subject to certain adjustments in the FSP formula available to other schools, and therefore do not receive extra funding they deserve.
The Education Code recognizes three classes of charter schools: home-rule school district charter schools, campus or campus program charter schools, and open-enrollment charter schools.289 The trial court and the Charter School Plaintiffs focused on open-enrollment charter schools funded directly by the State.
The trial court held that because the ISD Plaintiffs established that the level of school funding was constitutionally inadequate and because charter school funding is based on state averages of ISD funding, the Charter School Plaintiffs prevailed on their claim that funding for open-enrollment charter schools was constitutionally inadequate.290 The Charter School Plaintiffs also argue that because funding formulae are unsuitable as to the ISD Plaintiffs, those same formulae render the system constitutionally unsuitable as to charter schools. But as we hold above, the evidence of inadequate funding is not a basis for declaring the school system constitutionally inadequate or unsuitable. For the same reasons we have concluded that system-wide adequacy and suitability claims fail for the ISD Plaintiffs, these claims also fail with respect to the Charter School Plaintiffs.
The Charter Schools Plaintiffs further complain that the differential in funding between charter schools and school districts renders the school finance system unconstitutionally inadequate, unsuitable, and inefficient. Charter schools receive FSP funding but (1) they do not receive facilities funding from the State like other schools, and (2) they are subject to average funding formulae rather than receiving adjustments for factors that increase costs.
Charter schools sometimes receive less than other schools because of the way facilities are funded. School districts can impose local taxes to pay for facilities bonds under section '45.001(a) of the Education Code. Charter schools, lacking taxing authority, cannot utilize this provision. The State also provides direct facilities funding to districts to cover faeilities-relat-ed debt, but charter schools are not 61⅛-⅜ ble for these funds.291
Funding inequalities are addressed in our caselaw under the financial *880efficiency doctrine, which considers whether equality of funding is provided by similar tax efforts. The claim here is hard to fit into our caselaw on financial efficiency, the element that focuses on disparities in funding. Beginning with Edgewood, I, financial efficiency requires a “direct and close correlation between a district’s tax effort and the educational resources available to it; in other words, districts must have substantially equal access to similar revenues per pupil at similar levels of tax effort.” 292 But charter schools, having no tax base, do not have to raise taxes at all. In a sense, they'receive a better bargain because school district funding from the State comes largely in the form of matching funds under the two tiers of the FSP. But it is all essentially free taxpayer money for charter schools, requiring no matching effort on their part.
Conceivably a difference in the way similarly situated charter schools and other schools are funded might render the system constitutionally inefficient, inadequate, or unsuitable. The standard is whether this difference in treatment is arbitrary.293
The trial court found substantial differences between the way charter schools and other schools are operated that might affect funding needs.294 Charter schools are by legislative purpose an avenue to “encourage different and innovative learning methods.” 295 The Charter School Plaintiffs themselves point out that “charter schools find a way to do more with less,” and “have achieved remarkable success with their limited resources.” That is a strength of their innovative approaches. Compared to school districts, charter schools have much more flexibility in personnel. They can hire at-will employees, there is no minimum salary scale, and teachers need only a high school diploma and need not be certified.296 Some disciplinary and placement procedures of Chapter 37 of the Education Code do not apply. Class size limitations and student-teacher ratios do not apply.297 And the State points out that charter schools do receive some funding for facilities, under the “new infrastructure facility allotment” (NIFA) program.298 Ultimately, we see something of an apples-and-oranges comparison. Charter school operators must find a facility for the school and do not get taxpayer support for the facility, but once they have a facility they get free taxpayer funding from the State without having to match. We agree with the trial court that the differences in facilities funding are not so arbitrary as to rise to a constitutional violation.
As for funding adjustments, the State uses statewide averages, rather than making individual determinations for each charter school, for certain statutory adjustments — those relating to the cost of education, district size, district population sparsity, and enrichment funding.299 The Charter School Plaintiffs are asking for more fine tuning — school by school — than *881districts receive. But the administrative challenges of such fine tuning are not subject to a constitutional mandate under our “very deferential” review of claims under article VII, section l.300 The Commissioner, of Education testified that “when you create a charter, it’s like creating a whole new school district” that “adds to the level of workload to the agency.”301 He testified that without a cap, an influx of a large number of charter schools could strain the Education Agency’s resources. Further, the State correctly points out that some important funding adjustments are made for charter schools, including adjustments for ELL and educationally disadvantaged students.302 Again, the Charter School Plaintiffs have not shown a difference in treatment so arbitrary as to rise to a constitutional violation.
We accordingly reject all the claims of the Charter School Plaintiffs, except that we remand their claim for attorney fees along with all other claims for such fees, as discussed below.303
J. Statewide Ad Valorem Tax Claim
The trial court.held the current system imposed a statewide ad valorem tax in violation of article VIII, section 1-e. We disagree.
Article VIII, section 1-e provides: “No State ad valorem taxes shall be levied upon any property within this State.” We explained in Edgewood III: “An ad valorem tax is a state tax when it is imposed directly by the State or when the State so completely controls the levy, assessment and disbursement of revenue, either directly or indirectly, that the authority employed is without meaningful discretion.” 304 In Edgewood IV, we warned of a future where “some districts may be forced to tax at the maximum allowable rate just to provide a general diffusion of knowledge.” 305 If the cap set by the Legislature “were to become in effect a floor as well as a ceiling, the conclusion that the Legislature had set a statewide ad valorem tax would appear to be unavoidable because the districts would then have lost all meaningful discretion in setting the tax rate.”306
Analysis of this issue does not submit to simple rules or formulae. “Each case must turn on its own particulars.”307 “Meaningful discretion” in this context “is an admittedly imprecise standard.”308
In WOC II, we held the Legislature had imposed a statewide ad valorem tax. The maximum M & 0 rate at the time was $1.50. We relied on evidence that 48% of districts with 59% of students were taxing at the cap.309 Further, 67% of districts with 81% of students were taxing at or above $1.45, within five cents of the cap. As for looking at tax rates within a few cents of the cap, we explained in WOC *882/that “[a] district taxing a few cents below the maximum rate that can no longer provide an accredited education or a general diffusion of knowledge even by raising the rate to the maximum need not do so just to prove the point.”310
In today’s case, the comparable numbers are different. The maximum M & 0 rate is $1.1.7. About 24% of districts with only about 13% of students tax at the cap. About 69% of districts educating about 76% of students tax at or below $1.04, 13 cents below the cap. These figures by themselves do not suggest that districts as a whole have lost meaningful discretion.
In deciding this issue, the trial court held that the ISDs as a whole had “lost meaningful discretion to set their M & 0 rates” and “the ISD Plaintiffs collectively have also, established a systemic violation” of article VIII, section 1-e.311 The court considered the number of districts taxing at or above $1.04, finding that over 90% tax in this range. The court looked to the $1.04 rate because under current law a tax rollback election (TRE) is required to increase the rate above $1.04, and “a constitutionally adequate education cannot be left to the discretion of voters.”312 The ISD Plaintiffs also argue the $1.04 rate should be used in analyzing this issue because taxing above this rate is politically difficult for various reasons; for example, revenues raised by taxing above this rate are sometimes subject to recapture.
We do not agree with this reasoning. Article VII, section 3(e) of the Constitution states in part: “[T]he Legislature may authorize an additional ad valorem tax to be levied and collected within all school districts for the further maintenance of public free schools, and for the erection and equipment of school buildings therein; provided that a majority of the qualified voters of the district voting at an election to be held for that purpose, shall approve the tax.” The Legislature’s decision to require a THE is consistent with article VII, section 3(e)’s provision for a local election to approve such ad valorem taxes, and with its stated intent to leave such decisions to local as opposed to state authorities.
As we stated in Edgewood III, “Clearly, if the State merely authorized a tax but left the decision whether to levy it entirely up to local authorities, to be approved by voters if necessary, then the tax would not be a state tax.”313 That is what the Legislature did here. And in Edgewood IV, we concluded that a TRE requirement — requiring a rollback election for rate increases of more than six cents-was irrelevant to the statewide ad valorem tax issue because “[t]he requirement of rollback elections ... [a]t worst ... may only slow a district’s efforts to reach a desired rate.”314 Obtaining voter approval for a tax increase may pose political challenges, but we agree with the State that letting “local voters decide whéther to raise taxes is the exact ■opposite of a state-imposed property-tax rate.” To argue that the current system is the product of, and operates in, the rough- and-tumble world of Texas politics in no way distinguishes the current system. Further, the ISDs implicitly recognize that the Legislature’s actions in 2006 to compress rates, increase State funding, and impose the $1.17 cap had the effect, at least by original design, of increasing school district taxing discretion. They gay in their joint post-submission letter brief: *883“The Legislature met in special session in spring 2006 and passed legislation that provided school districts with additional money and additional taxing discretion.” Proper analysis of this issue in a manner consistent with our precedents should consider how many districts are taxing at or near the statutory cap,, whether or not taxing at the cap requires a TRE.
In WOC //we also considered the system’s “capacity,” citing evidence that districts as a whole were spending over 97% of the revenue that would be available if every district taxed at the cap.315 The evidence was apropos to whether districts had lost meaningful discretion in setting the tax rate. The trial court in today’s case also looked to several measures of capacity, but we find the analysis unconvincing. First, the trial court analyzed capacity based on what the system could raise if districts taxed at the $1.04 rate and comparing that amount to measures of the costs of an adequate education. Again, we think the $1.04 rate has no special significance in this context.
The trial court also considered what the system could raise if districts taxed at $1.17. But here we see a second problem with the trial court’s capacity analysis. It measured the taxing capacity of the system against what it had decided was the amount needed to fund an adequate system under the “cost-of-adequacy” views of experts Odden and Moak.316 As discussed above with respect to the adequacy issue, we do not agree that the trial court properly calculated the dollar cost of an adequate education based on the views of these experts. As with the financial efficiency analysis, this problem again bleeds over into another issue.
A better approach to capacity in today’s case, employed in WOC II, is to simply consider the additional revenue that would be available if every district taxed at the maximum rate. The State’s expert Lisa Dawn-Fisher provided such data, showing that an additional $2.3 billion in capacity, or 6-7% of total capacity of the FSP, was available in 2012. So the untapped capacity has roughly doubled from the 3% noted in WOC Z/.317 And since that calculation the 83rd and 84th Legislatures have added more funding to the system.
In our 2005 analysis of this issue in WOC II, addressing the system as it existed in 2003-04, we also considered the more than $1 billion in local tax revenues that was recaptured and redistributed by the State,318 because deciding whether a tax is a state tax turns in part on the State’s control over the distribution of the revenue raised by the tax.319 “The State’s control of this [recaptured] local revenue is a significant factor in considering whether local taxes have become a state property tax.” 320 The State points to evidence that the percentage of FSP funding that is subject to recapture has declined from 4.3% of total FSP funding in 2006 to a projected 2.8% "in 2015. The average amount recaptured per district has declined from about $9 million to about $5 million.
*884The Plaintiffs make varied arguments to the effect that the current system imposes a statewide ad valorem tax claim as to individual districts. Much of this argument, more or less, restates the arguments that the current system is not meeting the adequacy requirement, arguments we have rejected above. In WOC I, we held that “a single district states a claim under article VIII, section 1-e if it alleges that it is constrained by the State to tax at a particular rate.”321 The trial court in today’s ease did not actually hold that article VIII, section 1-e was violated as to any particular district. It made findings of fact as to 35 focus districts,322 12 of which were taxing at the maximum $1.17 rate, and found that every one lacked “meaningful discretion to raise tax rates to provide local enrichment programs to its students.” But it issued only two conclusions of law, repeated in its final judgment, on this claim. In COL 76, it concluded that the ISD Plaintiffs had established a violation of article VIII, section 1-e, because “the ISD Plaintiffs” — all of them — had “lost meaningful discretion to set their M & 0 rates.” In COL 77, it concluded that the ISD Plaintiffs had also established a “systemic” violation of this provision.323 The parties are also unclear in explaining what the appropriate relief would be in a case where the trial court held that only an individual district or some number of them had established a violation of article VIII, section 1-e. But we need not dwell on these difficulties.
In WOC II, we held that a systemic violation of article VIII, section 1-e had been established, although in the course of our analysis we noted evidence of “focus districts” struggling to meet accreditation requirements.324 We do not foreclose the possibility that a violation of article VIII, section 1-e can be established by a single district. But to establish such a claim, a single district would have to show that it has no meaningful discretion but to tax at or near the State-imposed cap simply to provide a general diffusion of knowledge (which, under the Legislature’s current scheme, means meeting State-imposed accreditation or other requirements). Some spending on optional enrichment programs might not necessarily defeat the claim. As we stated in WOC II, “The State cannot provide for local supplementation [by statute], pressure most of the districts by increasing accreditation standards in an environment of increasing costs to tax at maximum rates in order to afford any supplementation at all, and then argue that it is not controlling local tax rates.” 325 We also noted a distinction between a lack of meaningful discretion and a lack of any discretion whatsoever.326 But we think it follows from our precedents that if a district has meaningful discretion to spend money on optional enrichment programs and is in fact exercising that discretion, that district has not made the requisite showing. In this case, the *885Plaintiffs point to no district or districts that both (1) are taxing at or near the cap of $1.17, and (2) established that they had been forced to forgo most or all desired enrichment expenditures and were instead compelled to spend all or nearly all their resources to meet accreditation requirements or other State-imposed mandates. A breakdown of actual total spending on required versus optional spending, showing little or none of the latter, was not provided by any district taxing at or near $1.17. We recognize that such a breakdown might be difficult and might involve subjective allocations, but it was not presented here.
Determination of this issue is difficult and does not lend itself to simple numerical solutions.327 But it is, ultimately, a question of law we must decide. We therefore hold the current system does not impose a statewide ad valorem tax.
K. Attorney Fees
The State asks that the case be remanded to the trial court for reconsideration of attorney fees. The trial court awarded millions of dollars in attorney fees to the four ISD Plaintiffs. It denied fees to the State, the Charter School Plaintiffs, and the Intervenors. Fees were awarded under section 37.009 of the Civil Practice and Remedies Code, allowing for the award of “reasonable and necessary” attorney fees in declaratory judgment actions “as are equitable and just.” In its judgment, the trial court concluded that the awarded fees would be equitable and just even if on appeal the ISD Plaintiffs do not prevail on one or more of their claims, “because they have made significant .contributions to the public debate on school finance law through this lawsuit.” 328 The court denied fees to the State, Charter School Plaintiffs, and Intervenors “because they were predominantly non-prevailing parties,” which we take to mean, conversely, that the ISD Plaintiffs were awarded fees at least in part because they were predominantly prevailing parties.
We have recognized that an award of attorney fees is within the discretion of the trial court, but the question of whether fees are equitable and just is a question of law.329 The Plaintiffs have not prevailed on any of their claims, but we have stated that the award of attorney fees in declaratory judgment actions “is not dependent on a finding that a party ‘substantially prevailed.’ ” 330 Where the extent to which a party prevailed has changed on appeal, our practice has been to remand the issue of attorney fees to the trial court for reconsideration of what is equitable and just.331 We do so here as to all parties.
L. Continuing Jurisdiction
The trial court in its judgment states: “The Court will retain continuing jurisdiction over this matter -until the Court has determined that the State Defendants have fully and properly complied with its *886judgment and orders.” 332 The State contends there is no procedural authority under Texas law for a trial court to grant itself continuing jurisdiction, of the sort sometimes exercised by federal courts, in these circumstances.
Texas courts undoubtedly have the power to enforce their judgments.333 We have also held that a trial court generally retains jurisdiction to modify a permanent injunction if circumstances change.334 CCISD suggests the trial court retained continuing jurisdiction because it recognized that its injunction would have to be modified or vacated once the constitutional violations it found were remedied. But there is no basis for continuing jurisdiction where a take-nothing judgment — one dissolving the injunction — has been rendered on appeal. The only issue left for the trial court to consider in light of our holdings is attorney fees. There is no need for the trial court to monitor compliance with its judgment through ongoing jurisdiction where there is no judgment requiring the defendants to engage in or refrain from activities specified in the judgment. The judgment we render on all constitutional claims and all claims for injunctive relief dispenses with the need for any sort of general continuing trial court jurisdiction over this matter beyond the jurisdiction courts always have to enforce their judgments. We reverse this portion of the trial court judgment insofar as it directs otherwise.
III. Conclusion
“it shall be the duty of the Legislature ...”335
The Framers of our Texas Constitution . placed the responsibility for education poli-eymaking squarely with the Legislature. Those decisions are not immune from judicial review. Lawmakers decide if laws pass, and judges decide if those laws pass muster. But our lenient standard of review in this policy-laden area counsels modesty. The judicial role is not to second-guess whether our system is optimal, but whether it is constitutional.
Our Byzantine school funding “system” is undeniably imperfect, with immense room for improvement. But it satisfies minimum constitutional requirements. Accordingly, we decline to usurp legislative authority by issuing reform diktats from on- high, supplanting lawmakers’ policy wisdom with our own.
The Texas Legislature, the center of policymaking gravity, is not similarly bound. And smartly so. Our Constitution endows the people’s elected representatives with vast discretion in fulfilling their constitutional duty to fashion a school system fit for our dynamic and fast-growing State’s unique characteristics. We hope lawmakers will seize this urgent challenge and upend an ossified regime ill-suited for 21st century Texas.
The trial court’s judgment is affirmed in part and reversed in part. We remand the issue of attorney fees to the trial court. We render a take-nothing judgment on all of the Plaintiffs’ and Intervenors’ other claims.
JUSTICE GUZMAN filed a concurring opinion, in which JUSTICE LEHRMANN joined. JUSTICE BOYD filed a concurring opinion, in which JUSTICE LEHRMANN and JUSTICE DEVINE joined.. Tex. Educ. Code § 4.001(a).
. Id. § 28.001.
. 176 S.W.3d 746, 764 (Tex.2005).
. Tex. Educ. Code §§ 12.111(a)(1), 28.002(a)(1). Social studies consists of Texas, United States, and world history, government, economics, and geography. Id. § 28.002(a)(1)(D).
. Id. § 28.002(a)(2).
. Id. § 29.051.
. The SBOE consists of 15 elected members. Id. § 7.101.
. Id. § 28.002(c).
. 176 S.W.3d at 765.
. Tex. Educ. Code § 28.008(d).
. Id. § 39.024(a).
. Id. § 28.002(c).
. 176 S.W.3d at 765.
. Id.
. See Tex. Educ. Code § 39.023(c-4).
. See id. § 39.023(c).
. The Commissioner of Education is appointed by the Governor and heads the Texas Education Agency (TEA). Id. §§ 7.002,7.051.
. 176 S.W.3d at 766.
. Tex. Educ. Code § 39.054(a).
. Id. §§ 39.053(c), 39.054(b).
. Id. § 39.052(b)(1).
. Id. § 39.052(f).
. Id. §§ 39.102, 39.104.
. Id. § 39.103.
. WOC II, 176 S.W.3d at 758.
. Tex. Educ. Code § 42.002(a)(1).
. Id. § 42.002(b)(1).
. See id. § 42.2516(a).
. Id. § 45.003(d).
. Tex. Tax Code § 26.08.
. See Tex. Educ. Code § 42.101.
. Id. § 42.102. The trial court found that “[t]he CEI is based on five school district characteristics that were measured in 1989— 90 — district size, type, percentage of low income students, average beginning salary in surrounding districts, and location in a county with a population of less than 40,000." Trial court Finding of Fact (FOF) 597.
. Id. §§ 42.103, 42.105.
. Id. § 42.101(c).
. Id.
. Id. §§ 42.151 — . 160.
. Id. § 42.152. The statute refers to "educationally disadvantaged" students, but we conform to the parties’ reference to these students as economically disadvantaged.
. Id. § 42.153.
. Id. § 42.252.
. FOF 42.
. See Tex. Educ Code § 42.302(a).
. Id. § 42.302.
. See id. § 42.302(a-l)(l).
. Id. § 42.302(a-l)(2).
. 917 S.W.2d 717 (Tex.1995).
. Tex. Educ. Code § 41.003; Edgewood IV, 917 S.W.2d at 728.
. See Tex. Educ. Code § 41.002(a); FOFs 45-46.
. FOF46.
. See Tex. Educ. Code § 41.002(a)(3).
. See FOF 50; see also Tex. Educ. Code § 42.2516.
. See Tex. Educ. Code § 45.003.
. Id. § 45.0031(a).
. Id. §§ 46.001-.003.
. FOF 227.
. See Tex. Educ. Code §§ 46.032-034.
. The Texas Taxpayer and Student Fairness Coalition; Alief ISD; Canutillo ISD; Elgin ISD; Greenville ISD; Hillsboro ISD; Hutto ISD; Lake Worth ISD; Little Elm ISD; Nac-ogdoches ISD; Paris ISD; Pflugerville ISD; Quinlan ISD; Stamford ISD; Saij Antonio ISD; Taylor ISD; Van ISD; Randy Pittenger; Chip Langston; Norman Baker; Brad King; and Shelby Davidson, individually and as next friend of her three children. TTSFC is composed of 443 school districts. FOF 2.
. Calhoun County ISD; Abernathy ISD; Ar-ansas County ISD; Frisco ISD; Lewisville ISD; and Richardson ISD.
. Edgewood ISD; Harlingen Consolidated ISD; La Feria ISD; McAllen ISD; San Benito Consolidated ISD; Yolanda Canales, individually and on behalf of her children; Arturo Robles, individually and on behalf of his child; Jessica Romero, individually and on behalf of her children; and Araceli Vasquez, individually and on behalf of her children.
. Fort Bend ISD; Abilene ISD; Albany ISD; Allen ISD; Amarillo ISD; Angleton ISD; Austin ISD; Balmorhea ISD; Beaumont ISD; Bluff Dale ISD; Brazosport ISD; Carthage ' ISD; Channelview ISD; Clear Creek ISD; Cleveland ISD; College Station ISD; Coppell ISD; Corsicana ISD; Crosby ISD; Cypress-*840Fairbanks ISD; Dallas ISD; Damon ISD; Decatur ISD; Deer Park ISD; Denton ISD; Dumas ISD; Duncanville ISD; East Central ISD; Ector County ISD; Edna ISD; Fort Worth ISD; Galena Park ISD; Goose Creek Consolidated ISD; Graford ISD; Hardin-Jefferson ISD; Hays Consolidated ISD; Hemp-stead ISD; Highland ISD; Houston ISD; Huffman ISD; Humble ISD; Katy ISD; Keller ISD; Kenedy ISD; Kerrville ISD; Kings-ville ISD; Klein ISD; La Marque ISD; La Porte ISD; Lamar Consolidated ISD; Leggett ISD; Liberty ISD; McKinney ISD; Midland ISD; New Caney ISD; North East ISD; Northside ISD; Pampa ISD; Pasadena ISD; Pearland ISD; Perrin-Whitt Consolidated ISD; Pine Tree ISD; Pleasant Grove ISD; Rice Consolidated ISD; Rockdale ISD; Round Rock ISD; Royal ISD; Santa Fe ISD; Schertz-Cibolo-Universal City ISD; Shary-land ISD; Sheldon ISD; Splendora ISD; Spring Branch ISD; Stafford Municipal School District; Sudan ISD; Sweeny ISD; Trent ISD; Troup ISD; Waco ISD; Weather-ford ISD; West Orange-Cove Consolidated ISD; and Woodville ISD.
.Hence, we refer to the ISD Plaintiffs other than CCISD as the Financial Efficiency Plaintiffs in our discussion of the financial efficiency issue below.
. Texas Charter Schools Association; Mario Flores, individually and as parent and next friend of his child; Christopher Baerga, individually and as parent and next friend of his child; Dana Allen, individually and as parent and next friend of her child; Jason Christensen and Sarah Christensen, individually and as parents and next friends of their children; and Brooks Flemister, individually and as parent and next friend of his child.
. For convenience we sometimes refer to all of the Plaintiffs and Intervenors, or the subset of these parties who raised a particular argument, simply as “Plaintiffs.”
. Joyce Coleman, individually and as next friend of her children; Danessa Bolling, individually and as next friend of her child; Lee Beall and Aliena Beall, individually and as next friends of their children; Joel Smedsh-ammer and Andrea Smedshammer, individually and as next friends of their children; Darlene Menn, individually and as next friend of her child; Texans for Real Efficiency and Equity in Education; and Texas Association of Business.
. Judgment ¶¶ 11(1), 11(2), 111(4).
. Tex Educ. Code § 29.052(1). See also FOF 332.
. FOF 13.
. FOF 245.
. The Findings of Fact and Conclusions of Law are found at 2014 WL 4254969.
. Judgment ¶ IX(1); COL 95.
. Id. KXI.
. See Tex. Gov't Code § 22.001(c).
. See W. Orange-Cove Consol. Indep. Sch. Dist. v. Alanis, 107 S.W.3d 558, 585 (Tex. 2003) ("We do not agree with the dissent that the importance of stare decisis can be minimized in this area. For fourteen years the Legislature has worked to bring the public school finance system into conformity with constitutional requirements as declared by this Court. To announce now that we have simply changed our minds on matters that have been crucial to the development of the public education system would not only threaten havoc to the system, but would, far more importantly, undermine the rule of law to which the Court is firmly pledged.”).
. Edgewood Ind. Sch. Dist. v. Kirby, 111 S.W.2d 391 (1989).
. Id. at 392.
. Id. at 398.
. Edgewood Indep. Sch. Dist. v. Kirby, 804 S.W.2d 491 (Tex.1991).
. Id. at 496 (citation, footnote omitted).
. Id. at 499.
. Carrollton-Farmers Branch Indep. Sch. Dist. v. Edgewood Indep. Sch. Dist., 826 S.W.2d 489 (Tex.1992).
.Id. at 498.
. Id. at 500.
. Id.
. Id. at 502.
. Edgewood Indep. Sch. Dist. v. Meno, 917 S.W.2d 717 (Tex.1995).
. Id. at 727-28.
. Id. at 730.
. Id.
. Id. at 738.
. W. Orange-Cove Consol. Indep. Sch. Dist. v. Alanis, 107 S.W.3d 558 (Tex.2003).
. Id. at 576.
. Id. at 578 (quoting Edgewood III, 826 S.W.2d at 502).
. Id. at 578-83.
. Neeley v. W. Orange-Cove Consol. Indep. Sch. Dist., 176 S.W.3d 746 (Tex.2005).
. Id. at 772-83.
. Id. at 752-53.
. Id. at 753.
. Id. at 790.
. Id. at 794-98.
. Id. at 795.
. For example, high school students Zaakir Tameez and Amy Fan, with the help of other students, have filed an excellent amicus brief, stating: "We encourage the justices to remember their time as children, and reflect on what role music, drama, art, and sports played in their lives as kids.... We are willing to bet most justices of this Court remember their good and bad teachers from high school to this day.” We do remember.
. Tex. R. App. P. 57.2; see also Tex. Gov’t Code § 22.001(c).
. WOC II, 176 S.W.3d at 785.
. Id.; see also Tex. Workers’ Comp. Comm'n v. Garcia, 893 S.W.2d 504, 520 (Tex.1995) (holding that trial court’s fact findings play a "limited role” in reviewing the constitutionality of a legislative enactment and whether "the Legislature has acted arbitrarily”).
. WOC II, 176 S.W.3d at 785.
. Edgewood TV, 917 S.W.2d at 725; Edgewood III, 826 S.W.2d at 493.
. 176 S.W.3d at 784-85.
. Id. at 790.
. Id. at 785.
. Id.
. Id. at 778.
. 777 S.W.2d at 394.
. Id.
. 107 S.W.3d at 563-64, quoted in WOC II, 176 S.W.3d at 777.
. 176 S.W.3d at 778.
. Id.
. Edgewood IV, 917 S.W.2d at 730 n. 8.
. WOC II, 176 S.W.3d at 774.
. Id. at 776 (noting that individuals, including taxpayers, "were free to intervene” and that “individuals would have standing to raise the claims in this case”).
. Heckman v. Williamson Cnty., 369 S.W.3d 137, 154 (Tex.2012).
. Id. at 155.
. See, e.g., WOC II, 176 S.W.3d at 753-54 (describing trial court injunction); Edgewood III, 826 S.W.2d at 523 n. 42; Edgewood II, 804 S.W.2d at 493, 498-99 & nn. 16-17 (describing trial court's injunction in Edgewood I, as modified by the Court in Edgewood II ).
. WOC II, 176 S.W.3d at 771 (quoting trial court injunction enjoining the State from “giving any force and effect to the sections of the Education Code relating to the financing of public school education (Chapters 41 and 42 of the Education Code) and from distributing any money under the current Texas school financing system until the constitutional violations are remedied’’).
. See Andrade v. NAACP of Austin, 345 S.W.3d 1, 16 (Tex.2011) (stating that courts cannot direct what laws the Legislature must enact to comply with a constitutional provision); Terrazas v. Ramirez, 829 S.W.2d 712, 720 (Tex.1991) (“After a legislative plan has been invalidated, respect for the separation of powers explicitly recognized in article II, section 1 of our Constitution requires that the Legislature be given a reasonable opportunity to enact a substitute statute.”).
. See, e.g., Edgewood IV, 917 S.W.2d at 726-27 (describing legislative responses to Edgewood I, Edgewood II, and Edgewood III decisions).
. Perry v. Del Rio, 66 S.W.3d 239, 250 (Tex.2001).
. WOC II, 176 S.W.3d at 753.
. Id. at 787.
. Id. at 788.
. Id. at 787; WOC I, 107 S.W.3d at 581; Edgewood IV, 917 S.W.2d at 730 & n. 8.
. WOC II, 176 S.W.3d at 790.
. See FOFs 610-20.
. FOFs 619-20.
. FOF 621.
. FOF 622.
. FOF 631.
. FOF 632.
. See, e.g., FOFs 633-35; COL 33; Judgment ¶ 111(2).
. FOF 636.
. WOC II, 176 S.W.3d at 788.
. For example, the second adequacy declaration states:
The ISD Plaintiffs have shown that the cost of meeting the constitutional mandate of adequacy (the “general diffusion of knowledge”) exceeds the maximum amount of funding that is available to them at the $1.04 M & O tax rate (the highest rate accessible without a TRE). Accordingly, THIS COURT DECLARES the State’s school finance system fails to satisfy the Article VII, Section I adequacy requirement *851as to the ISD Plaintiffs districts. The ISD Plaintiffs also have shown that the cost of meeting the constitutional mandate of adequacy exceeds the amount of funding that is or would be available to them at the maximum $1.17 M & O tax rate. Accordingly, THIS COURT DECLARES the State's school finance system fails to satisfy the Article VII, Section I adequacy requirement as to the ISD Plaintiffs districts.
Judgment ¶ 11(3).
. WOC II, 176 S.W.3d at 785.
. 777 S.W.2d at 393.
. 176 S.W.3d at 788.
. 411 U.S. 1, 93 S.Ct. 1278, 36 L.Ed.2d 16 (1973).
. Id. at 42-43, 93 S.Ct. 1278.
. James S. Coleman et al„ Equality of Educational Opportunity (U.S. Office of Education 1966), available at http://files.eric.ed.gov/ fulltext/EDO 12275 .pdf.
. Barbara J. Kiviat, The Social Side of Schooling, Johns Hopkins Magazine (April 2000), available at http://pages.jh.edu/jhumag/ 0400web/l 8.html.
. Colehjan, supra note 146, at 8. See also id. at 325 ("Taking all these results together, one implication stands out above all: That schools bring little influence to bear on a child's achievement that is independent of his background and general social context; and that this very lack of an independent effect means that the inequalities imposed on children by their home, neighborhood, and peer environment are carried along to become the inequalities with which they confront adult life at the end of school.").
. Id. at 21-22.
. Adam Gamoran & Daniel A. Long, Equality of Educational Opportunity: A 40-Year Retrospective 3, 19 (Wis. Ctr. for Educ. Research, Working Paper No. 2006-9, 2006), available at http://wcer-web.ad.education. wisc.edu/docs/working-papers/Working_ Paper_No_2006_09.pdf.
. Id. at 8.
. For example, as support for "a significant causal relationship between school funding and improvements in long-term educational outcomes," amici curiae Center for Public Policy Priorities et al. cite C. Kirabo Jackson et al., The Effect of School Finance Reforms on the Distribution of Spending, Academic Achievement, and Adult Outcomes (Nat’l Bureau of Econ. Research, Working Paper No.’ 20118, 2014). Amici Curiae Senator Phil Gramm and Stacy Hock quote another study for the opposite conclusion, that "there is essentially no link between state education spending (which has exploded) and the performance of students at the end of high school (which has generally stagnated or declined).” Andrew J. Coulson, State Education Trends: Academic Peifonnance and Spending over the Past 40 Years 4 (Cato Inst. Policy Analysis No. 746, March 18, 2014), available at http:// obj ect. cato. org/sites/cato .org/files/pubs/pdl/pa 746.pdf. Amicus Curiae The Goldwater Institute quotes the conclusion of another researcher and expert witness for the Interve-nors that “no currently available evidence shows that past judicial actions about school finance — either related to equity or adequacy- — -have had a beneficial effect on student performance,” and that court-ordered spending increases have not boosted student performance for this "quite obvious” reason: "Measures of school resources do not provide guidance either about the current quality of schools or about the potential for improving matters.” Eric A. Hanushek, Introduction, in Courting Failure: How School Finance Lawsuits Exploit Judges’ Good Intentions and Harm Our Children xxiii-xxiv (Eric A. Hanushek *853ed.2006), available at http://hanushek. stanford.edu/sites/defaulVfiles/publications/ courting_failure.pdf. As an expert in today’s case, Hanushek offered nationwide data showing that while, over several decades, real expenditures per student have steadily increased, test scores have essentially remained flat. Gorman and Long, supra note 150, cite seven Hanushek articles in their bibliography, more than any other author including Coleman himself. As with Coleman, we do not endorse Hanushek’s views, but there is little denying he is a recognized scholar in the field of educational funding. The United States Supreme Court has cited Hanushek as its lead authority in noting “a growing consensus in education research that increased funding alone does not improve student achievement.” Horne v. Flores, 557 U.S. 433, 464 n. 17, 129 S.Ct. 2579, 174 L.Ed.2d 406 (2009). The trial court, however, adjudged Hanushek's views unpersuasive. E.g., FOFs 648-52.
. 917 S.W.2d at 731 n. 10.
. FOF 628.
. Edgewood IV, 917 S.W.2d at 731 & n. 12.
. Nor was this issue tried in the trial court. Justice Spector's dissent explained that the trial court had "severed out what it called the 'adequacy issues' including the issue of 'whether the legislature appropriates sufficient funds for districts to provide a constitutionally, minimally acceptable education.” Id. at 768 (Spector, J., dissenting).
. Id. at 730.
. E.g., id. at 731 (recognizing that districts had "the funds necessary for a general diffusion of knowledge”); WOC II, 176 S.W.3d at 785 ("It would be arbitrary, for example, for the Legislature to define the goals for accomplishing the constitutionally required general diffusion of knowledge, and then to provide insufficient means for achieving those goals.”); id. at 788 ("[I]t is useful to consider how funding levels and mechanisms relate to better-educated students.”).
. For example, COL 32 holds:
This Court rejects the notion that the general diffusion of knowledge requires expenditures only in the instructional program described in statute and that other expenditures are merely “extraneous.” A district cannot provide a constitutionally adequate education without a sufficient support network, which may include, among other things, (a) adequate and well-maintained facilities, (b) nurses to keep students healthy, (c) security guards in certain schools to keep students safe, (d) guidance counselors to help students with course selection and with planning for college and careers, (e) paraprofessionals to provide vital assistance to teachers, (f) libraries with both print and electronic resources and librarians to assist students and teachers in using these resources, (g) tutors to help struggling students, and (h) transportation. (See supra Part I.C.3.d (FOF 575, et seq.).) In some districts, the general diffusion of knowledge may additionally require programs designed to keep students in school until graduation.
. The trial court referred to best practices in FOFs 488, 612, 615, 1190, 1191, 1197, 1199, and COL 31. By way of example, COL 31 holds:
Measures that superintendents and other experts have identified as best practices to -attain the legislatively mandated outcome objective of college and career readiness include, among other things, (a) manageable class sizes, particularly for economically disadvantaged and ELL populations, (b) preschool programs of sufficient quality to provide a “head start” to special needs students, (c) remedial and literacy programs to help ELL, economically disadvantaged, and other special needs students, including *856summer school and after school programs, (d) salaries that can attract and retain sufficient numbers of qualified teachers, and (e) vocational and career courses to give those students that cannot attend college an opportunity to succeed in post-secondary employment settings.... The Court identifies these practices as examples of ways to accomplish the general diffusion of knowledge, not to order the Legislature to adopt these practices as per se constitutional; however, where, research supports a practice as effective, an approach that undermines those practices, without replacing them with another approach that is supported by research as reasonable, could be considered arbitrary and unconstitutional.
. For example, FOF 615 finds:
Some of the key strategies recommended by Dr. Odden's evidence-based approach include (1) core teachers for class sizes of fifteen in kindergarten through third grade and of twenty-five in grades four through twelve, (2) full-day kindergarten; (3) specialist teachers at 20% of core teachers at elementary and middle schools and 33% at high school, and (4) instructional coaches to provide professional development, including classroom observation and feedback for teachers.... Dr. Odden's evidence-based model provides additional resources, including tutors and summer school, which are targeted toward struggling students.... These strategies are supported by the evidence as “best practices” and are credible factors for determining the cost of education. Dr. Odden testified that Texas is unlikely to substantially improve student performance without implementing the core interventions recommended by his evidence-based model.
. FOF 620.
. FOF 615.
. COL 30.
. Edgeood IV, 917 S.W.2d at 725 ("[W]e begin with the presumption that [the system] is constitutional; the burden of proof is on those parties challenging this presumption.”).
.See Tex. Educ. Agency, Academic Excellence Indicator System, http://ritter.tea.state. tx.us/perfreport/aeis.
.The trial court's Judgment at paragraph 11(2) declared:
The Edgewood ISD Plaintiffs have further shown that the costs of providing a general diffusion of knowledge to economically disadvantaged and English Language Learner students exceed the funding provided through the current system, due to the arbitrarily designed and insufficient weights for those students. This defect coupled with the arbitrarily designed and insufficient Foundation School Program funding made available to districts like the Edgewood ISD Plaintiffs cumulatively prevent those districts from generating sufficient resources to accomplish a general diffusion of knowledge for the State’s economically disadvantaged and English Language Learner students. Accordingly, THIS COURT DECLARES that the Texas school finance system violates the "make suitable provision" clause in Article VII, Section 1 of the Texas Constitution because the system is not "structured, operated, and funded so that it can accomplish its purpose [of providing a general diffusion of knowledge] for [economically disadvantaged and English Language Learner] children." WOC II, 176 S.W.3d at 753.
Paragraph 111(4) declared:
The Edgewood ISD Plaintiffs, the TTSFC Plaintiffs, and the Fort Bend ISD Plaintiffs have further shown that economically disadvantaged students and English Language Learner students are not achieving a general diffusion of knowledge and that the cost of providing a general diffusion of knowledge to these students exceeds the amount of funding made available for their education under the current school finance system. The Court concludes the funding for economically disadvantaged and English Language Learner students is inadequate and arbitrary. Accordingly, THIS COURT DECLARES the current public school finance system is inadequate for the provision of a general diffusion of knowledge for economically disadvantaged and English Language Learner students under Article VII, Section 1 of the Texas Constitution.
. 176 S.W.3d at 789.
. Id. at 790 (emphasis added, footnote and internal quotation marks omitted).
.Id. at 788 (emphasis added).
. Id. at 801 (Brister, J., dissenting).
. South Carolina v. United States, 199 U.S. 437, 449, 26 S.Ct. 110, 50 L.Ed. 261 (1905).
. For example, FOF 520 finds:
Based on the output data described above in Parts 'I.C.2.a.iii (FOF 298, et seq.) and I.C.2.b.iii (FOF 349, et seq.), the Court finds that economically disadvantaged and ELL students are not achieving a general diffusion of knowledge. The inability of districts to offer the necessary interventions {see supra Part I.C.2.C (FOF 379, et seq.)) to help these populations overcome the educational obstacles they face {see supra Parts I.C.2.a.i (FOF 277, et seq.) and I.C.2.b.i (FOF 333, et seq.)) means that school districts are not able to provide these students with a meaningful opportunity to achieve a general diffusion of knowledge. Therefore, the Court finds that the education system is constitutionally inadequate as to economically disadvantaged and ELL students.
.The Report states:
This analysis concentrated on the educational opportunities offered by the schools in terms of their student body composition, facilities, curriculums, and teachers. This emphasis, while entirely appropriate as a response to the legislation calling for the survey, nevertheless neglects *860important factors in the variability between individual pupils within the same school; this variability is roughly four times as large as the variability between schools. For example, a pupil attitude factor, which appears to have a stronger relationship to achievement than do all the "school" factors together, is the extent to which an individual feels that he has some control over his own destiny.
Coleman, supra note 146, at 22-23.
. WOC II, 176 S.W.3d at 788.
. Tex. Educ. Code § 42.152(a).
. Id. § 42.153(a).
. See Tex. Educ. Code §§ 42.102 (CEI adjustment), 42.103-.104 (small and mid-sized district adjustment), 42.105 (sparcity adjustment).
. See supra note 32 and accompanying text.
. FOF 598.
. WOC II, 176 S.W.3d at 788.
. FOF 1470.
. COLs 41, 78-79. While the trial court issued these conclusions in assessing suitability rather than adequacy, the two requirements overlap, as we note in the next part of our opinion. The two elements particularly overlap in this case because the trial court's suitability analysis was tied to its adequacy determination that the school system was underfunded. COLs 41, 78, and 79 all expressly concluded that the system was unsuitable because it did not provide “sufficient resources to accomplish a general diffusion of knowledge.”
. FOF 404.
. WOC II, 176 S.W.3d at 753 (emphasis added).
. Id. at 788.
. Judgment ¶ III(l).
. WOC II, 176 S.W.3d at 778.
. Id.
. Id. at 790.
. Tex. Educ. Code §§ 39.054(b), 39.053(c)(1).
. Id. §§ 39.052, 39.104.
. 917 S.W.2d at 730.
. 107 S.W.3d at 581.
. Id.
. Edgewood IV, 917 S.W.2d at 730 n. 8 ("This is not to say that the Legislature may define what constitutes a general diffusion of knowledge so low as to avoid its obligation to make suitable provision imposed by article VII, section 1.”); WOC I, 107 S.W.3d at 581 ("[A]n accredited education may provide more than a general diffusion of knowledge, or vice versa[.]”).
. 2014 Accountability System State Summary, http://ritter.lea.state.tx.us/perfreport/ account/2 014/statesummary .html.
. 2015 Accountability System State Summary, http://ritter.tea.state.tx.us/perfreport/ account/2015/ statesummary.html.
. 176 S.W.3d at 769.
. FOF 117.
. 176 S.W.3d at 788.
. FOF 120 n.32.
. 176 S.W.3d at 766-67.
. Id. at 789.
. Id. at 766.
. FOF 89.
. FOF 130.
. Id.
. FOF 140.
. WOC II, 176 S.W.3d at 768.
. Id. at 789.
. The director of the TEA student assessment division testified:
I think any time we move to a new — a new system, whether it is the assessment system, whether it’s the curriculum standards, there’s a certain sort of settling in period, you know, if you will. And we also move— *866every time we move to a new assessment system or new curriculum standards, generally it’s because it becomes more rigorous. And educators need time in order to adapt to the new cognitive demands. They need time for professional development. Students need time to become familiar with the assessments. And so we allow a phase-in'' period so that all of those things can happen before we would hold students'to the final recommended standards.'
She also testified that "student performance in the first year of a new [testing] program tends to be lower than it was in the last year of the previous program, and we see increases over time.”
. FOF 140.
. WOC II, 176 S.W.3d at 788 (emphasis added).
. Texas Education Agency, Class of 2015 STAAR End-of-Course Exam Passing Rate Hits 92 Percent (May 29, 2015), htlp://tea. texas. go v/About_TEA/News_ancLMultimedia/ Press_Releases/2015/Class_of_2015_ STAARB>_end-of-course_exam_passing_rate_ hits_92_percent/. This passage rate is the graduation passing rate, a phase-in rate, not the final Level II standard passing rate.
. Id.
. 176 S.W.3d at 768-69, 789.
. FOFs 170-175.
. Tex. Educ. Code §§ 39.052(b)(1), 39.053(c)(3), 39.054(b).
. Fort Bend points to other data that the Texas graduation rate was 75.4% and ranked 28th among the States. However, this data is from 2009 and the expert presenting it testified that “[t]he calculation method that the U.S. Department of Education is presently using ... is the 'superior way to calculate graduation rates.”
. http://nces.ed.gov/ccd/tables/ACGR_2010-ll_to_2012-13.asp.
. http://nces.ed.gov/ccd/tables/ACGR_RE_ and_characteristics_2012-13 .asp.
. 176 S.W.3d at 769; see also id. at 789 (noting high dropout rates in assessing adequacy requirement).
. See supra note 220 and infra.
. WOC II, 176 S.W.3d at 789-90.
. Id. at 790.
. Id. at 793.
. Id. at 794.
. Edgewood IV, 917 S.W.2d at 736.
. WOC II, 176 S.W.3d at 793.
. Id. at 753.
. Id. at 793.
. WOC I, 107 S.W.3d at 584.
. WOC II, 176 S.W.3d at 794.
. Id. at 793 (emphasis added).
. WOC I, 107 S.W.3d at 563-64.
. The Edgewood Plaintiffs appear to agree, contending that the suitability requirement focuses on whether "the entire system” is accomplishing it goals.
. 176 S.W.3d at 794.
. Tex. Educ. Code § 28.002(c).
. Id. §§ 39.052(f), 39.102-.104.
. See FOF 125 (finding that system was unsuitable because "[t]he Legislature failed to provide additional financial support with the introduction of the STAAR regime”); COL 41 (holding system unsuitable because it does not "generate sufficient resources to accomplish a general diffusion of knowledge”); COL 42 (holding system unsuitable because it "bears no relationship to the actual cost of providing access to a constitutionally adequate education”); Judgment ¶ 11(1) (declaring system unsuitable because "the costs of providing a general diffusion of knowledge exceed the funding provided”).
. We refer to those Plaintiffs alleging violation of the financial efficiency requirement as the Financial Efficiency Plaintiffs.
. 777 S.W.2d at 397.
. 804 S.W.2d at 496.
. See Edgewood I, 777 S.W.2d at 397 (holding that financial efficiency does not "require a per capita distribution”).
. Edgewood IV, 917 S.W.2d at 730-31.
. Edgewood II, 804 S.W.2d at 500.
. 777 S.W.2d at 392-93.
. 917 S.W.2d at 731.
. Id.
. Id. at 728, 731.
. Id. at 731 n. 12.
. 176 S.W.3d at 792.
. Id. at 756. .
. Id. at 762.
. The briefing indicates there is some confusion about this evidence from WOC II. Our review of the WOC II record indicates that these numbers reflect actual spending based on tax rates the districts had chosen. The numbers are weighted averages and were not *872limited to M & O revenues; they also included facilities funding.
. TTSFC contends, for example, that “the evidence is clear that under the current system, funding for adequacy is so low that there is no enrichment for a majority of districts.” Fort Bend says, “Even if districts taxed at the statutory cap, the vast majority of districts cannot raise the money necessary to provide a general diffusion of knowledge.” Edgewood says, “The reality, however, is that property-poor and even some property-wealthy districts rely on Tier II funding to provide the minimum adequate education mandated by the Education Code.” Presumably, excluding any revenues spent on optional enrichment from the calculation of ratios would decrease the ratios and indicate greater financial efficiency, because rich districts have more funds for enrichment than poor districts.
. Clark calculated weighted-average figures for 2013-14, using funding formulae and target revenue calculations for 2013-14 and known 2012 M & O tax rates.
. For these calculations, Clark did not use the same method she had previously used of assigning an equal number of districts to each decile, so the calculations are not strictly comparable.
.The differences between the Pierce and Dawn-Fisher numbers appear to be due to the way their exhibits identify the top and bottom 5% and whether they use a weighted or simple average. As we understand the methods, Dawn-Fisher took the number of districts — 51—comprising 5% of the total number of districts, and then calculated a weighted-average revenue per WADA figure for the richest and poorest 51 districts. The weighting adjusted for the differing sizes of districts, by dividing the total revenue for the group by the total WADA number. Pierce took .the number of poorest and richest districts needed to comprise approximately 5% of students measured by WADA. So he used' 46 districts for the bottom 5% and 111 districts for the top 5%. He then calculated a simple average revenue number for these districts, by adding up the separate per-WADA funding for each district in the group, and then dividing by the number of districts in the group. The parties dispute which method is better. The State and CCISD argue that Pierce was the only expert out of six who did not use weighted averages, and that his simple averages are distorted by a few wealthy *873districts with small numbers of students. We need not decide who has the better measure. We also note that some of the numbers we discuss are based on projections. Pierce took data for one year and then applied "parameters” for the next year “to demonstrate the effects' of the 83rd Legislature,” which, for years after 2013, changed certain “funding parameters” and "funding elements” such as "the basic allotment and the equalized wealth level and those sorts of things.” Pierce and Dawn-Fisher, for example, used fiscal year 2013 as a "data year” and fiscal year 2014 as a "parameter year” for some of their calculations.
. FOFs 1426-35.
. 176 S.W.3d at 791.
. FOF 1376.
. Edgewood I, 777 S.W.2d at 397.
. See id. at 393 (describing the richest districts as spending an average of $7,233 per students at an average tax rate of 47 cents, while the poorest spent an average of $2,978 at an average tax rate of 74.5 cents).
. 917 S.W.2d at 730-31.
. Id. at 131 n. 12.
. As described supra note 260, there are numerous parameters that can be used in making these sorts of calculations, and depending on the precise methods employed, the ultimate resulting ratios can vary.
. E.g., FOFs 1205-06, 1210-13, 1217-19, 1222-33, 1237-40, 1278, 1444-47.
. E.g., FOFs 1289-1300.
. 176 S.W.3d at 792.
. 777 S.W.2d at 395.
. 917 S.W.2d at 730.
. 176 S.W.3d at 752-53.
. Id. at 753.
. Id. at 787.
. WOC I, 107 S.W.3d at 571.
. The State says in its briefing that, setting aside its jurisdictional arguments, "the system conceivably could produce the required results and thus be constitutionally adequate, yet be inefficient because it arbitrarily wastes too many resources to achieve those results.”
. WOC II, 176 S.W.3d at 777 (quoting WOC I, 107 S.W.3d at 563-64).
. Edgewood IV, 917 S.W.2d at 726.
. Id. at 747 (quoting Edgewood I, Ill S.W.2d at 399).
. See Tex. Educ. Code §§ 61.221-.230.
. Edgewood IV, 917 S.W.2d at 726.
. WOC II, 116 S.W.3d at 779.
. Id. at 802 (Brister, J., dissenting).
. Edgewood IV, 917 S.W.2d at 730 n. 8.
. The Charter School Plaintiffs also complained to the trial court about the cap, but do not urge this claim on appeal.
. Tex. Educ. Code § 12.101(b-l), (b-2).
. See id. §§ 12.002, 12.011-.04 (home-rule school district charters), 12.051-065 (campus or campus program charters); 12.101 — . 136 (open-enrollment charters). A fourth category of charter schools, "university” charter schools, are generally subject to provisions governing open-enrollment charter schools. Seeid.§§ 12.151-156.
. COLs 61, 89.
. Tex. Educ. Code §§ 46.003, .012, .032, .036.
. 777 S.W.2d at 397 (emphasis added).
. WOC II, 176 S.W.3d at 785.
. FOF 1497.
. Tex. Educ. Code § 12.001(5).
. See id. §§ 12.103-.104 (limiting requirements of Education Code on open-enrollment charter schools to specified provisions), 21.002(a) (contract requirements), 21.055 (teacher certification and degree requirements), 21.402 (salary requirements).
. See id. § 25.111 (student-teacher ratios), 25.112 (class size).
. See id. § 42.158.
. See id. §§ 12.106(a-l) & (a-2), 42.102-.105, 42.302.
. WOC II, 176 S.W.3d at 790.
. FOF 1495.
. See Tex. Educ. Code §§ 12.106(a), 42.151-160.
. The State argues, as a separate ground for rejecting some of the claims of the Charter School Plaintiffs, that the claims are seeking to reform the charter schools’ contracts with the State, and that suits on a contract are barred by sovereign immunity. We do not reach this issue.
. 826 S.W.2d at 502.
. 917 S.W.2d at 738.
. Id.
. Edgewood III, 826 S.W.2d at 503.
. WOC II, 176 S.W.3d at 796.
. Id. at 794.
. 107 S.W.3d at 583.
. COLs 76-77.
. FOF 213.
. 826 S.W.2d at 503 (emphasis added).
. 917 S.W.2d at 737.
. 176 S.W.3d at 763, 796.
. FOFs 215-18. The court expressly referenced the analyses of Odden and Moak. It did not reference Baker, but relied on another expert, Dr. Catherine Clark, who like Baker arrived at a cost for an adequate education by adjusting for inflation the $3,500 figure given in Edgewood IV ’s footnote 10.
. 176 S.W.3d at 796.
. Id. at 797.
. Edgewood III, 826 S.W.2d at 500, 502.
. WOC II, 176 S.W.3d at 797.
. 107 S.W.3d at 579.
. As we understand the findings, the trial court made reference to 36 "focus” districts but only made adequacy and statewide tax findings as to 35 of them, excluding Hillsboro ISD. See FOFs 9, 210, 680-1187.
. The trial court also held on page 3 of its judgment that "local districts do not have meaningful discretion in the levy, assessment, and disbursement of property taxes; therefore, the Texas school finance system imposes an unconstitutional state property tax.” Again, the court did not break out any particular district or districts and separately hold that it or them had established a violation of article VIII, section 1-e.
. 176 S.W.3d at 796.
. Id. at 797.
. Id. at 795-96.
. See id. at 796 (reaffirming Court’s view that "State influence on district taxing and spending cannot be measured exactly but must be gauged along a spectrum of possibilities”).
. Judgment ¶ X.
. City of Lorena v. BMTP Holdings, L.P., 409 S.W.3d 634, 646 (Tex.2013); Bocquet v. Herring, 972 S.W.2d 19, 21 (Tex.1998).
. Barshop v. Medina Underground Water Conservation Dist., 925 S.W.2d 618, 637 (Tex.1996).
. WOC II, 116 S.W.3d at 799 (remanding fee issue where appellate court rules that a party awarded fees is "entitled to only a part of the relief granted by the district court” or “no relief”); City of Lorena, 409 S.W.3d at 646; Edwards Aquifer Auth. v. Chem. Lime, Ltd., 291 S.W.3d 392, 405 (Tex.2009); Barshop, 925 S.W.2d at 637-38.
. Judgment ¶ XI.
. See Tex. R. Civ. P. 308.
. City of San Antonio v. Singleton, 858 S.W.2d 411, 412 (Tex.1993).
. Tex. Const, art. VII, § 1.