Methodology

The methodology defines all variables, explains why we chose them, and describes how we calculated and coded them. It also describes how we calculated the final state rankings. The variables are organized into four main categories that influence the degree of education liberty in each state: (1) Education Choice, (2) Regulation, (3) Transparency, and (4) Spending. Within some main categories, we identified second-level indicators, which are the sub-categories of these main categories of education liberty.

1. Education Choice

States with more education choice have more educational liberty. “Education Choice” has five sub-categories: (a) Private School Choice, (b) Private School Choice Program Design, (c) Charter Schools, (d) Homeschooling, and (e) Public School Choice.

(a) Private School Choice. Following are components of “private school choice” variable.

  • Education savings accounts (ESAs). This variable is a binary measure of whether a state offers ESAs. These include state-funded ESAs and tax-credit-funded ESAs. The code is “1” if the state offers an ESA and “0” if the state does not. ESAs are the future of education liberty, and they are a key measure of whether states are giving parents the freedom to customize their children’s education, not just to choose their school. These data come from pages 9 to 12 of EdChoice’s 2022 ABCs of School Choice.
  • Proportion of eligible students. This variable is a measure of the proportion of all K–12 students in the state who are eligible for a private-school-choice program. Private-school-choice programs include school vouchers, ESAs, and tax-credit scholarships. We did not include individual tax credits and deductions. This variable is calculated as a proportion between 0 and 1 where the numerator is the number of students eligible for a school choice program, and the denominator is the total number of public, private, and homeschooled K–12 students in the state. Total public school and private school enrollment data by state are from the National Center for Education Statistics. These data come from pages 15 to 150 of EdChoice’s 2022 ABCs of School Choice.
  • Proportion of students who participate. This variable a measure of the proportion of all K–12 students in the state who are participating in a private-school-choice program. Private-school-choice programs include school vouchers, ESAs, and tax-credit scholarships. We did not include individual tax credits and deductions. This variable is calculated as a proportion between 0 and 1 where the numerator is the number of students participating in a school choice program, and the denominator is the total number of public, private, and homeschooled K–12 students in the state. These data come from pages 15 to 150 of EdChoice’s 2022 ABCs of School Choice.

(b) Private-School-Choice Program Design. This variable indicates whether states have a private-school-choice program that requires participating private schools to receive accreditation from a state, regional, or national accrediting body and/or to conduct an approved standardized test. These regulations restrict the freedom of private schools to create innovative education options that families desire. The variable is coded as “3” if a state has a private-school-choice program(s) and none of the programs have accreditation or testing requirements; the variable is coded as “2” if the state has a private-school-choice program(s) and the program(s) have either an accreditation or a testing requirement; the variable is coded as “1” if the state has a private-school-choice program(s) and the program(s) have both an accreditation and a testing requirement; and the variable is coded “0” if the state does not have any private-school-choice programs. These data come from pages 155 to 160 and 165 to 168 of EdChoice’s 2022 ABCs of School Choice.

(c) Charter Schools. Following are the indicators that comprise the “charter schools” variable.

  • Proportion of public schools that are charter schools. This variable is calculated as a proportion between 0 and 1 where the numerator is the number of charter schools in the state and the denominator is the total number of public schools in the state. Charter schools sever the tie between ZIP code and school, providing more options and freedom for parents to choose the most appropriate education for their child. These data are from Table 216.90 of the 2021 NCES Digest of Education Statistics, which measured the 2019–2020 school year.
  • Charter school law rating. This variable is a rating of the quality of a state’s charter school law from the Center for Education Reform’s (CER’s) 2021 National Charter School Law Rankings & Scorecard. As the CER states, “[C]harter school success depends on the policy environments in which charter schools operate,” and “overregulation and underfunding force charters to behave as district schools by another name.” This report card rewards states for increasing “operational autonomy” for charter schools. CER rated the 46 states with charter laws as A, B, C, D, or F. We coded these 4, 3, 2, 1, and 0, respectively. If a state did not have a charter law according to CER, we coded it as –
  • Educational Freedom Institute (EFI) charter ecosystem ranking. This variable measures the quality of the charter school ecosystem in each state from the Educational Freedom Institute’s 2022 Charter School Ecosystem Rankings. This report is a continuous variable of the raw score calculated by the EFI authors that considers access to charter schools, growth in charter schools, and growth in achievement outcomes, among other factors. A higher raw score translates to a lower-quality charter ecosystem. The scores range from 48 to 206. We then multiplied all scores by –1 so that the higher scores reflect higher-quality charter ecosystems. We coded states that do not have a charter environment to – Scores are from Table 10 on page 33 of the EFI report.

(d) Homeschooling. Following are the indicators that comprise the “homeschooling” variable.

  • Percentage of homeschooled students. This variable is calculated as a proportion between 0 and 1 where the numerator is the adjusted number of homeschooled K–12 students, and the denominator is the total number of public, private, and homeschooled K–12 students in the state.
  • Home School Legal Defense Association (HSLDA) rating (ranking of state homeschool laws). This variable measures the quality of state homeschool laws by the Home School Legal Defense Association. The HSLDA categorized states as “No regulation,” “Low regulation,” “Moderate regulation,” and “High regulation” which we coded as 3, 2, 1, 0, respectively. Less regulation in the homeschooling process means parents have more freedom to customize their child’s education. More regulations, such as mandatory testing and curriculum, limit education freedom. These data were collected on February 22, 2022.

(e) Public School Choice. Following are indicators that comprise the “public school choice” variable.

  • Open-enrollment policy. The 2018 report titled 50-State Comparison: Open Enrollment Policies from the Education Commission of the States documents whether states have mandatory intra-district enrollment and/or mandatory inter-district enrollment. Open enrollment policies allow families to choose public schools beyond the one assigned to their ZIP code. While not as impactful as private-school-choice programs, allowing students to attend other public schools injects more accountability into schools and give families more options. If a state requires districts to participate in both types of open enrollment, we coded the variable as 2. If a state requires one type, we coded it as 1, and if a state does not require open enrollment, we coded the variable as 0.
  • Students per district. This variable measures the number of K–12 public school students per public school district in each state. Larger, more consolidated, districts reduce parental control over how and where their child is educated. This variable is a continuous variable where the numerator is the total number of public school students in the 2018–2019 school year from the NCES Digest Table 214.30, and the denominator is the total number of public K–12 districts in the state.
  • Square miles per district. This variable measures the number of square miles of land mass per K–12 public school district in each state. Larger, more consolidated districts reduce parental control over how and where their child is educated. This variable is a continuous variable where the numerator is the total land area from the 2010 Census, and the denominator is the total number of public K–12 districts in the state.

2. Regulation

The burden of regulation on education providers limits the innovative options they can provide. An education environment of low regulation would maximize options and allow parents to choose the best education for their children. Within “Regulation,” there are three sub-categories and variables: (a) Barriers to Teaching, (b) Chief Diversity Officers, and (c) Common Core Testing Requirement.

(a) Barriers to Teaching. Following are indicators that comprise the “barriers to teaching” variable.

  • Proportion of alternative-certified teachers. This variable measures the proportion of new teachers who are certified through alternative pathways. Traditional teacher licensing is a major barrier to entering the teaching profession. States that allow many paths to certification have more options for good teachers to get jobs. The numerator of this proportion is the number of people who completed an alternative-teaching certification program, and the denominator is the number of people who completed any teaching certification, traditional or alternative. Data are from the Department of Education for the 2018–2019 academic year as part of the 2020 Title II Report.
  • Full reciprocity for out-of-state teachers. This variable is a measure of whether a state has a statute that grants full reciprocity to teachers who complete their certification in other states. This binary variable is coded as 1 if a state has a statute that grants full reciprocity for teachers with an out-of-state certification, and it is coded as 0 if a state does not have such a statute. These data are from the Education Commission of the States report titled, 50-State Comparison: Teacher License Reciprocity.

(b) Chief Diversity Officers. This variable measures the proportion of school districts in the state with 15,000 students or more that have a chief diversity officer (CDO). As Jay Greene and James Paul point out in their study, CDOs “may be best understood as political activists who articulate and enforce an ideological orthodoxy within school districts.”1 CDOs likely limit academic freedom in the classroom. The numerator is the number of districts that employ a CDO by the total number of districts for each state. If a state does not have any districts with 15,000 or more students, we coded it as 0. Data were collected from Greene’s and Paul’s Heritage Foundation Backgrounder, “Equity Elementary: ‘Diversity, Equity, and Inclusion’ Staff in Public Schools.”

(c) Common Core Testing Requirement. This binary variable captures whether a state administers either the Partnership for Assessment of Readiness for College and Careers (PARCC) or the Smarter Balanced Common Core exam. Standardized testing lowers standards and homogeneity in education options. Families in states that do not administer a Common Core exam have more curricular options and, therefore, more education liberty. We coded this variable as 1 if the state did not administer a Common Core exam in spring 2019 and coded it as 0 if the state did administer a Common Core exam. These data are from the 2018–2019 school year and were collected by Education Week.

3. Transparency

Transparency of the education system gives parents information to make the best decisions about their children’s education. “Transparency” has three subcategories: (a) Strong Critical Race Theory Law, (b) Parental Empowerment, and (c) Accountability.

(a) Strong Critical Race Theory Law. This variable indicates whether a state has a regulation reaffirming that compelled speech is unconstitutional and/or stating that violations of the Civil Rights Act of 1964 in K–12 schools are illegal. The variable is coded as 2 if the state has a regulation(s) with both components, 1 if the state has only one component, and 0 if the state has neither component. Heritage researchers used the following sources: Education Week, Heritage Foundation Critical Race Theory: Legislation Tracker, and the Brookings Institution.

(b) Parental Empowerment. The “parental empowerment” variable includes the following indicators.

  • Parental organizations per pupil. This variable is a count of the number of grassroots parent organizations in each state as defined by the Parents Defending Education IndoctriNation Map. Parents care deeply about the education of their children and the better organized they are, the more influence they have on making policy changes that increase education liberty. This count was divided by the number of public school students in the state to arrive at the number of parental organizations per pupil. Data were collected by Heritage researchers on January 24, 2022.
  • Indoctrination incidents per pupil. This variable is a count of the number of incidents of indoctrination in each state as defined by the Parents Defending Education IndoctriNation Map. Examples of incidents include, “Santa Barbara Unified School District teacher exposes highly racialized and political curriculum that is being withheld from parents.” This count was divided by the number of public school students in the state to arrive at the number of incidents per pupil. This number was then converted to a negative number by multiplying it by –1 as more incidents of indoctrination reflect less education freedom. Data were collected by Heritage researchers on January 24, 2022.

(c) Accountability. The “accountability” variable is comprised of the following indicators.

  • Parent access to curricula and materials. This variable indicates whether a state has a policy that that clearly states that parents must be allowed access to their child’s curricula and learning materials. Increased transparency of curricula and lesson plans gives parents information to make the best decisions about the education of their children. This variable is coded as 1 if the state has a policy that specifically states that a parent must have access to curricula and materials and 0 if the state has no such policy. Sources: Education Liberty Alliance and Tennessee’s Textbook Transparency Act. Thanks to Matt Beienburg, director of education policy at the Goldwater Institute, for confirming these data.
  • Public comment in school board meetings. This variable indicates whether a state has a requirement to allow public comment at school board meetings. While the freedom of exit encourages the greatest accountability to parents, the ability to communicate with one’s community at local school board meetings is a valuable indicator of transparency. This variable is coded as 1 if the state has a policy that specifically states that parents must have access to curricula and materials and 0 if the state has no such policy. Source: Reporters Committee for Freedom of the Press.
  • Freedom of Information Act (FOIA) law rating. This variable is a rating of the state FOIA laws, which captures the ability of the public to access information about their public school system. Rating criteria include response type, response time, costs, and contact ease. Easier access to information about their public schools gives parents knowledge about their education options. States were rated as “Good,” “Average,” or “Poor,” which Heritage researchers coded as 2, 1, and 0, respectively. Source: Logikcull.
  • School board election cycle. This variable indicates whether states hold school board elections on cycle with the general election. Teacher’s unions often lobby for off-cycle elections so they can ensure a larger relative turnout from their members. This makes it more likely that union-friendly school board members are elected. Researchers collected data on dates of school board elections in each state and the District of Columbia. This variable is coded as 1 if school board elections are held in November of even years. For most states, the date is decided at the state level and is the same for all districts. For some states, election dates are decided at the local level and thus might vary within a state. If any district in a state holds school board elections off-cycle, we coded it as 0. States where any districts appoint, rather than elect, their school board are coded as 0. We collected the data in February 2022.

4. Spending

Education liberty includes the freedom of taxpayers to keep more of their money. Private-school-choice programs also tend to spend less money for the same or better outcomes. “Spending” has four subcategories and variables: (a) Per-Pupil Spending, (b) Return on Investment, (c) Teacher-to-Non-Teacher Ratio, and (d) Unfunded Pension Liability.

(a) Per-Pupil Spending. The “per-pupil-spending” variable is comprised of the following indicators.

  • Nominal per-pupil spending. This variable measures the average K–12 spending per pupil in each state. These data represent the total “current expenditures, capital expenditures, and interest on school debt per pupil” from Table 236.75 of the NCES Digest of Education Statistics for the 2018–2019 school year.
  • Cost of living adjusted per-pupil spending. This variable takes the nominal per-pupil spending variable and adjusts it for differences in cost of living by state. To calculate the cost of living adjusted per-pupil spending, Heritage researchers multiplied the nominal per-pupil spending variable by the regional price parities by state for 2018 from the Bureau of Economic Analysis.

(b) Return on Investment (ROI). This variable measures ROI in terms of academic achievement for the investment in terms of spending per pupil. The numerator is the average score on the National Assessment of Education Progress (NAEP) across fourth grade and eighth grade and reading and math. The denominator is the nominal per-pupil spending variable. Data on NAEP scores are from spring 2019 and were collected from the NCES Digest of Education Statistics Tables 222.50, 222.60, 221.40, and 221.60.

(c) Teacher-to-Non-Teacher Ratio. Much of the K–12 public school spending has resulted in more school administrators, not investments in teachers and resources to improve education quality. This variable was calculated by dividing the number of teachers by the number of non-teaching staff in each state in fall 2019. Data are from the NCES Digest of Education Statistics Table 213.20.

(d) Unfunded Pension Liabilities. This variable measures the unfunded pension liability as a percentage of state gross domestic product. In many states, unfunded pension liabilities are a burden on taxpayers and economic growth. They also represent a concentration of teacher union power as well as deficient political leaders who make decisions regarding public education. Data are from page 11 of the downloadable data files for the State of Pensions 2021 Report by the Equable Institute.

State Ranking Calculations

We coded all variables so that a higher value corresponds with more education liberty. First, we standardized each variable by calculating a z-score. If a sub-category had multiple variables, we averaged z-scores within that sub-category. For example, we calculated the average of z-scores of the four variables in the Private School Choice sub-category; we then calculated the z-score of that average.

Next, we averaged the z-scores across sub-categories and top-level variables. For example, we calculated the average of the z-scores from the Private School Choice, Private School Choice Program Design, Charter Schools, Homeschooling, and Public School Choice sub-categories to arrive at a single number for the main “School Choice” category. We then calculated a z-score of that average. Once we calculated a z-score for each of the four main categories, we then took the average of those four z-scores and calculated the z-score of that average to arrive at a single z-score for each state. The state with the highest final z-score is ranked first in the Education Freedom Report Card.


[1]  Jay P. Greene and James Paul, “Equity Elementary: ‘Diversity, Equity, and Inclusion’ Staff in Public Schools,” Heritage Foundation Backgrounder No. 3666, October 19, 2021, https://www.heritage.org/education/report/equity-elementary-diversity-equity-and-inclusion-staff-public-schools.