School Choice: Public Opinion in Five Battleground States
The Manhattan Institute commissioned Rasmussen Reports to include nine questions related to school choice and charter schools in their late August–early September polling of likely voters in Pennsylvania, Ohio, Wisconsin, Michigan, and North Carolina. Among these states, 46%–52% of the respondents said that they believe that giving parents the right to choose their children’s school raises the overall quality of K–12 education for students; 18%–20% believe that it lowers educational quality. Black respondents were more likely to believe that school choice raises educational quality.
Between 66% and 70% of all respondents supported the concept of publicly funded K–12 school choice. This support was higher among black respondents in all states but North Carolina.
Between 51% and 62% of all respondents supported state funding of charter schools as an alternative to traditional local district-managed public schools. This support was higher for black respondents in all states and ranged from 58% to 67%.
Across the five states, actual enrollment in charter schools accounts for 7.3% of all students in charter and district-managed public schools. In Michigan, 9.6% of public school students attend charters; in Wisconsin, 4.9% do. Across the five states, 17.3% of black public school students attend charters, over 3.8 times the rate of white students. In four of the surveyed states, black students are the group most likely to enroll in charters. In North Carolina, the number of white students enrolled in charters is slightly higher than that of black students. In Michigan, 27% of black public school students attend charters; in Pennsylvania, 23% of black students in public schools attend charters.
National research indicates that urban charter schools achieve significantly higher levels of growth in math and reading than their district school peers. This is particularly true for lower-income children of color and those with special needs. Charters in nonurban areas, and those operating as virtual schools, do not show this advantage.
Has the public turned against charter schools and school choice in recent years, after the tremendous growth of charters across the country in the first 15 years of the century? Under President Obama, the Democratic Party was committed to the growth of charter schools as a means of expanding educational opportunities to traditionally underserved communities. This year’s Democratic Party platform, however, calls for increased scrutiny of charter schools:
The Republican Party has not released a detailed platform for the coming election, but Republican members of Congress and the Trump administration have supported school choice in various forms, particularly during discussions of COVID relief aid packages. While the administration has given a high priority to forms of school choice that include public support of private and religious schools, it has not proposed any significant expansion of federal regulation of existing or new charter schools.
In August, the Manhattan Institute contracted with Rasmussen Reports to survey public opinion in five states about school choice policies generally, as well as charter schools specifically. Working with MI, Rasmussen crafted nine questions that were added to the organization’s regular polling of likely voters for presidential and senatorial races in Pennsylvania, Ohio, North Carolina, Michigan, and Wisconsin. This paper will discuss five key questions from those surveys and place the results in the context of what we know about charter schools and choice in these states and nationwide. A companion report, being released by Rasmussen Reports, presents detailed results of all nine education-related questions from the surveys conducted between August 25 and September 10, 2020.
Public Attitudes Toward School Choice and Charter Schools
Among the likely voters surveyed, there was little support for the notion that school choice is harmful to education in grades K–12. Across the five states, 46%–52% of respondents said that they believed that giving parents the right to choose their children’s school raises the overall quality of K–12 education for students (see the questions and responses in Figure 1). Conversely, 18%–20% believe that it lowers educational quality. Among black respondents, 43%–57% believed that parental choice improves educational quality (Ohio was at the lower band of that range). In all states but Ohio, black respondents were less likely to believe that parental choice lowers educational quality than was the general population.
The survey asked individuals to express their support for the concept of publicly funded K–12 school choice. Of all respondents, 66%–70% either strongly or somewhat supported this concept. At 65%–77%, support for publicly funded school choice was higher among blacks.
Between 51% and 62% of respondents either strongly support or somewhat support state funding of charter schools as an alternative to traditional district-managed public schools; 29%–39% of respondents were somewhat opposed or strongly opposed to such funding. In each state, support for state funding of charter schools was higher among black respondents; 58%–67% favored it, while 21%–34% were opposed.
Few respondents reported that they believed that their state gave parents too much choice when deciding where their children would attend school; 8%–11% of all respondents felt this way, with 4%–11% of blacks feeling this way. Among blacks, Michigan was the outlier, with 11% reporting that they thought parents were given too much choice. Between 29% and 39% of all respondents reported that they believed that their state gave parents too little school choice, with North Carolina being the outlier, at 39%. Between 34% and 45% of black respondents reported that they felt parents were given too little choice, with Pennsylvania schools being the outlier, at 45%.
As another measure of interest in school choice, we asked individuals who reported having school-age children at home what type of school they would choose if they had the choice. In each state, the plurality of respondents said that they would choose a district-managed public school. There was a strong interest in private schools, with 14%–22% selecting this option; 10%–15% reported that they would choose a religious school, and 10%–19% said that they would choose a publicly funded charter school.
There are two important caveats for these responses. Because the survey asked this question only of individuals with schoolage children at home, the sample was much smaller and contains a higher margin of error. Also, the options for private and religious schools did not describe them as publicly funded, so these results cannot be viewed as a measure of support for public funding of private and religious schools. In all states but North Carolina, black respondents were less likely to say that they would choose a district-managed public school, but here the sample was quite small and difficult to interpret.
Profiles of School Choice in the Five Survey States
Overall, 7.4 million students in the states we surveyed attend public schools (Figure 2). The vast majority, 92.7%, attend district- managed schools. Overall, 7.3% of public school students attended charter schools, ranging from 4.9% in Wisconsin to 9.6% in Michigan. Considering only the charter sector across the five states, black students are the most likely to be enrolled; 17.3% of them attend, more than 3.8 times the rate of white students. The only state of the five in which black students were not the most likely to use charters was North Carolina, where they were edged out by whites, 7.4% to 6.7%. In Michigan, 27% of black students in public schools attend charters; in Pennsylvania, 23% do. White students are the least likely (4.5%) to attend charters. Among Hispanics, 8.1% of students are enrolled in charters, as are 6.7% of Asians.
The Effectiveness of Charter Schools
It is difficult to make blanket statements on the relative performance of charter and district-run public schools. Student performance varies in both sectors, and the characteristics of students attending individual schools differ as well. Many district- managed schools in higher-income suburbs are models of effectiveness, serving as ready pipelines to elite and selective universities. It would be unfair to compare either charter or district-run schools serving lower-income communities with these schools. At the same time, even though charters tend to admit students through random lotteries when they are oversubscribed, concerns remain about the alleged cream-skimming of students most likely to succeed. The best research studies control for the demographic differences in student populations. These studies are complicated to conduct; but fortunately, some point us to an understanding of the relative performance of charter schools, compared with public schools managed by districts.
In 2015, the Center for Research on Education Options (CREDO) at Stanford University published a study of charter school performance in 41 urban areas. It concluded that “urban charter schools in the aggregate provide significantly higher levels of annual growth in both math and reading compared with their TPS [traditional public school] peers.” The researchers interpreted their statistical findings to mean that “charter students [are] receiving the equivalent of roughly 40 days of additional learning per year in math and 28 additional days of learning per year in reading.” This does not mean that charter schools have longer annual schedules, though some do. Rather, the test-score advantage associated with charter school attendance was the equivalent of what the average student would learn in those numbers of days. The 2015 CREDO study also found that “learning gains for charter school students are larger by significant amounts for black, Hispanic, low-income, and special education students in both math and reading.” The achievement advantages for students who were in two of these subgroups (black and special education) were quite large, amounting to “months of additional learning per year.”
This study of urban areas followed one that CREDO published in 2009, one that examined charter school performance in 16 states across the country. This earlier study had much more sobering results for charter supporters and is often quoted by charter opponents. It found: “Charter school students on average see a decrease in their academic growth in reading of .01 standard deviations compared with their traditional school peers. In math, their learning lags by .03 standard deviations on average. While the magnitude of these effects is small, they are both statistically significant.” The report further found significant variability across states in terms of the charter school effect.
In January of this year, Boston University’s Marcus Winters employed a randomization procedure in the unified application system used for charters as well as district schools in Newark, New Jersey, in order to replicate a randomized control study on the impact of charters. His study, published by the Manhattan Institute (where he is a senior fellow), found a strong impact of charter school attendance. It found that:
Winters also examined the contention that the growth of charter schools harms the performance of surrounding district-run schools. In June 2020, his Manhattan Institute study examined the impact of charter school growth on district schools across the country. His assessment:
Winters’s report does not assess causality; rather, it concludes that “the general pattern of test-score outcomes over this period is simply not consistent with the claim that charter school exposure for a meaningful period of time produces declines in the performance of traditional public schools.”
All the CREDO studies found variability in charter school performance that tended to mute the differences between charters and district-managed schools. In other words, within the charter sector, as with the traditional district sector, there are high and low performers. This is clear in the performance of charter schools within the five states in which we asked Rasmussen to survey public opinion about charters. We did not select these states based on the performance of their charter schools. Rather, in a national election year, these critical states are likely to draw the attention of both political parties as well as the national media. The performance of charters in each of these states is mixed, with performance varying by urban vs. rural location; and by type of school—virtual vs. brick-and-mortar. Performance varies within those categories as well.
Research on Charters in the Five Survey States
In 2019, CREDO published a report on charter school performance in Pennsylvania. It showed that a typical charter school student in a year achieves similar gains in reading and weaker gains in math, compared with what the student would have produced in a traditional public school (TPS). CREDO estimates that a typical charter school student in Pennsylvania experiences the equivalent of 30 fewer days of learning in math, compared with a student in a traditional school setting.
However, a deeper analysis of the results reveals that Pennsylvania’s charter school landscape is defined by notable bright spots and failures. Urban charter school students exhibit growth in reading equivalent to 35 more days of learning, compared with urban TPS counterparts. Black charter school students experienced a statistically significant increase in learning gains (24 more days) in reading, compared with black TPS students.
CREDO also analyzed the difference in learning experiences in brick-and-mortar versus online charters. Brick-and-mortar charter schools had a statistically significant positive impact on students’ reading growth, while online charters depressed students’ learning in reading and math. Students in online charters had learning gaps equivalent to 106 fewer days in reading and 118 fewer days in math, compared with TPS students. Because the enrollment in online charters was roughly a quarter of the charter sector in the state at the time the study was published, the online sector is likely weighing down the results of Pennsylvania charter schools.
Online charter schools are often the biggest blemish academically in a state’s charter school system, and Ohio’s charter sector follows this trend. A 2016 report published by the Thomas B. Fordham Institute details that while Ohio’s virtual charter schools were growing faster than all other public education sectors in terms of enrollment, their students performed significantly worse on state standardized tests, even after controlling for prior achievement. The effects were particularly profound regarding math achievement. Students in online math courses were far less likely to take advanced math courses and far more likely to enroll in remedial math classes. State data also show that while 24% of all face-to-face math courses are advanced, only 6% of online math courses fall under the same category.
When compared with district schools, brick-and-mortar charters in Ohio have a positive effect on students in grades 4–8 and a mixed effect on high school students.
One feature of the charter school sector is its ability and willingness to close low-performing schools. In 2005, the Ohio General Assembly passed a law requiring all charter schools that continually fell below a year-to-year performance threshold to close. A 2016 study (by professors at the University of Oklahoma and Ohio State) estimates the causal effects of school closures on student achievement. Two years after low-performing charter schools face mandatory closure (three years after a school is informed that it will be required to close), students from closed schools scored 0.2–0.3 standard deviations higher in reading and math than students from schools who just score above the performance mark mandating closure. These results suggest that accountability measures should be combined with school choice to guarantee that children have access to high-quality schools.
A 2019 study by professors at Oakland University used longitudinal data from 2003 to 2012 to determine whether Michigan charter schools outperformed traditional public schools on standardized testing. Unlike those in most other states, charters in the Wolverine State are largely managed by for-profit education management organizations. The researchers analyzed reading and math pass rates on the Michigan Educational Assessment Program (MEAP) in 4th, 7th, and 11th grades., and pass rates on the Michigan Merit Examination (MME) among 11th graders. In all grades and subjects, charter schools early in the study period lagged far behind traditional public schools. The initial disparity in pass rates, depending on grade and subject, was generally 6–14 percentage points. However, by 2012, charters had largely caught up to traditional public schools in all grades, nearly eliminating achievement differences. Improvements in the charter sector during 2003–12 could be due to the closure of ineffective charters, or the ability of charters to be more creative and adaptive in their approaches. However, the researchers conclude that their results infer that charters in Michigan “do no harm” to student outcomes, rather than provide substantive improvements.
An earlier study by CREDO cast a different light on Michigan charters. Researchers used data from the 2005–06 to the 2010–11 school years to determine whether students who attend charter schools make significantly more learning gains than they would in traditional public schools. Using a Virtual Control Record method to compare learning gains made by students in charter schools with learning gains made by a synthesized “twin” in a traditional public school setting, they found that Michigan charter schools have a significantly positive effect on reading and math gains among black students, Hispanic students, and students who have repeated grades. While charter schools in Michigan are largely concentrated in urban settings, urban as well as rural charter schools have a significant, positive effect on learning gains in both math and reading. The charter sector’s impact on a student’s achievement grows as a student stays in a charter school—students in their fourth and fifth years of attendance see larger gains in math and reading, compared with students in their first year.
A 2012 study published by the University of Arkansas compared four-year achievement gains between Milwaukee independent charter school students and students in the Milwaukee Public Schools (MPS) district. The researchers took a sample of students who were enrolled in one of Milwaukee’s independent charters in 2006, and matched them to an MPS sample, controlling for variables such as race, free-lunch status, and prior achievement. Previous research showed little to no differences in achievement gains in math and reading between charter and district schools after one and two years, but a clear advantage emerged for charters in the third year. The study, however, concluded that while four-year achievement gains are positive for students in independent charter schools, the results do not maintain statistical significance between those in charters and MPS students.
One promising takeaway for charters is that charter schools outperformed their public school counterparts among “stayers”— students who remained in either the charter or district sector throughout the period studied. Stayers in charter schools, on average, significantly outpaced stayers in the MPS in reading and math gains, by approximately three-tenths and four-tenths of a standard deviation, respectively. This indicates that long-term exposure to the charter sector in Milwaukee has a large effect on student achievement.
Additionally, that study revealed that students in conversion charters—schools that were once private but converted to charter schools—consistently outpaced MPS students in reading. While these advantages might be due to the selectivity of the private school before conversion, the gains realized by conversion charters but not realized by non-conversion charters may also be attributed to the schools’ prior experience in operation and not having to undergo a “start-up” phase like traditional charter schools.
A 2019 study from the Wisconsin Institute for Law & Liberty found that certain types of charter schools in Wisconsin outperformed their local traditional district schools on state examinations of math and English Language Arts (ELA). In Milwaukee, “non-instrumentality” charters (under district control, but with nonunionized teachers hired by the school, not the district) outperformed MPS by 13 points in math and 12 points in ELA; independent charters (no district control) outperformed the district schools by 8 points in math. Further, charters authorized by the University of Wisconsin–Milwaukee outperformed MPS by 8 points in ELA and 10 points in math.
In 2015, researchers at Duke University published a study on the evolution of the charter school sector in North Carolina. The paper delved into the timeline of how charter schools have shifted demographically and academically. The researchers note that charter schools in the Tarheel State initially contained a disproportionately high percentage of black students; but over time, minority shares of charter enrollment declined and the percentage of white students surpassed the percentage of white students in traditional public schools. In addition, during 1998–2014, the percentage of students in predominantly white charters nearly doubled, from 24.2% to 47.1%.
The researchers also note that student achievement has gradually improved in North Carolina charter schools over time. From 1999 to 2003, charter schools lagged in learning gains for both reading and math. More recently, they began to surpass traditional public schools, which remained stagnant. By 2012, charter schools had exceeded the average achievement gains in reading and math in traditional public schools. Moreover, the variation in average student performance across charter schools also declined over time, now matching the variation observed in traditional public schools. Despite the initial positive results, the researchers explain that the positive trend should be interpreted with caution. While they attribute this growth to a departure of low-performing schools and an entrance of better-performing models, they also infer that growth occurred because of the enrollment of more academically motivated students in charter schools. New entrants into charter schools, on average, had 20% fewer absences than those who remained in traditional public schools. Moreover, when analyzing student achievement using student fixed effects, they conclude that charter schools in North Carolina overall are no better at raising achievement levels than traditional public schools.