See the full results of this study here.
Over the past 25 years, charter schools have offered an increasing number of families an alternative to their local district schools. The charter option has proven particularly popular in large cities, but charter-school growth is often constrained by state laws that limit the number of students the sector can serve. In the 2016 election, for example, voters in Massachusetts rejected a ballot question that would have allowed further expansion of charters in communities that had reached the state’s enrollment ceiling.
As a result, charter schools remain the smallest of the sectors that serve K–12 students. While district-operated schools still serve more than 80 percent of the U.S. school-age population, and private schools serve close to 10 percent, charters serve only about 6 percent (a share that is just slightly larger than that of the home-schooling sector). Yet the charter sector is the most rapidly growing segment of the education marketplace, and nationwide, the number of student names on charter-school waiting lists now exceeds one million, according to the National Alliance for Public Charter Schools.
Despite this rapid growth in the charter sector, little is known about the views of parents who are making use of these schools. Are charter-school parents more—or less—satisfied than parents in the district and private sectors with teacher quality, student discipline, and other characteristics of their children’s schools? Do they perceive more misbehavior there, or less? Are communications between parents and schools more or less extensive? And to what extent do parents’ perceptions of these issues vary within each sector?
To examine parental perceptions of charter, district, and private schools, we administered, in 2016, a survey to nationally representative samples of parents with children enrolled in each of these sectors. To our knowledge, this study, together with a companion investigation by Albert Cheng and Paul E. Peterson (see “How Satisfied are Parents with Their Children’s Schools?” features, Spring 2017), are the first to report results from nationally representative surveys of parents in these three sectors.
We find that charter parents are more satisfied with important aspects of their schools—such as teacher quality, school discipline, and character instruction—than are district-school parents, but they are less satisfied than private-school parents. Charter parents are also less likely to perceive serious problems in their children’s schools than are district-school parents. Charter parents report more extensive communications with their children’s schools than parents in the other two sectors, but they also express greater concern about a paucity of extracurricular activities.
To be clear, these findings speak only to how parents experience their children’s schools; they do not necessarily reflect the on-the-ground reality within each sector. Even so, parents’ opinions affect their choices among schools as well as, likely, the political pressure they may exert on educators and policymakers. What parents think of their children’s schools, therefore, has important implications for the future of the charter-school movement.
Charter schools have features in common with both private schools and those operated by the nation’s local school districts. Like district schools, charter schools receive most of their funding from public sources and are subject to state regulation. Also like district schools, they may not charge tuition and must admit all students who apply, unless they are oversubscribed, in which case they must hold an admissions lottery. But, like private schools, charter schools are operated by nongovernmental entities, and students attend only if their family selects the school.
As mentioned, there is no published comparison of parental perceptions of school life across the charter, district, and private sectors nationwide. However, the National Center for Education Statistics (NCES), in a series of surveys conducted between 1993 and 2012, reported on parental satisfaction with assigned public schools, public schools chosen by parents, private religious schools, and private nonreligious schools. These reports reveal that private-school parents are generally more satisfied than are those with children in the public sector. They do not, however, present information about parents with children in charter schools. In the companion study to ours, Cheng and Peterson report results for charter parents from the 2012 NCES survey, providing a portrait of differences across sectors that complements the one presented here.
Two existing studies do shed some light on the views of parents with students in charter schools. In 2010, Mathematica Policy Research compared the satisfaction of parents at 36 charter middle schools that held lotteries for admission with that of parents whose children had applied but had not won the lottery. Parents of lottery winners were 33 percentage points more likely than parents of lottery losers to rate their students’ schools as excellent and 10 percentage points more likely to agree that their children liked school. However, the schools, though broadly scattered throughout the United States, do not constitute a representative sample of parental opinion in the charter sector.
From 2001 to 2004, Jack Buckley and Mark Schneider tracked the satisfaction of parents whose children attended public and district schools in Washington, D.C. They found that, initially, charter-school parents rated their children’s schools more highly than their public-school counterparts did. However, this difference diminished over the course of the study, leading the authors to suggest that satisfaction with schools in the two sectors is unlikely to differ in the long run. Again, the study cannot be generalized to the nation as a whole.
Data and Methods
Our results are based on data from the nationally representative 2016 Education Next survey of public opinion, in which we oversampled parents with children in the three school sectors. The survey was conducted from May 6 to June 13, 2016, by the polling firm Knowledge Networks (KN), a GfK company. KN maintains a national probability panel of some 40,000 adults who agree to participate in a number of online surveys.
The Education Next survey was administered to a randomly selected subset of the KN panel, including an oversample of 1,571 respondents who were already identified as currently having school-age children (ages 5 through 18) living in their household. All parents were asked the same set of questions about their children’s schools, modified only to specify the relevant school sector (see methodology sidebar).
The background characteristics of parents in the three sectors differ along many dimensions, including race and ethnicity, education, income, and place of residence. Most notably, parents of charter-school students are more likely to be of minority background than are parents of either district- or private-school students (see Figure 1). Given the differences among families who use the three sectors, we report in an interactive graphic on the Education Next website all survey results both with and without adjustments for background characteristics. In the figures accompanying this essay, we report only the actual, raw percentages of parents in each sector giving a particular response. We do, however, adjust for background characteristics when testing whether differences in parents’ responses across the three sectors are statistically significant.
We find that in all three sectors an overwhelming majority of parents report that they are satisfied with their schools. Even so, the percentage of parents who say they are “very satisfied” with a given aspect of the school varies markedly across sectors (see Figure 2).
We inquired about parental satisfaction with respect to five key school characteristics where, based on previous studies comparing private- and public-school parents, we expected to see clear differences across those two sectors: teacher quality, discipline, expectations for achievement, safety, and instruction in character and values. We also asked about three other characteristics where we expected less differentiation: ethnic and racial diversity, facilities, and location. Consistent with prior research, we find that parents in the private sector are far more satisfied with most aspects of their children’s schools than are parents with children in district schools. The assessments of charter-school parents, meanwhile, typically fall between those of parents using schools in the district and private sectors.
Charter schools vs. district schools. Charter parents are considerably more satisfied with their schools than are district-school parents. The difference in the share of parents who are very satisfied for the five key characteristics is, on average, 13 percentage points. Smaller differences are observed for racial and ethnic diversity (9 percentage points), buildings and facilities (2 points), and location (-2 points); only the difference for racial and ethnic diversity is statistically significant.
Charter schools vs. private schools. We observe lower levels of satisfaction with charter schools than with private schools. For the five key characteristics, the private-school advantage is 12 percentage points, on average. In three cases (expectations for student achievement, safety, and instruction in character or values), these differences remain statistically significant after adjusting for the characteristics of the parents whom the two sectors serve. We also find a sizable but statistically insignificant difference in satisfaction with school location (13 percentage points). We do not find noteworthy differences between the two sectors with respect to satisfaction with the school building and facilities (4 points) or with ethnic and racial diversity (2 points).
Few parents acknowledge the existence of serious problems at their children’s schools. Some 47 percent of charter-school parents did perceive a lack of extracurricular activities to be a serious problem, but this was the highest share observed for any item across all three sectors. Nonetheless, the percentage of parents who indicate on a three-point scale that a given problem was either “serious” or “very serious” varies widely between sectors (see Figure 3).
District-school parents are generally more likely to say that various problems are either serious or very serious at their school than are private-school parents. The differences are substantial for four problems: students using drugs, destroying property, fighting, and missing class. Meanwhile, perceptions of charter-school parents again tend to fall somewhere in between.
Charter schools vs. district schools. Charter-school parents report fewer social problems than do district-school parents. On the four indicators of social disruption mentioned above, the district–charter difference in the percentage identifying a problem as serious or very serious is on average 8 percentage points. Each of these differences is statistically significant after adjusting for differences in the background characteristics of parents using district and charter schools.
Although charter parents perceive less social disruption in their schools, they express greater concern about the paucity of extracurricular activities. As compared to parents with children in district schools, charter-school parents are 14 percentage points more likely to perceive a lack of extracurricular activities as a serious problem.
Charter schools vs. private schools. A greater share of parents reported serious problems in the charter sector than in the private sector. However, after we controlled for background characteristics, these differences were statistically significant for only three problems: fighting, students with different abilities being placed in the same classroom, and a lack of extracurricular activities.
In sum, charter parents are more likely to identify serious problems with student behavior at their children’s schools than are private-school parents, but less likely to do so than district-school parents. Charters appear to provide fewer extracurricular activities than either private or district schools, perhaps because they are newer and often have less-lavish facilities and limited space for playgrounds and sports activities.
Most parents say they are in communication with staff at their children’s schools. Either a majority or a near majority within each sector say they have spoken to a school staff member at least once within the past year about each of the following: their child’s achievements and accomplishments; their child’s schoolwork or homework; their child’s behavioral problems; volunteering; the quality of teaching; and the behavior of other students at school. On some of these items there is little variation across sectors, but on others charter-school parents seem to be in closer contact with their school than parents in either the district or private sector (see Figure 4).
Charter schools vs. district schools. As compared to parents of children in district schools, charter parents are 15 percentage points more likely to say they have communicated with the school about volunteering, and 7 percentage points more likely to report having spoken to school officials about their child’s accomplishments.
Charter schools vs. private schools. School communications in the charter sector are also perceived by parents to be more extensive than those in the private sector. While private-school parents are as likely as charter parents to report having discussed volunteering, charter parents report that they have communicated with school officials more frequently than private-school parents about their child’s schoolwork or homework, the behavior of other students, their child’s behavioral problems, and the quality of teaching at the school. Only the difference with respect to communications about schoolwork or homework is statistically significant after adjusting for differences in background characteristics. Judging from parental perceptions, however, charter schools appear to have built a more extensive communication system with parents than schools in either the district or private sector.
Variation within Sectors
Charter schools are overseen by autonomous boards free of many state regulations, giving rise to a wide variety of approaches and emphases. District schools are operated by more than 14,000 locally elected boards, and this decentralization also creates the potential for wide disparities in school practice. As for the private sector, it offers parents an extraordinary variety of models from which to choose: Catholic schools, other Christian schools, Jewish schools, Muslim schools, Waldorf schools, Montessori schools, and many more.
But is the degree of variation in parents’ experiences with schools greater in one of these sectors than in the others? To find out, we calculated a metric that gauges the degree of variation in parental perceptions within each sector (see the methodology sidebar for a discussion of our approach). This topic has never been explored previously, making this aspect of our investigation especially groundbreaking.
Variation in parental satisfaction. On four items—school discipline, expectations for student achievement, school building and facilities, and the racial and ethnic diversity among students—we find no significant difference in the variation in satisfaction across sectors. However, we find that charter parents vary more in their satisfaction with the location of their school than do parents of students in district schools. Because district-school enrollment is usually determined by neighborhood while charter schools typically draw students from larger areas, this difference is to be expected.
Charter parents also vary more in their satisfaction with teacher quality than do district-school parents. This is also not surprising, given that teacher hiring in charter schools is often less tightly regulated than it is in the district sector. Some charter schools may use their flexibility to recruit outstanding teachers, while others fall well short of that mark.
When it comes to instruction in character and values, parental satisfaction within the private sector varies less than it does in either the charter or district sector. It may be that private schools consistently live up to their reputation for providing a moral as well as an academic education.
Satisfaction with school safety also varies less in the private-school sector than in the charter arena. It may be that private schools are able to create a more or less consistently safe environment, whereas charter schools, concentrated in areas that serve disadvantaged students, vary substantially in this regard.
Variation in the seriousness of problems. Our results concerning variation in how parents judge the seriousness of potential problems at their children’s schools tell a consistent story: the private sector is easily the most homogeneous of the three sectors. Meanwhile, parents’ perceptions of the seriousness of problems at district schools and charter schools are equally varied.
Variation in school–parent communications. When it comes to discussions of children’s achievements, the behavior of other students at the school, and even volunteering at school, we do not find significant differences across sectors in the degree to which parental perceptions vary. However, there is greater variation among charter parents in how frequently they report communicating about teacher quality than among parents in either private or district schools.
So which sector is the most homogeneous? The award goes to the private sector—at least insofar as parental perceptions are concerned. Private schools are very much alike with respect to how parents gauge the seriousness of potential problems at the school; they are also more homogeneous than the other two sectors in their instruction in character and values, as perceived by parents. And private schools are more homogeneous when it comes to discussions about children’s behavioral problems at school.
On most matters, charters and district schools are equally varied, but we do see greater variation within the charter sector in parents’ satisfaction with school location and teacher quality. We also see greater variation in the amount of parent–school communication about schoolwork and homework among charter parents.
So it is not altogether wrong to emphasize variation in the charter world, but on most of the school characteristics we find no significant difference between the variability in parental perceptions in the charter and district-school sectors. Perhaps the biggest surprise is the markedly greater homogeneity in the perceptions of private-school parents. If private schools operate in response to market demands, while district and charter schools operate in response to government expectations, then one might conclude that the marketplace expects certain fundamentals from all schools. Whatever their differences, all private schools appear to have learned that they must satisfy parental demands with respect to the basics.
What should one make of these results? As noted, our data speak only to parental perceptions and cannot be linked with certainty to the actual conditions in schools. Yet the differences we document are consistent with two classic ethnographic studies of schools in the private and district sectors.
Private schools. One of these studies, a 1993 book by Anthony Bryk and colleagues titled Catholic Schools and the Common Good, provides an unrivaled analysis of Catholic-school life. One cannot generalize to all private schools from an account of Catholic schools, of course. Nonetheless, roughly 50 percent of all private-school students today attend Catholic schools, and we have found a good deal of homogeneity in perceptions within the private sector. Further, many non-Catholic private schools have their own religious heritage that informs institutional practice. Even the elite members of the National Association of Independent Schools, which educate 10 to 15 percent of students in the private sector, usually have historical religious affiliations. We therefore treat Bryk’s account as a window on private schools more generally.
Central to that account is the idea that private schools create a strong sense of community. Teachers play an important role in students’ lives beyond the classroom. Catholic-school teachers also develop strong relationships with one another, both professionally and socially, and these bonds help facilitate harmonious decisionmaking.
The curriculum is largely uniform and based on “a long-standing Catholic tradition about what constitutes a proper humanistic education.” A common academic core does not preclude extensive extracurricular activities, however. These are numerous, and “provide more informal occasions for interactions between students and adults.”
Surprisingly, all this informality and collegiality does not necessarily lead to extensive parental involvement in the life of the school. Parents participate in fundraising and attend school events, but they tend to stay out of day-to-day operations and decisionmaking, putting their trust in the staff to provide a sound academic and moral education.
Trust is key to the Catholic enterprise, which is based on “a set of shared beliefs about what students should learn, about proper norms of instruction, and about how people should relate to one another.” Furthermore, the Catholic-school system is “highly decentralized,” and “considerable deference is accorded to the principal.”
In sum, Bryk’s account suggests that private schools create closed communities of like-minded people. Only those who seem to meet the requisite criteria—both educational and moral—are admitted. Families must pay tuition, unless their student is deemed worthy of a scholarship. Collective bargaining is virtually unknown within the private sector, so teachers can be dismissed if they do not perform to the school’s standards. Students are expected to focus on their studies, perform at a minimally acceptable standard, and conduct themselves according to the school’s code. Students remain at the school throughout their educational careers unless the family situation changes or they are expelled for misbehavior. The mission and ethos of the school are perpetuated by rituals and rites that celebrate the school’s traditions. Strong ties bind parents, students, faculty, and support staff together in service to a common set of values. But parents do not have extensive communication with the school; they trust the institution, which shares their values, to provide appropriate instruction. The role of the parent is to volunteer in support of fundraising and other activities.
District schools. An equally impressive account of district-school life is to be found in The Shopping Mall High School, written by sociologist Arthur Powell and his colleagues in 1985. Its focus on high schools, especially suburban ones, precludes easy extrapolation to middle and elementary schools. Yet these authors seem to have captured fundamental aspects of the district-school system as a whole.
The authors emphasize the grand variety of the student body. “Student inclusiveness is the reality most high schools must cope with: the students are different, and they are there,” they write, and schools accommodate this situation by providing “something for everyone.”
Students are often told that it’s up to them to make the most of school. This produces “a neutral environment where a do-your-own thing attitude prevails. High schools take few stands on what is educationally or morally important. Yet one thing they cannot be neutral about is diversity itself.… But tolerance … precludes schools’ celebrating more focused notions of education or of character. ‘Community’ has come to mean differences peacefully coexisting rather than people working together toward some serious end.”
At the shopping-mall high school, extracurricular activities play a central role. “There is nothing extra about the extracurriculum whether schools are rich or poor, public or private, large or small,” the authors write. “It supplements the rest of the curriculum and … is as integral to high schools as food service and celebrity appearances are integral to shopping malls.”
Parental involvement varies as much as everything else at the shopping-mall school. “Only a few parents understood that it was legitimate to be … active, and fewer still … would actually initiate an appointment with a teacher,” the authors observe. And for any involvement to happen, the initiative had to come from the parent.
If the Powell account remains accurate today, public schools are open communities made up of students from different backgrounds and with diverse values. Admission is entirely open, and the school operates within a framework created by state law and federal requirements. Teachers acquire tenure and seniority rights after a few years’ service, creating a challenge for administrators. Students have legal rights that limit administrators’ capacity to expel or suspend them from school. Rules and procedures provide the structure needed to reduce the likelihood of internal conflict within this diverse community. To ensure fairness in the open, diverse community, a complex bureaucratic structure becomes essential, but students come to resent the requirements that are imposed, which they see as inflexible and onerous. High performance by either teacher or pupil is seldom rewarded. Informal treaties bargained between teachers and students keep the peace. Parental engagement is determined by the parent not the school.
Admittedly, both the Bryk and Powell studies were conducted during the waning decades of the last century, and much has changed, especially the electronic revolution that has produced computers, cell phones, and Internet games. But we suspect that these changes have only accentuated features of the shopping-mall high school Powell captured a quarter century ago.
Our survey results are quite consistent with these ethnographic studies, and suggest that charter schools generally fall somewhere in between those in the district and private sectors. We find higher levels of satisfaction among parents of children attending charter schools than among those attending district schools, but lower levels of satisfaction than among those whose children attend private schools. We find that parents report less social disruption at charter schools than at district schools. Charter-school parents also report more extensive communication with school officials but lament the paucity of extracurricular activities at their children’s schools. As for the extent of variation in parents’ perceptions, it is the private sector that is the most homogeneous, while charter and district schools are for the most part similarly heterogeneous.
None of these results can necessarily be interpreted as identifying the actual characteristics of schools. Parents’ perceptions may be distorted by a lack of knowledge about what really goes on at a school or by an understandable tendency to view life at their own children’s schools through rose-tinted glasses. Nor can the results be interpreted as causal. We do not have experimental evidence as to the impact of attending schools in one sector rather than another. Parents have exercised choice in selecting a charter or private-sector school rather than a district school, making it impossible to say whether parental perceptions of the school are caused by actual school characteristics in each sector or some other factor.
This study, and the companion study by Cheng and Peterson, nonetheless provide the first descriptive accounts of differences in perceptions across the charter, district, and private sectors. Both studies find that charter-school parents’ assessments of their schools generally fall somewhere between those of parents in the district and private sectors. Charter parents’ greater satisfaction with their schools as compared to district parents, and their perceptions of fewer problems and more communication, have important implications for their school choices and, perhaps, their political behavior. If the number of charter schools continues to increase, the parents who use these schools may form a growing constituency in support of the charter-school option.
Samuel Barrows is a postdoctoral fellow at the Program on Education Policy and Governance (PEPG) at the Harvard Kennedy School. Paul E. Peterson, senior editor of Education Next, is professor of government at Harvard University and director of PEPG. Martin R. West, editor-in-chief of Education Next, is associate professor at the Harvard Graduate School of Education and deputy director of PEPG.
For a look at findings using data from a U.S. Department of Education survey, see “How Satisfied are Parents with Their Children’s Schools?.”
* * *
To obtain representative samples of parents with children in the three school sectors, we oversampled them during the administration of the larger 2016 Education Next survey of public opinion on education policies and practices in the United States. The survey was conducted from May 6 to June 13, 2016, by the polling firm Knowledge Networks (KN), a GfK company. KN maintains a national probability panel of some 40,000 adults who agree to participate in a number of online surveys. Members of this panel are recruited through address-based sampling from a frame of residential addresses covering approximately 97 percent of U.S. households. Internet access is provided to members of the panel who lack it.
The Education Next survey was administered to a randomly selected subset of the KN panel, including an oversample of 1,571 respondents who were already identified in the panel as currently having school-age children (ages 5 through 18) living in their household. After verifying the presence of school-age children in the household, respondents were asked how many of these children currently attend schools in a variety of sectors: district school, charter school, private school, and home school. In order to maximize the number of responses to questions concerning charter and private schools, respondents were classified as charter-school parents if they currently had a child in a charter school, even if they had other children who attended other school types; as private-school parents if they currently had a child in a private school but not in a charter school; or as district-school parents if they had a child in a district school but not in either the charter or private sector. This question served as a screen to allow oversampling of charter-school and private-school parents until we reached numbers large enough to estimate perceptions within each sector with a small margin of error. This process yielded a sample with 774 district-school parents, 428 private-school parents, and 317 charter-school parents. KN provided sample weights to adjust the parent oversample to the demographic profile of the U.S. population with school-age children.
All parents were asked the same set of survey questions, with only a slight variation in language to specify the sector of the school about which they were being asked. To simplify the presentation, we report results in a binary fashion, even though scales with three to five response options were constructed for each item. When testing the statistical significance of differences between sectors, however, we take into account the full distribution of responses across all options. Specifically, we fit proportional odds models that include indicators for the respondent’s child being at a private or district school, while charter-school parents serve as a baseline. These models also include controls for the respondent’s education, income, race, homeowner status, region,
and whether the respondent lives in an urban area. The significance of the coefficients on the private- and district-school indicators allows us to test whether there is a statistically significant difference between charter-school parents and parents from either of the other sectors, after adjusting for differences in the observable background characteristics of the parents they serve.
In order to explore how variation in parents’ responses differs across sectors, we calculate a measure called Leik’s D for each question and sector. We then test for a statistically significant difference in the variation of responses across sectors by estimating bootstrapped confidence intervals for the differences in Leik’s D between sectors.