My Kids, Your Kids, Our Kids: What Parents and the Public Want From Schools

by Jon Valant & Daniel A. Newark - 2017

Background/Context: School choice reforms could strengthen parents’ influence on school behaviors, since schools must appeal to parents in order to operate. If parents’ desires for schools differ from the broader public’s desires for schools, then schools might pursue different goals and activities in systems emphasizing school choice. One popular hypothesis is that school-choosing parents, more than the public, want schools to prioritize their own students’ private interests over more collective social, economic, and political interests.

Purpose/Objective/Research Question/Focus of Study: We compare parents’ desires for their own children’s schools with the U.S. public’s desires for public schools. We make these comparisons with respect to the abstract goals that schools pursue, as well as schools’ more tangible behaviors.

Population/Participants/Subjects: We administered an online survey to nationally representative samples of parents and adults. We administered a second online survey to a national sample of adults.

Intervention/Program/Practice: The article consists of two studies. Study 1 compares parents’ and the public’s beliefs about which abstract goals schools should prioritize. Respondents were randomly assigned to consider either schools in their community, schools around the country, or, if they had children, their own children’s schools. They chose from goals that prioritized their students’ professional achievement (“Private Success”), the economy’s needs (“Shared Economic Health”), and more collective social and political needs (“Democratic Character”). Study 2 compares parents’ and the public’s beliefs about how schools should actually behave. Respondents were randomly assigned to consider either schools in their community, schools around the country, or their own children’s schools. We asked about the basic structure and content of the school day, how schools should teach, and how to evaluate school performance.

Research Design: The studies consist of randomized experiments and related statistical analysis.

Findings/Results: We find remarkably little difference between parents’ desires for their children’s schools and the public’s desires for public schools. This is true both for the abstract goals that schools pursue and for schools’ more tangible behaviors.

Conclusions/Recommendations: Our findings suggest that the hypothesis that parents want schools to focus on their students’ private success at the expense of more collective goals is oversimplified. It may be, for example, that parents want their children to be well rounded in ways that also serve more collective social, political, and economic interests. We observe divisions in Americans’ views of the goals that schools should pursue, but these divisions are more connected to their political affiliation than parent status (with Republicans more focused than Democrats on Private Success).

Few U.S. education policy initiatives have been as consequential, or controversial, as the school choice reforms of recent decades. In 20132014, approximately 2.5 million U.S. students attended a charter school, 2.6 million attended a magnet school, and 300,000 students enrolled in private schools through school voucher, scholarship tax credit, or education savings account programs (American Federation for Children, 2016; U.S. Department of Education, 2015). These reforms have tended to empower parents. They expand parents discretion over which schools their children attend and potentially make schools more responsive to parents desires, since schools funding depends on their ability to attract school-choosing families. The shift toward empowering parents may disempower the broader public, since U.S. public schools have historically been governed by school districts led by publicly elected boards.

At the core of the debate about school choice reforms is a governance question: Should schools be primarily accountable to parents or the public? To the extent that we hold schools accountable to parents (e.g., through school choice programs), we should expect them to attend to parents interests and demands. To the extent that we hold schools accountable to the public (e.g., through public elections and referenda), we should expect them to attend to the publics interests and demands. This makes it important to understand what parents and the public want from schools. One hypothesis is that these groups interests differ sharply, with parents wanting schools to pursue their own childrens private benefits at the expense of public benefits valued by society more broadly (e.g., Belfield & Levin, 2005; Friedman, 1955; Krashinsky, 1986; Labaree, 1997; Levin, 1991). In particular, school choosers may prioritize their childrens college and career prospects, whereas the broader public prioritizes schools role in our shared social, political, and economic well-being.

Labarees (1997) articulation of how school choosers and the public may value private and public aims differently has been particularly influential. He asserted that, historically, American schools have pursued three goals, which he called social mobility, social efficiency, and democratic equality. As consumers, school choosers might be drawn primarily to the pursuit of social mobility, which emphasizes preparing students to succeed relative to others in society (e.g., by obtaining desirable employment and social positions). The broader public, however, is more inclined to value social efficiency (in its roles as taxpayers) and democratic equality (in its roles as citizens). Social efficiency emphasizes preparing workers to fill the roles necessary for a strong economy, whereas democratic equality emphasizes preparing children, regardless of their backgrounds, for healthy participation in democratic society. In Labarees view, as U.S. education policy has increasingly empowered parental school choosers, schools have felt growing pressure to make education a commodity valuable to the particular students attending. He argued that the U.S. has shifted from treating education as a good intended to serve the country as a whole to one intended to serve the private interests of students pursuing desirable professional and social positions.

Though these ideas are well reasoned, they have lacked empirical verification. They also do not present a clear explanation of how purported differences in parents and the publics desires for the abstract goals that schools pursue might give rise to tangibly different school systems depending on how schools are governed. There are, however, related literatures that may offer insight into whether school choosers and the publics desires differ as hypothesized and how the posited differences in these groups abstract desires might tangibly affect schools.

There are two principal ways that parents, thinking about their own childrens schools, and the public, thinking about schools around its communities and country, could diverge in what they want from schools. First, what parents want for themselves (and their own children) could differ from what they want for society more broadly. In other words, the same person might have different preferences depending on whether she is operating on behalf of herself and her own children or others and others children. For example, the not-in-my-backyard (NIMBY) syndrome describes a tendency for people to want others, but not themselves, to incur the costs of having essential facilities like prisons or landfills near their homes (Dear, 1992). Perhaps parents would like students in general to make the sacrifices necessary to develop a society of ethical citizens and productive workers but would like their own children to focus on acquiring privileged social positions. If this is so, then parents might favor schools that pursue public goals when voting (the case of democratic governance), and schools that pursue private goals when deciding where their children will enroll (the case of school choice). Second, school-choosing parents and the publics desires could diverge not because different systems of governance evoke different desires within parents, but rather because different governing systems empower distinct portions of the population with differences between them. Govern through an educational marketplace, and only parents of school-age children have a voice. Govern through democratic institutions, and the entire voting public weighs in on what schools should do. Even if each individual wants the same for herself and society more generally, school-choosing parents are a discrete subset of the American population with potentially distinct social and political views.

Despite these two potential ways in which desires for schools could differ, it remains an open question whether school-choosing parents desires for their own childrens schools and the publics desires for schools actually diverge as hypothesized. For one, parents desires for their childrens relative prosperity exist alongside many other desires that parents have for their children (e.g., see LeVine, 1974). In fact, parents goals for their children may be more compatible with serving shared social, political, and economic interests than is often believed. Suizzo (2007), reporting on surveys of 343 parents from cities in the southwestern U.S., found that parents of every ethnic group prioritized their childrens self-direction, benevolence, and prosocial dispositions over power and achievement. Perhaps most parents want their children to become the types of people the public wants for a healthy society. Additionally, just as parents desires may be more compatible with public goals than is typically assumed, the publics desires may also be more consumer oriented. The emergence in recent decades of the new public management ideology reflects growing tendencies to treat the public sector more like the private sector, citizens more like customers and clients, and administration more like management (Hood, 1995; Schedler & Felix, 2000). In part, this has meant increased focus on individual, rather than collective, concerns. Perhaps todays U.S. public wants schools to attend to students private interests more than the current literature suggests.

Although there is little direct evidence of how parents and the public value schools pursuit of private and public goals, related literatures examine which school attributes people value when choosing schools and how they feel about issues in education. Surveys of school-choosing parents typically find that parents prioritize academic quality above other criteria (Harris & Larsen, 2014; Stein, Goldring, & Cravens, 2011), although physical safety can be parents dominant concern in areas with particularly high crime (Stewart & Wolf, 2014). Studies of families school-choosing behaviors and ultimate school choices echo the importance of academic quality, while also highlighting the importance of characteristics like proximity to home (Harris & Larsen, 2014; Hastings, Kane, & Staiger, 2009), neighborhood type (Bell, 2009), extracurricular offerings (Harris & Larsen, 2014), student satisfaction (Jacob & Lefgren, 2007), and peer demographics (Hastings et al., 2009; Schneider & Buckley, 2002).

Surveys of the U.S. public also highlight the importance of academic quality. Rothstein and Jacobsen (2006) reported that American adults, school board members, state legislators, and school superintendents valued a variety of school goals but rated Basic Academic Skills in Core Subjects as the most important among them. Moreover, most Americans believe that core subjects like math and English are the most valuable subjects they studied in school (Jones, 2013). Recent surveys conducted by Phi Delta Kappan and Gallup (Bushaw & Calderon, 2014) and Education Next (Henderson & Peterson, 2014) offer intriguing comparisons between the beliefs of U.S. parents and the broader public. These surveys suggest that parents and the public think similarly about U.S. schools on a wide variety of subjects, but they ask parents about schools generally and not specifically about their own childrens schools. At most, these findings hint that parents may not be a subgroup of the population with unique desires. The findings cannot speak to whether parents feel different about their own childrens schools than they, or the broader public, feel about schools across their communities and country.

In addition to a dearth of empirical evidence on whether parents and the public have different desires for schools pursuit of private and public goals, little attention has been paid to how these differences would substantively affect students. Students experiences are shaped by how time is allocated across subjects and activities and how those subjects are taught. While educational theory has posited that parents and the broader public may want schools to pursue different high-level priorities, it has largely ignored whether these groups have different desires for tangible school behaviors or how tangible behaviors and abstract, high-level priorities relate. For instance, would we expect private-minded school choosers to demand longer, more challenging school days to obtain an advantage relative to other children, or would we expect the public to make these demands to obtain a better return on its investment in others children?

Furthermore, there is a real possibility, supported by organizational theory research, that desires for the concrete behaviors of schools may be relatively uniform. Neo-institutional theory describes the tendency of organizations in a shared field to adopt similar, established structures and practices largely as a way of achieving legitimacy in the eyes of their constituents (Powell & DiMaggio, 1991). Organizations tend to adhere to the same scripts and rationalized myths in order to appear valid and credible, thereby improving their chances of survival (Meyer & Rowan, 1977). This could be especially true in a field like education, where organizational goals are relatively ambiguous, the technologies for achieving those goals are relatively uncertain, and evaluation of whether the goals have been achieved can be difficult (DiMaggio & Powell, 1983; March, 1978). Metz (1989) observed this phenomenon in the context of American high schools. After studying teachers work lives in public high schools, Metz described schools adherence to a common script (p. 81). She noted that schools followed similar schedules, settled on common class sizes, and used similar textbooks, scope and sequences, and instructional techniques. Metz interprets conformity with this common script as a desire on the part of the schools to convey that these were Real Schools with Real Teachers and Real Students doing whatever it is that real schools do. If school choosers and the public both want schools to act like schools have acted historicallyor like the schools they remember from their youththen any differences in their ideological desires might not correspond to differences in how they want schools to actually behave.

This article compares school choosers and the publics desires for the ideological goals that schools pursue, schools more tangible behaviors, and the performance metrics used to evaluate school quality. We introduce data from surveys and survey experiments, some drawn from nationally representative samples, to a long-standing scholarly conversation light on empirical testing. In doing so, we find strikingly little evidence that parents desires for their own childrens schools differ from the broader publics desires for its schools. Preferences for abstract goals are virtually identical, with 39.5% of American school choosers and 39.5% of the broader public preferring that schools prioritize their students private success over more public goals related to shared economic health and democratic character. Similarities also appear in these groups desires for schools tangible behaviors, as evidenced by their preferences for school day length, allocation of hours across subjects and activities, instructional approaches, homework demands, and performance metrics.

OVERVIEW OF STUDIES

We conducted two studies to examine whether school choosers desires for schools differ markedly from the broader publics desires for schools. In particular, we addressed the following three research questions:

"

Study 1:

o

RQ #1. Do parents desires for their own childrens schools differ from the U.S. publics desires for schools with respect to the abstract, ideological goals that schools pursue?

o

RQ #2. Do parents, themselves, have different desires for their own childrens schools and for schools in their communities and country with respect to the abstract, ideological goals that schools pursue?

"

Study 2:

o

RQ #3. Do parents desires for their own childrens schools differ from the U.S. publics desires for schools with respect to schools tangible behaviors?

In both Study 1 and Study 2, we assigned respondents to share their desires for their own childrens schools (school choosers group) or schools more generally (broader public group). Within each study, we also divided the broader public group in two, randomly assigning respondents to consider schools in either their local communities or around the U.S. This division reflects governing authority being spread across the local, state, and federal levels and the possibility that people think differently about matters of local education politics (e.g., electing school board members) and national education politics (e.g., electing presidents). We report both disaggregated results and results for an aggregated broader public group.¹

We asked respondents in both studies specifically about high schools. We did so for a few reasons. First, elementary schools and high schools differ in what they teach and how they teach it, so asking specifically about one of the two reduced ambiguity about which questions to ask and how to interpret respondents answers. Second, although it is reasonable to ask parents of elementary school students what they will look for in a high school, it is nonsensical to ask parents of high school students what they will look for in an elementary school. Asking about high schools made it plausible for both elementary and high school parents to take the perspective of school choosers. Finally, although some parents of young children will see their beliefs evolve over time, their current beliefs could affect their future high school choices. We would expect this to happen if, for example, parents carefully choose an elementary school and then choose middle and high schools based on which schools their childrens peers will attend. Focusing on high schools provided the clearest path to a coherent, meaningful survey.

DATA

Data come from a series of online surveys. For Study 1, we included items on two omnibus surveys administered by YouGov, a firm specializing in academic survey research and online political polling. We asked the same questions on these two surveys in order to increase our sample size, particularly of parents with children 18 years old or younger. Each survey had 1,000 respondents, but in one survey, we asked questions only of the 264 respondents with at least one child 18 or younger, yielding a total sample size of 1,264 adults. YouGov used a matching and weighting strategy to construct analytical samples approximately representative of the U.S. adult population. YouGov has a panel of approximately 1.2 million U.S. residents who fill out online surveys in exchange for modest participation incentives. For our study, YouGov used demographic and political variables to match members of its panel to a random sample drawn from the 2010 American Community Survey (ACS), invited matched panelists to complete our survey, and used respondent weights to further improve the alignment between our survey respondents and the ACS sample. YouGov has performed well in predicting national election outcomes using these methods (Rivers, 2012; Silver, 2012) and now partners with The New York Times and CBS News for their election polling (Cohn, 2014). Still, there is no foolproof way to construct a nationally representative sample, and YouGovs success in doing so hinges on its ability to construct a sample that mirrors the U.S. adult population using only known variables for its opt-in respondent pool.²

In Study 2, we surveyed a sample of respondents from amazon.coms Mechanical Turk (MTurk). MTurk is an online marketplace where a Requester can post a task (e.g., a survey or an audio transcription request) to be completed for pay by a Worker. Adults from around the world sign up to complete tasks as contract work. MTurk has become a popular source for academic survey respondents because of its affordability, unobtrusiveness, and the desirable properties of its respondent pool (Berinsky, Huber, & Lenz, 2012; Buhrmester, Kwang, & Gosling, 2011; Paolacci, Chandler, & Ipeirotis, 2010). We restricted our sample to American respondents, paying $1 to each of the 165 Workers who took our survey. Although we had fewer respondents in Study 2 than Study 1, we had more time per respondent in Study 2 (median response time = 7.5 minutes), enabling us to ask more and richer questions.

Table 1 contains the observable characteristics of our Study 1 (YouGov) and Study 2 (MTurk) samples, as well as the U.S. adult population. In general, the Study 1 sample is similar to the U.S. adult population on all available variables: gender, race and ethnicity, age, educational attainment, and political identification. The Study 2 sample, which was not matched and weighted to be nationally representative, is diverse, but more likely to be Asian, younger, more formally educated, and more inclined toward the Democratic Party.

Table 1. Demographic Data for the U.S. Adult Population, YouGov Sample, and MTurk Sample
	U.S. population (ages 18+)	YouGov sample (weighted)	YouGov sample (unweighted)	MTurk sample (unweighted)
Gender
Female	51.3%	53.2%	53.1%	49.1%
Male	48.7%	46.8%	46.9%	50.9%
Race/Ethnicity
Asian, non-Hispanic	5.2%	3.0%	1.5%	11.4%
Black, non-Hispanic	11.9%	13.1%	11.6%	4.8%
Hispanic	15.0%	12.8%	10.5%	7.8%
White, non-Hispanic	65.7%	66.4%	71.3%	73.5%
Other race	2.2%	4.8%	5.1%	2.4%
Age
Mean age (years)	46.9	46.0	47.7	32.5
Educational Attainment
No college	42.1%	41.0%	40.7%	13.4%
Some college / 2-year degree	29.0%	30.9%	33.1%	33.5%
Completed 4-year degree	28.9%	28.0%	26.3%	53.0%
Political Identification
Leans Democrat	47%	42.7%	44.7%	49.1%
Leans Republican	41%	30.3%	34.2%	14.5%
Independent (no lean)	--	22.2%	17.6%	32.1%
Not reported	--	4.8%	3.4%	4.2%
# of observations	N/A	1,264	1,264	165
Notes. For the U.S. population data, respondents were asked to identify themselves as Democrat, Republican, or Independent. Those who identified as Independent were then asked whether they lean to the Democratic Party or Republican Party and have been classified according to their reported lean. For the YouGov and MTurk data, respondents were classified as Leans Democrat if they reported a political affiliation of Strong Democrat, Not very strong Democrat, or Lean Democrat, and classified as Leans Republican if they reported Strong Republican, Not very strong Republican, or Lean Republican. We include respondent weights for all analyses using YouGov data. Data on gender, race/ethnicity, age, and educational attainment come from the U.S. Census Bureau (2013a, 2013b). Data on political affiliation come from Gallup (Jones, 2014).

The MTurk sampling process does not allow for claims of national representativeness, but this sample is meaningful for our purposes nonetheless. All the evidence presented in Study 2 is the product of randomized experiments, and experimental results from MTurk studies (based on treatment-control comparisons) have generalized well to populations of interest even when the sample and populations are observably different (Berinsky et al., 2012). Also of note, when we administered our primary survey item from Study 1 to a separate MTurk sample of parents with children 18 or younger, we obtained similar results to those found in the YouGov sample.

STUDY 1: PARENT AND PUBLIC DESIRES FOR ABSTRACT SCHOOL GOALS

Study 1 features a single substantive item that allows us to (a) compare parents desires for their childrens schools with the broader publics desires for schools around its communities and country (RQ #1) and (b) compare parents desires for their childrens schools with parents desires for schools around their communities and country (RQ #2) with respect to the abstract, ideological goals that schools pursue. Based on the existing literature, we provided respondents with three goals that high schools could pursue and asked them to select the most important. We asked some respondents about their own childrens schools, others about schools in their communities, and others about schools around the United States.

In deciding which respondents would answer for which sets of schools, we aimed to minimize survey response bias, maximize the precision of our estimates, and generate the data necessary to answer our research questions. Study 1 respondents come from the two YouGov samples. In the first sample (n = 1,000), we assigned all respondents with children 18 or younger to consider their own childrens schools, and we randomly assigned the remaining respondents to consider either schools in their local community or schools around the United States. This way, we could compare what parents desire from their own childrens schools (relevant if parents are empowered through school choice) with what the public desires from schools around its community and country (relevant if the public is empowered through democratic governance). In the first sample, we refrained from asking any parents about schools around their community or country in order to secure a large enough sample of parents reporting on their own childrens schools to make precise comparisons for RQ #1.

A drawback to assigning all parents with children 18 or under to the school chooser group is that it does not generate the data needed to examine whether parents themselves have different desires for their own childrens schools compared with schools around their communities and country (RQ #2). It also keeps parents with young children from being represented in our broader public group. For these reasons, we administered the same item, almost simultaneously, to a second sample of YouGov respondents with children 18 or younger. This second sample (n = 264) augmented our parent sample and generated the data needed for RQ #2. By randomly assigning parents to one of the three conditions, we could test whether the same individualsparents with school-age childrenhave different desires for their own childrens schools and schools around their communities and country.

Parents and the public look to schools to serve many purposes. In choosing which goals to present, we settled on three goals that are prominent in this literature. We call them Private Success, Shared Economic Health, and Democratic Character. These goals are analogous to Labarees social mobility, social efficiency, and democratic equality, respectively (1997), with our adaptations aimed at translating Labarees goals into a coherent survey item. The Private Success goal prioritizes enrolled students opportunities for personal and professional success and social status. Private Success is the only goal focused on generating value for the student, with no mention of society more broadly. The Shared Economic Health goal prioritizes the economys needs even when doing so is not the most direct way to serve the private interests of the students enrolled. While Private Success focuses on helping particular students get ahead (succeed relative to others in society), Shared Economic Health focuses on building a strong economy for all. The Democratic Character goal prioritizes collective political and social interests by building an ethical, compassionate, engaged citizenry. Though parents might desire their childrens character development for their own childrens sake, we classify Democratic Character as a public goal because it almost certainly benefits society more broadly. In the item shown in Figure 1, the first response corresponds to Private Success, the second to Shared Economic Health, and the third to Democratic Character.

Figure 1. Primary survey item from Study 1

Which of the following do you think is most important for [your childs (or childrens) high school / high schools in your local community / high schools around the U.S.] to do well?  (Choose one.)

"

Prepare students for the careers that are best for them personally (e.g., high-paying, prestigious, or enjoyable jobs), even if other careers would benefit the economy more.

"

Prepare students for the careers that are best for the economy, even if other careers would be better for the students themselves.

"

Prepare students to be engaged, ethical members of their communities who treat others with kindness and respect.

Our analysis begins by comparing respondents desires for these three goals (Private Success, Shared Economic Health, and Democratic Character) across all three sets of schools (ones own childrens schools, schools in ones community, and schools around the country). We call this our full, 3x3 analysis. To statistically assess whether responses vary by which set of schools the respondents were asked about, we ran chi-square tests of independence that incorporate the respondent weights provided by YouGov and make the necessary statistical corrections described by Rao and Scott (1984). We provide full, 3x3 results for both the full sample (for RQ #1) and for the subset of our sample with a child 18 or younger (for RQ #2).

In addition to providing full, 3x3 results, we collapse our findings into 2x2 results that examine the hypothesized private-versus-public interests distinction between school choosers and the broader public. Here, we combined responses for those answering about schools in their communities and schools across the United States into a single broader public group. We compare this group with the parents who constitute our school chooser group. We also combined responses for the Shared Economic Health and Democratic Character goals into a single public goals response, which we compare with the Private Success response. These 2x2 results present a simplified look at whether parents thinking about their own childrens schools prioritize private and public goals differently than people thinking about schools more broadly.

Our final empirical test for Study 1 examines whether observable characteristics of our respondents predict their proclivity to select the Private Success goal. Here, we are interested in which subgroups of our respondents, if any, are the strongest proponents of schools prioritizing their students private interests. Our data enable us to examine whether significant relationships exist between respondents probability of choosing Private Success, and their gender, race and ethnicity, educational attainment, age, political orientation, or parent status (regardless of whether the respondents were asked about their childrens, communities, or countrys schools). We report the results as odds-ratios from a basic logistic regression model. Like all our other analyses for Study 1, this model incorporates the respondent weights provided by YouGov.

RESULTS FOR IDEOLOGICAL SCHOOL GOALS

The results from Study 1 show strikingly little difference between parents desires for which goals their childrens schools pursue and the broader publics desires for which goals schools around its communities and country pursue (RQ #1). We also see no stark differences between what parents desire from their own childrens schools and what they desire from schools more generally (RQ #2). When we examine which respondent characteristics predict choosing Private Success, the only significant predictor is identification with the Republican Party.

The Full Sample panel of Table 2 shows the full, 3x3 results for RQ #1. There is no statistically significant relationship between respondents selection of the Private Success, Shared Economic Health, or Democratic Character goal and whether they were asked about their childrens, communities, or countrys schools. The Private Success goal was selected by 40% of those asked about their own childrens schools, 41% of those asked about schools in their communities, and 38% of those asked about schools around the country. The Shared Economic Health goal was selected by 16% of those asked about their own childrens schools, 10% of those asked about schools in their communities, and 15% of those asked about schools around the country. The Democratic Character goal was selected by 45% of those asked about their own childrens schools, 49% of those asked about schools in their communities, and 47% of those asked about schools around the country.

Table 2. Desires for Abstract Goals by Whether Asked About Own Children's Schools, Schools Around the Community, or Schools Around the Country (Study 1)
	Own child(ren)'s schools	Schools around community	Schools around country	Test statistic	p value
A. Full Sample				F(3.91,4932)=0.87	p=0.479
Desires "Private Success"	39.5%	41.2%	37.8%
Desires "Shared Economic Health"	15.9%	10.0%	14.8%
Desires "Democratic Character"	44.6%	48.9%	47.4%
# of observations	450	415	399
B. Parents-Only Subsample				F(3.83,2349)=0.73	p=0.563
Desires "Private Success"	39.5%	43.7%	35.3%
Desires "Shared Economic Health"	15.9%	19.1%	27.0%
Desires "Democratic Character"	44.6%	37.3%	37.6%
# of observations	450	87	77
Notes. In each sample, only parents were asked about their own children's schools. We report Pearson design-based F statistics (and the accompanying p values) to account for survey weights in accordance with Rao and Scott (1984). All samples include the respondent weights provided by YouGov.

The similarity between what parents desire, ideologically, from their own childrens schools and what the public desires from its schools is even clearer in the pooled, 2x2 responses (Table 3). These results combine those asked about schools around their communities and schools around the United States into a single broader public group, and combine the Shared Economic Health and Democratic Character goals into a single public goals response. Of the 450 parents asked about their own childrens schools, 39.5% selected private goals as the most important. Of the 814 respondents asked about schools around their communities and country, again, 39.5% selected private goals. The remaining 60.5% of each group selected one of the public goals. This comparison, of course, shows no statistically significant relationship. Our sample provides reasonably precise estimates for these comparisons, with a standard error of 4.0 percentage points on this comparison between school choosers and the broader public. Multivariate models controlling for observed demographic and political characteristics are virtually identical to these simple comparisons in both their estimated coefficients and standard errors.

Table 3.  Desires for Abstract Goals by Whether Asked About Own Children's Schools or Schools Around the Community/Country (Study 1)
	Own child(ren)'s schools	Schools around community/ country (aggregated)	Difference (own children community/ country)	Test statistic	p value
A. Full Sample
Desires private goals	39.5%	39.5%	0.1	F(1,1263)=0.00	p=0.990
Desires public goals	60.5%	60.5%	(4.0)
# of observations	450	814
B. Parents-Only Subsample
Desires private goals	39.5%	39.7%	-0.2	F(1, 613)=0.00	p=0.974
Desires public goals	60.5%	60.3%	(6.4)
# of observations	450	164
Notes. Standard errors appear in parentheses. In each sample, only parents were asked about their own children's schools. Respondents classified as "Desire private goals" selected the Private Success response, while respondents classified as "Desire public goals" selected either the Shared Economic Health or Democratic Character response. All samples include the respondent weights provided by YouGov. ***p < .01. **p < .05. *p < .10.

Our findings for RQ #2examining whether parents themselves have different desires for their own childrens schools and for schools around their communities and countryalso show no significant differences across groups. The Parents-Only Subsample panel of Table 2 shows the full, 3x3 results. Forty percent (40%) of parents selected Private Success for their own childrens schools, 44% selected it for their communities schools, and 35% selected it for the countrys schools. Again, we see little evidence of differences across groups, though the smaller, disaggregated samples make for less precise estimates than we had for RQ #1. The pooled, 2x2 Parents-Only Subsample results reported in Table 3 offer more statistically precise estimates that are almost identical across groups. Of parents randomly assigned to answer about their own childrens schools, 39.5% preferred that schools pursue private goals, and 60.5% preferred public goals. Of parents randomly assigned to answer about schools in their communities or country, 39.7% preferred that schools pursue private goals, and 60.3% preferred public goals. These differences are far from statistically significant and reasonably precisely estimated (with a standard error of 6.4 percentage points).

With these survey responses being so similar across parents and the public, one might wonder whether respondents were dishonest in reporting their preferences or were not focused on the particular set of schools about which we asked. We took numerous precautions and performed a variety of checks to ensure this was not the case. Aware that respondents might find it socially undesirable to admit that they want something different from their own childrens schools from what they want from other childrens schools, we employed a between-subjects study design in which we asked each respondent about only one type of school. As such, no individual respondent had to reveal different preferences for her own and other children. Administering the survey online rather than in person further limited social desirability bias and interviewer effects (Duffy, Smith, Terhanian, & Bremer, 2005; Kreuter, Presser, & Tourangeau, 2008). To assess whether YouGov respondents internalized which set of schools they had been asked about, we conducted manipulation checks with 124 online survey respondents from amazon.coms MTurk. First, we administered the Study 1 task to these MTurk respondents, each of whom had at least one child 18 or younger. Then we asked respondents to complete a cognitively challenging estimation task before asking them to recall whether they had originally been asked about their own childrens schools, schools around their communities, or schools around the country. Nearly three fourths (73.3%) correctly identified their original assignment. These precautions and tests make us confident that respondents understood what they were being asked and answered honestly.³

Our analysis of which subgroups prefer Private Success bolsters confidence that respondents answered this item in accordance with their beliefs, along with offering additional insight into the dynamics of Americans desires for schools. Table 4 reports results from the logistic regression model predicting whether respondents selected Private Success by their observable demographic and political characteristics. The one significant predictor of whether respondents chose the Private Success goal was political orientation. Republicans were significantly more likely than Democrats to want schools to prioritize Private Success (odds ratio = 1.592; p = 0.011).⁴ This difference seems consistent with ideological differences between the Democratic and Republican Parties. Lakoff (2010) and Freeman (1986) argued that the Republican Partys worldview holds that societys collective interests are best served through the pursuit of individuals own self-interests. Although our Private Success response specified that the careers best for these students might not be best for the economy as a whole, Republican respondents might have been skeptical that policy makers could adequately identify and serve the broader public interest. The presence of a RepublicanDemocratic difference, consistent with the difference in party ideologies, suggests that our items captured variation in respondents views where such variation existed.⁵

Table 4.  Odds Ratios for Whether Selected Private Success Goal by Personal Characteristics (Study 1)
	Selected Private Success
Female	1.191
	(0.193)
Black	1.049
	(0.269)
Hispanic	1.071
	(0.282)
Asian	1.699
	(1.129)
Other race	1.067
	(0.393)
Did not attend college	0.772
	(0.154)
Attended college, no 4-year degree	1.128
	(0.227)
Age (years)	1.002
	(0.006)
Republican	1.592**
	(0.291)
Independent (politically)	1.154
	(0.263)
Parent w/ child(ren) 18 or under	1.003
	(0.190)
Constant	0.479
	(0.192)
# of observations	1264
Notes. Standard errors appear in parentheses. Omitted reference groups are males, Whites, respondents who completed a 4-year college degree, Democrats, and respondents with no children 18 or under. Samples include YouGov weights. ***p < .01. ** p <.05. *p < .10.

Results from Study 1 indicate that parents desires for which abstract goals their childrens schools pursue are similar to the broader publics desires for schools (RQ #1) and to parents desires for schools around their communities and country (RQ #2). This suggests that transferring authority from the public to school choosers might not, in fact, empower actors who demand a much sharper focus on students private success than the public would like. Likewise, it suggests that transferring authority from school choosers to the public might not, in fact, empower actors who demand a much sharper focus on public goals than school choosers would like. Still, this ideological agreement does not rule out the possibility that school choosers and the publics desires for schools differ in consequential ways. In particular, school choosers and the public might share similar ideological goals but disagree about how schools should actually behave or which metrics to use in evaluating school quality. In Study 2, we examine whether school choosers and the public express different desires about the more tangible aspects of schooling that define many students educational experiences: what schools do, how they do it, and how people assess whether they are doing it well.

STUDY 2: PARENT AND PUBLIC DESIRES FOR TANGIBLE SCHOOL BEHAVIORS

Our second study examines whether school-choosing parents and the broader U.S. public desire different tangible, everyday behaviors from schools and whether they value concrete performance indicators differently (RQ #3). While scholars have speculated whether parents and the broader public have different abstract goals for schools, there has been little discussion of whether or how preferences for what schools actually do might diverge. Furthermore, predicting preferences for tangible school behaviors based on peoples perspectives and ideological desires is difficult. For example, should we expect parents to desire more rigorous educational programs to give their children a relative advantage, or should we expect the public to demand greater rigor in order to get a strong return on their investment in public education? Should we expect differences in parents and the publics views on how much time students should spend in art class, how much homework they should get, or how social studies should be taught?

Here, we examine preferences for the basic structure of the school day, focusing on the characteristics of schooling that education researchers have identified as most consequential. Specifically, we examine these groups desires for what schools do (the basic structure and content of the school day), how they do it (approaches to teaching that content), and how people assess school quality (common metrics for high school performance). Incorporating school performance metrics is important because these metrics are sometimes more observable than actual school behaviors, and schools may feel pressure to perform well on the metrics that people value most. With more time per respondent than we had in Study 1, we could ask more detailed questions, and we took this opportunity to focus on the most substantively important content of schooling.

For Study 2, we randomly assigned respondents to consider their own childrens high schools (the school chooser condition), high schools in their communities, or high schools around the country. We allowed nonparents to be randomized into the school chooser condition, asking them first to imagine they had a child in high school.⁶ Of the 54 respondents assigned to the school chooser condition, 21 reported having at least one child (19 had a child age 18 or younger), and there were no statistically significant differences in any of our outcome variables between the responses of parents and nonparents assigned to the school chooser condition. Other respondents were asked to focus on either high schools in their local community (n = 56) or high schools around the United States (n = 55).

Regardless of a participants treatment assignment, her first task was to write a few sentences from the point of view of her assignment. The purpose of this task was to reinforce the respondents treatment condition and facilitate perspective-taking. The opening text was: In this study you will be asked what you want from [your childs high school/high schools in your local community/high schools in the United States]. Please take a moment now to consider what role you would like [your childs high school/high schools in your local community/U.S. high schools] to play. In the space below, please list the things that are most important for [your childs high school/high schools in your local community/high schools in the United States] to accomplish. As the survey progressed, we reminded respondents of the schools they should be thinking about in the text of each item.

Our first set of items asked how students should spend their time in school. We began with an open-ended item asking respondents how much time high school students should spend in school per week. To simplify the task and reduce the probability that respondent mistakes would generate outlier data, we volunteered that students currently spend about 35 hours per week in school. We then asked respondents to allocate hours across school subjects and activities. Respondents were asked, How many hours per week should [your child/high schools in your local community/U.S. high schools] spend on each of these subjects or activities? We identified nine subjects and activities based on a review of the curricula of dozens of districts and charter schools.⁷ Respondents saw a list of these nine subjects and activities, each with a box to indicate ones preferred number of hours. The default value in each box was zero. Even if a respondent reported that the school week should be shorter or longer than 35 hours, we asked her to allocate exactly 35 hours across nine subjects and activities. Although the survey program would accept submissions totaling more or less than 35 hours, respondents could see a running total of their hour allocations, and 98% of the submissions totaled exactly 35 hours. After this task, we asked a final, time-focused item: In your opinion, how many hours of homework should [your child/high school students in your local community/high school students in the United States] have per week? This item was open ended, and we did not provide any indication of how much homework is typical.

Having asked how schools should allocate time across subjects, we next turned to how they should teach those subjects. We selected three core subjects with prominent debates over how they should be taught. These subjects were mathematics, language and literature, and social studies. For each subject, we presented two instructional approaches and asked respondents to identify one as their higher priority. For math, the two approaches represented each side of the math wars (Schoenfeld, 2004). Respondents could choose either teach students to apply mathematics to real-world situations and challenges or teach students mathematical formulas, concepts, and step-by-step procedures. For language and literature, we asked whether English classes should prioritize literature and creative writing or informational text (Mathews, 2012). Respondents could choose either improve students appreciation and understanding of literature and creative writing (e.g., through poetry, classic books, and other fiction writing) or improve students ability to understand and analyze informational texts (e.g., through newspaper articles, government reports, and other nonfiction work). For social studies, we focused on the debate about whether social studies classes should prioritize the transmission of facts or socialization into societal norms and reflective thinking (Ross, 2006). Respondents could choose either teach students important facts about history, government, and related topics or teach students about their roles as citizens and the diversity of their world.

Finally, having asked what schools should teach and how they should teach it, we turned to the tangible performance metrics that people find most helpful in assessing school quality. Because publicly disseminated performance metrics can be more visible to school-choosing parents and the public than schools actual behaviors, how people prioritize these metrics can affect schools incentives. For example, if test scores are critically important to the public, then school board members and superintendents may have strong incentives to maximize test scores as they govern public schools. We asked respondents to rank four metrics beginning with the one that best helps you decide [which high school you want your child to attend/whether high schools in your local community are doing a good job/whether U.S. high schools are doing a good job]. To identify commonly available high school performance metrics, we reviewed government and third-party school performance reports. Respondents ranked four metrics: performance on state tests, especially in math and reading; graduation rates; results from surveys of students and parents about their satisfaction with the school; and performance on college entrance exams (e.g., the SAT). A respondents most preferred metric received a 1 and least preferred metric received a 4.

Like with Study 1, we tested for differences across the three treatment conditions (own childrens schools, schools in ones local community, and schools around the country) and for differences with the two broader public conditions pooled together. We report the results of each set of tests. For the tests with all three conditions, we ran either chi-square tests of independence, one-way analyses of variance (ANOVAs), or a multivariate analysis of variance (MANOVA). We ran chi-square tests of independence to assess whether respondents preferred instructional approaches in English, math, and social studies varied across our three treatment conditions. We ran one-way ANOVAs to test whether respondents hour allocations for school overall, homework, and individual subjects differed across conditions, and a MANOVA to test whether time allocation in these subjects (collectively) differed across conditions. For the comparisons between the school chooser condition and the pooled broader public condition, we ran t-tests. Taken together, these tests generate evidence of whether school choosers and the public have different desires for the tangible, everyday behaviors of schools (RQ #3).

RESULTS FOR TANGIBLE SCHOOL BEHAVIORS

The results from Study 2, like those from Study 1, show strong similarities in the preferences of school-choosing parents (in what they want for their own childrens schools) and the broader public (in what it wants for schools across its communities and country). Results appear in Table 5. To more directly assess differences between school choosers and the public, we focus our discussion on the analyses that aggregate the community and country conditions.

Table 5.  Desires for Tangible Behaviors by Whether Asked About Own Children's Schools, Schools Around the Community, or Schools Around the Country (Study 2)
	Mean responses by condition (means and standard deviations)							Tests of differences across conditions (test statistics and significance levels)
	Own child(ren)'s schools	Schools around community	Schools around country	Schools around community/ country (aggregated)			Own vs. community/ country (aggregated)		Own vs. community vs. country (disaggregated)
Hours in the school week (mean)	35.2	36.1	36.0	36.1			t(163)=-1.01		F(2,162)=0.51
	[4.8]	[5.7]	[5.7]	[5.7]
Mean hours by subject per week
Language & Literature	4.81	4.35	4.87	4.61			t(163)=0.95		F(2,162)=2.94*
	[1.16]	[1.24]	[1.32]	[1.30]
Mathematics	4.96	4.77	5.08	4.92			t(163)=0.18		F(2,162)=0.80
	[1.33]	[1.37]	[1.25]	[1.32]
Science	4.84	4.72	4.74	4.74			t(163)=0.46		F(2,162)=0.11
	[1.39]	[1.38]	[1.41]	[1.35]
Social Studies	3.95	3.93	3.90	3.91			t(163)=0.20		F(2,162)=0.03
	[1.05]	[1.44]	[1.07]	[1.26]
Lunch and other social/free time	3.73	3.75	3.91	3.83			t(163)=-0.36		F(2,162)=0.20
	[1.56]	[1.92]	[1.43]	[1.69]
Computers & Technology	4.01	4.50	3.70	4.10			t(163)=-0.31		F(2,162)=2.84*
	[1.94]	[1.88]	[1.54]	[1.76]
Foreign Language	2.82	2.74	2.90	2.82			t(163)=0.02		F(2,162)=0.18
	[1.44]	[1.37]	[1.35]	[1.36]
Physical Education	3.07	3.39	2.66	3.03			t(163)=0.19		F(2,162)=4.01**
	[1.33]	[1.48]	[1.25]	[1.41]
Visual & Performing Arts	2.70	2.94	2.62	2.78			t(163)=-0.30		F(2,162)=0.67
	[1.65]	[1.49]	[1.38]	[1.44]
Homework hours	8.07	7.95	8.07	8.01			t(163)=0.08		F(2,162)=0.01
	[4.42]	[5.63]	[3.94]	[4.85]
Instructional focus
Lang. & Lit.: informational text	29.6%	32.1%	32.7%	32.4%			t(163)=-0.36		Χ2(2,165)=0.07
Lang. & Lit.: lit. & creative writing	70.4%	67.9%	67.3%	67.6%
Math: real-world situations	33.3%	17.9%	14.6%	16.2%			t(163)=2.53**		Χ2(2,165)=3.28**
Math: formulas & procedures	66.7%	82.1%	85.5%	83.8%
Social Studies: facts	63.0%	67.9%	61.8%	64.9%			t(163)=-0.24		Χ2(2,165)=0.25
Social Studies: roles as citizens	37.0%	32.1%	38.2%	35.1%
School performance measures (mean rank, 1–4; 1=favorite)
State test scores	2.44	2.68	2.40	2.54			t(163)=-0.55		F(2,162)=1.13
	[1.02]	[1.08]	[1.05]	[1.07]
Graduation rates	2.65	2.13	2.00	2.06			t(163)=3.35***		F(2,162)=5.78***
	[1.12]	[0.99]	[1.05]	[1.02]
Surveys of students & parents	2.61	2.71	3.02	2.86			t(163)=-1.25		F(2,162)=1.63
	[1.23]	[1.29]	[1.15]	[1.22]
College entrance exam scores	2.30	2.48	2.58	2.53			t(163)=-1.36		F(2,162)=1.05
	[1.09]	[1.03]	[1.01]	[1.02]
# of observations	54	56	55	111
Notes. Standard deviations appear in brackets. ***p < .01. ** p <.05. *p < .10.

Preferences for how much time students should spend in school, in each subject and activity, and on homework were extremely consistent across respondents treatment conditions. Respondents asked about their own childrens schools desired 35.2-hour school weeks, on average, compared with 36.1 hours for those asked about schools around their communities or country (not a significant difference). Moreover, while respondents wanted more time on certain subjects (e.g., math, science, and English) than others, these preferences were similar for school-choosing parents and the broader public. In fact, none of the nine subject-specific hour allocations differed across school choosers and the broader public by even 15 minutes per week.⁸ Responses were also extremely similar across treatment conditions with respect to the number of homework hours desired. Respondents asked about their own childrens schools wanted, on average, 8.07 hours of homework per week, compared with 8.01 hours for respondents asked about schools around their communities and country. In sum, our questions about how schools should use time (the core content of the school day) reveal no meaningful differences between the desires of school choosers and the broader public.

We see modest differences across treatment conditions with respect to respondents preferred instructional approaches. Approximately two thirds of both the school chooser and broader public groups preferred that English classes prioritize literature and creative writing to informational text. Approximately two thirds of each group also preferred that social studies classes prioritize the transmission of facts to socializing students into their roles as citizens. In neither case was the difference significant. However, we do observe a statistically significant difference in math. Respondents from each condition preferred instruction grounded in procedures and formulas to instruction grounded in real-world applications, but this was especially true of those assigned to think about schools across their communities or country. One potential explanation for this finding is that rote learning seems less enjoyable to people, and parents care more about their childrens enjoyment in school than the public cares about student enjoyment more generally. Alternatively, we cannot rule out the possibility that this finding was a statistical artifact. With outcome measures as numerous as ours, some spurious, statistically significant differences are likely to arise.

Seeing little disagreement between school choosers and the public in how they want schools to spend their time and teach their classes, we look finally to which school performance metrics each group values. Here, too, we see broad similarities, with one notable exception: School choosers saw less value in graduation rates. In fact, graduation rates are the least preferred metric for school choosers and the most preferred metric for the broader public (p < .01).⁹ One plausible explanation for this is parents overconfidence that their children will complete high school. The Fordham Institute (2013) found that 99% of a nationally representative sample of parents expected their children to graduate from high school or earn a GED, while a Pew Research Center survey, also using a nationally representative sample, found that 94% of parents believe that at least one of their children will attend college (Taylor et al., 2011). These figures seem improbably optimistic given that estimates of the actual U.S. high school graduation rate run closer to 75%80% (Heckman & LaFontaine, 2010; Swanson & Lloyd, 2013). It could be that school choosers downplay the importance of graduation rates because they see their children as virtually certain to graduate. If so, then shifting authority from the public to school choosers might create incentives for schools to focus on improving metrics other than graduation rates.

Study 2 provides no strong evidence that school choosers and the broader public have markedly different desires for some of the defining, tangible, everyday realities of schooling. With numerous tests, some statistically significant differences are likely to arise by chance, and the responses from school choosers and the broader public are generally much more alike than different. In Study 1, we found no support for the hypothesis that school choosers demand that their childrens schools pursue their students private interestsat the expense of public intereststo a greater extent than the broader public would like. In Study 2, we find little reason to believe that school choosers and the broader public differ in their demands for the structure and substance of schooling.

DISCUSSION

The passions currently stirred by the school choice debate often stem from the belief that increasing parental school choice has the potential to fundamentally reshape the aims and content of schooling. Education systems relying on marketplaces where consumers decide which schools their children attend might pursue substantially different endsvia substantially different meansthan systems controlled by democratic institutions and the citizens who select representatives to control those institutions. One commonly stated hypothesis is that school-choosing parents could fixate on their own childrens private interests, especially their professional interests and social mobility, at the expense of more collective social, economic, and political interests (e.g., Labaree, 1997).

However, the hypothesis that parents want schools to prioritize their students private interests more than the public would like has not been subjected to careful empirical testing. Moreover, the question of how differences in abstract desires for schools might relate to desires for tangible school behaviors has received little attention of any sort, empirical or otherwise. This study brings empirical evidence to questions of what school choosers and the public desire from schools, examining both their desires for abstract, ideological school goals and their desires for the tangible, everyday behaviors of schooling.

We find strikingly little difference between parents and the broader publics desires for abstract school goals and tangible school behaviors. These findings might seem surprising in the context of the hypothesized differences between school choosers and public desires, and indeed, further study of these questions is essential to ensure that our findings were not unduly affected by decisions about how to operationalize variables and administer surveys. That said, we have taken great care in the design, administration, and analysis of this work. And although our findings challenge a prominent hypothesis, they are consistent with other theories and research suggesting that school-choosing parents and the public might not differ as posited (Hood, 1995; Powell & DiMaggio, 1991; Schedler & Felix, 2000; Suizzo, 2007).

A limitation of this study is that it does not capture the full complexity of the relationships between peoples desires, their behaviors, and how institutions respond to these desires and behaviors. There is certainly a pathway by which differences between the desires of school-choosing parents and the public could yield different pressures on schools. However, governance is complicated, and there are many ways in which the pathway from parent or public desires to school behaviors could be disrupted. For example, if policy makers are more responsive to interest groups than the public, then the publics influence could be modest. The same could be true if peoples behaviors are disconnected from their own desires (e.g., because they lack the information needed to identify which schools or candidates best serve their interests). Even in a democracy, just having the public desire something is not enough to guarantee that its government will provide it.

What, then, does this study imply for school choice and educational governance? Above all, our findings suggest reframing a school choice debate often predicated on the assumption that parents differ from the public in how much they prioritize their childrens private success compared with more shared, collective values. This is not to say that parents and the public do not differ in potentially consequential ways. For one, parents and the public may apply different pressures to schools not because they disagree about the proper balance of goals or behaviors but because they rely on different information. For example, parents likely acquire more textured, firsthand knowledge about schools through their own direct experience and the direct experience of those in their social networks. In addition to performing well on publicly disseminated performance measures, schools operating in market settings might feel stronger pressure to increase parent satisfaction or otherwise tend to their local reputations. Second, school choice reforms could alter the pressures schools face by empowering a groupparentsthat is perhaps more concerned about school quality than the broader public. For better or worse, parents may be willing to invest considerably more time and energy into ensuring that their desires for schools are realized. This could mean that schools in market settings, where parental scrutiny and involvement can be more direct, experience different pressures than schools in more traditionally governed settings.

An additional implication of this work is that traditional understandings of private and public goals in education may need to be revisited. Our findings do not refute that parents want their own children to attend the best colleges and achieve the greatest professional success. They just may want their children to be equally exceptional in their moral conduct, social relationships, and well roundedness. To restrict our understanding of parents private interests in their childrens education to academic and professional success is to overlook the many other personal attributes that parents hope to cultivate in their children. At the same time, to restrict our understanding of the publics desires for schools to the advancement of collective social, economic, and political interests is to overlook the possibility that the public wants schools to primarily serve the needs of the students enrolleda possibility increasingly likely as public institutions become more and more customer oriented.

Our findings also indicate that the question of whether schools should pursue private or collective interests could be divisive, but divisive along partisan lines rather than parentpublic lines. When we examined which characteristics predict whether respondents prioritized students Private Success, we found that the only statistically significant predictor was identification with the Republican Party. Republicans, who have historically been more supportive of school choice than Democrats, are considerably more inclined to want schoolsboth their own childrens schools and schools more generallyto pursue students private success ahead of more collective economic and democratic interests. This is perhaps consistent with a Pew Research Center (2014) finding that 86% of politically liberal Americans say it is important to teach children empathy, whereas only 55% of conservatives say the same.

The fact that a subset of the American population that favors school choice also favors schools pursuit of private interests does not imply that school choosers, as a whole, want schools to pursue their childrens private interests any more than the broader American public would like. However, with Republicans more likely than Democrats to favor schools pursuit of students private success, the political affiliation of those setting education policy could be consequential. Moreover, with Americans increasingly living in communities that share their political views (Bishop, 2009), we might expect further divergence in the pursuits of schools located in conservative and liberal areas.

Future research on school choice would do well to try to connect the poetry of educational aspirations to the prose of daily schooling. Regardless of whether parents and the public agree on the relative importance of private and public goals, it is critical to understand how they would like to see those goals achieved on the ground. We found few differences in what parents and the public desire when it comes to the everyday behaviors of schools that matter most. Nonetheless, examining which schooling behaviors people see as contributing to which goals, and in what ways, would further enrich our understanding of the relationship between abstract aims and tangible behaviors.   

Over 100 years ago, John Dewey wrote, What the best and wisest parent wants for his own child, that must the community want for all of its children. Any other ideal for our schools is narrow and unlovely; acted upon, it destroys our democracy (1900, p. 19). One wonders if the reverse statement might also be true: What the best and wisest community wants for all of its children, that must the parent want for his or her own child. In the present research, we have found notable similarities between what parents and the public want from schools. Consensus should not be confused for wisdom, but this consensus does suggest that parents and the publics wishes for schools may not differ in the ways commonly presumed.

Notes

1. There could be overlap between ones own childrens schools and the schools in ones local community, given that most children attend schools near their homes, and many communities have only a few schools. This overlap could lead respondents to answer similarly to survey questions about their own childrens schools and schools in their local communities. However, the overlap could also lead people to place similar demands on their own childrens schools (when making market choices) and on schools in their local communities (when participating in local education governance). Comparisons between what parents desire from their own childrens schools and what people desire from schools across the country consider more distinct sets of schools, but the comparison is not necessarily more relevant to policy.

2. A key assumption is conditional ignorability. Conditional on the variables included in the matching and weighting, we must assume that sample selection is independent of the surveys outcome variables. YouGov matched respondents based on gender, age, race, education, party identification, ideology, and political interest. Additional details of the YouGov process and underlying assumptions appear in Rivers (2007). Moreover, there is one potential distortion of national representativeness from the study design (beyond the assumptions underlying YouGovs process). Because we needed to assign parents differently from nonparents, parents constitute a disproportionate share of the sample for some analyses. To address this, we recalibrated the YouGov-provided weights such that our weighted sample would reflect the U.S. adult population in its share of parents with children 18 or younger. When we reran our analyses with these new weights, the results were virtually identical (and the interpretations the same). For simplicity, we report our findings using only the original YouGov-provided weights.

3. To further assess how problematic social desirability bias could be, we asked MTurk respondents the following question: Depending on what questions they are asked, people differ in how willing they are to report their honest opinions in surveys. Please think about the item that we showed you earlier in this survey. (The item is at the bottom of this page for reference.) Do you think most people would answer this question honestly or dishonestly? Of the 113 people who answered this item, 108 (95.6%) reported believing that most people would answer honestly.

4. Respondents were classified as Leans Democrat if they reported a political affiliation of Strong Democrat, Not very strong Democrat, or Lean Democrat. They were classified as Leans Republican if they reported Strong Republican, Not very strong Republican, or Lean Republican.

5. The 2012 Democratic platforms section on education begins with a statement that concisely incorporates Private Success, Shared Economic Health, and Democratic Character: Democrats believe that getting an education is the surest path to the middle class, giving all students the opportunity to fulfill their dreams and contribute to our economy and democracy (Democratic National Committee, 2012). For comparison, the 2012 Republican platforms section on education begins, Parents are responsible for the education of their children (Republican National Committee, 2012).

6. In this way, Study 2 more closely resembles a scenario study than Study 1. We did this to preserve the integrity of the experiment by having a single survey that MTurk Workers could select (and within which the randomization occurred). To make the school chooser condition coherent, those assigned to the condition saw the following text on the studys opening screen: If you are not the parent of a high school student, please imagine that you are for the duration of this study. In other words, please answer the questions throughout this study as if you are the parent of a high school student.

7. The subjects were written as follows: Language & Literature (English), Mathematics, Science, Social Studies (e.g., history, civics, government, world cultures), Lunch and other social or free time, Computers & Technology, Foreign Language, Physical Education (e.g., gym, health), and Visual & Performing Arts (e.g., painting, drama, music).

8. To account for the relationships between the dependent variables in the subject/activity hour allocationsand to adjust for the increased possibility of Type I error present with nine different subjects/activities testedwe conducted a one-way MANOVA (with the three treatment conditions disaggregated). It shows no significant relationship between respondents treatment assignment and subject/activity hour allocations, Wilks λ=0.873, F(18,308)=1.20, p = 0.257.

9. Post-hoc analysis using Tukey's HSD indicates that there is no statistically significant difference between the local community group (M = 2.13, SD = 0.99) and country group (M = 2.00, SD = 1.05) with respect to the mean rank that respondents assigned to graduation rates. There are significant differences between the school chooser group (M = 2.65, SD = 1.12) and local community group (p = 0.027) and between the school chooser and country groups (p = 0.005) for this measure.

Acknowledgment

We are grateful to David Labaree, Susanna Loeb, Terry Moe, Sean Reardon, Paul Sniderman, and Michael Tomz for their insights and suggestions. Funding was provided by the Institute of Education Sciences Predoctoral Training Program (Award R305B090016), the National Academy of Education/Spencer Foundation Dissertation Fellowship Program, the Regina Casper Stanford Graduate Fellowship in Science and Engineering, and the Stanford University Laboratory for the Study of American Values. The views expressed do not necessarily reflect those of the studys funders. All errors are our own.

References

American Federation for Children. (2016). Ed Choice 101: Facts. Retrieved from http://www.federationforchildren.org/ed-choice-101/facts/

Belfield, C., & Levin, H. M. (2005). Vouchers and public policy: When ideology trumps evidence. American Journal of Education, 111(4), 548567.

Bell, C. (2009). Geography in parental choice. American Journal of Education, 115(4), 493521.

Berinsky, A. J., Huber, G. A., & Lenz, G. S. (2012). Evaluating online labor markets for experimental research: amazon.com's Mechanical Turk. Political Analysis, 20(3), 351368.

Bishop, B. (2009). The big sort: Why the clustering of like-minded America is tearing us apart.

New York, NY: Houghton Mifflin Harcourt.

Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon's Mechanical Turk a new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6(1), 35.

Bushaw, W. J., & Calderon, V. J. (2014). The 46th annual PDK/Gallup poll of the publics attitudes toward the public schools. Retrieved from http://pdkintl.org/noindex/PDK_Poll46_2014.pdf

Cohn, N. (2014, July 27). Explaining online panels and the 2014 midterms. New York Times. Retrieved from http://www.nytimes.com/2014/07/28/upshot/explaining-online-panels-and-the-2014-midterms.html

Dear, M. (1992). Understanding and overcoming the NIMBY syndrome. Journal of the American Planning Association, 58(3), 288300.

Democratic National Committee. (2012). Moving America forward: 2012 Democratic National Platform. Retrieved from https://www.democrats.org/party-platform

Dewey, J. (1900). The school and society. Chicago, IL: University of Chicago Press

DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2), 147160.

Duffy, B., Smith, K., Terhanian, G., & Bremer, J. (2005). Comparing data from online and face-to-face surveys. International Journal of Market Research, 47(6), 615639.

Fordham Institute. (2013). What parents want: Education preferences and trade-offs. A national survey of K-12 parents. Retrieved from

Freeman, J. (1986). The political culture of the Democratic and Republican parties. Political Science Quarterly, 101(3), 327356.

Friedman, M. (1955). The role of government in education. In R. A. Solo (Ed.), Economics and the public interest (pp. 123144). New Brunswick, NJ: Rutgers University Press.

Harris, D. N., & Larsen, M. F. (2014). What schools do families want (and why)? School demand and information before and after the New Orleans post-Katrina school reforms. Retrieved from http://educationresearchalliancenola.org/files/publications/Technical-Report-Final-Combined.pdf

Hastings, J. S., Kane, T. J., & Staiger, D. O. (2009). Heterogeneous preferences and the efficacy of public school choice (Working paper). Retrieved from http://www.justinehastings.com/images/downloads/HKS_Combined_201006.pdf

Heckman, J. J., & LaFontaine, P. A. (2010). The American high school graduation rate: Trends and levels. Review of Economics and Statistics, 92(2), 244262.

Henderson, M. B., & Peterson, P. E. (2014). The 2013 Education Next survey: Americans react to Common Core and other education policies. Education Next. Retrieved from

http://educationnext.org/the-2013-education-next-survey/

Hood, C. (1995). The New Public Management in the 1980s: Variations on a theme.

Accounting, Organizations and Society, 20(2), 93109.

Jacob, B. A., & Lefgren, L. (2007). What do parents value in education? An empirical investigation of parents revealed preferences for teachers. Quarterly Journal of Economics, 122(4), 16031637.

Jones, J. M. (2014, January 8). Record-high 42% of Americans identify as Independents. Gallup. Retrieved from http://www.gallup.com/poll/166763/record-high-americans-identify-independents.aspx

Jones, J. M. (2013, September 4). Americans grade math as the most valuable school subject. Gallup. Retrieved from http://www.gallup.com/poll/164249/americans-grade-math-valuable-school-subject.aspx

Krashinsky, M. (1986). Why educational vouchers may be bad economics. Teachers College

Record, 88(2), 139167.

Kreuter, F., Presser, S., & Tourangeau, R. (2008). Social desirability bias in CATI, IVR, and Web surveys: The effects of mode and question sensitivity. Public Opinion Quarterly, 72(5), 847865.

Labaree, D. F. (1997). Public goods, private goods: The American struggle over educational goals. American Educational Research Journal, 34(1), 3981.

Lakoff, G. (2010). Moral politics: How liberals and conservatives think. Chicago, IL: University of Chicago Press.

Levin, H. M. (1991). The economics of educational choice. Economics of Education Review, 10(2), 137158.

LeVine, R. (1974). Parental goals: A cross-cultural view. Teachers College Record, 76(2), 226239.

March, J. G. (1978). American public school administration: A short analysis. The School

Review, 86(2), 217250.

Mathews, J. (2012, October 17). Fiction vs. nonfiction smackdown. Washington Post.

Metz, M. H. (1989). Real school: A universal drama amid disparate experience. Politics of Education Association Yearbook, 4(5), 7591.

Meyer, J. W., & Rowan, B. (1977). Institutionalized organizations: Formal structure as myth and ceremony. American Journal of Sociology, 83(2), 340363.

Paolacci, G., Chandler, J., & Ipeirotis, P. G. (2010). Running experiments on Amazon Mechanical Turk. Judgment and Decision Making, 5(5), 411419.

Pew Research Center. (2014). Teaching the children: Sharp ideological differences, some common ground.

Powell, W. W., & DiMaggio, P. J. (1991). The new institutionalism in organizational analysis. Chicago, IL: University of Chicago Press.

Rao, J. N., & Scott, A. J. (1984). On chi-square tests for multiway contingency tables with cell proportions estimated from survey data. The Annals of Statistics, 12(1), 4660.

Republican National Committee. (2012). 2012 Republican Platform. Retrieved from http://www.gop.com/wp-content/uploads/2012/08/2012GOPPlatform.pdf

Rivers, D. (2007). Sampling for web surveys. Presented at the 2007 Joint Statistical Meetings, Salt Lake City, UT. Retrieved from http://www.laits.utexas.edu/txp_media/html/poll/files/Rivers_matching.pdf

Rivers, D. (2012). YouGov poll performance in the 2012 U.S. elections. Retrieved from https://today.yougov.com/news/2012/11/07/yougov-poll-performance-2012-us-elections/

Ross, E. W. (2006). The social studies curriculum: Purposes, problems, and possibilities (3rd ed.). Albany, NY: State University of New York Press.

Rothstein, R., & Jacobsen, R. (2006). The goals of education. Phi Delta Kappan, 88(4), 264272.

Schedler, K., & Felix, J. (2000). Quality in public management: The customer perspective. International Public Management Journal, 3(1), 125143.

Schneider, M., & Buckley, J. (2002). What do parents want from schools? Evidence from the Internet. Educational Evaluation and Policy Analysis, 24(2), 133144.

Schoenfeld, A. H. (2004). The math wars. Educational Policy, 18(1), 253286.

Silver, N. (2012, November 10). Which polls fared best (and worst) in the 2012 presidential race. New York Times.

Stein, M, L., Goldring, E. B., & Cravens, X. (2011). Do parents do as they say? Choosing Indianapolis charter schools. In M. Berends, M. Cannata, & E. B. Goldring (Eds.), School choice and school improvement (pp. 105124). Cambridge, MA: Harvard Education Press.

Stewart, T., & Wolf, P. J. (2014). The school choice journey: School vouchers and the empowerment of urban families. New York, NY: Palgrave Macmillan.

Suizzo, M. A. (2007). Parents' goals and values for children: Dimensions of independence and interdependence across four U.S. ethnic groups. Journal of Cross-Cultural Psychology, 38(4), 506530.

Swanson, C. B., & Lloyd, S. C. (2013, May 31). Nations graduation rate nears a milestone. Education Week. Retrieved from http://www.edweek.org/ew/articles/2013/06/06/34analysis.h32.html

Taylor, P., Parker, K., Fry, R., Cohn, D., Wang, W., Velasco, G., & Dockterman, D. (2011). Is college worth it? College presidents, public assess value, quality, and mission of higher education. Pew Research Center. Retrieved from http://www.pewsocialtrends.org/files/2011/05/higher-ed-report.pdf

U.S. Census Bureau. (2013a). National characteristics: Vintage 2013. Retrieved from http://www.census.gov/popest/data/national/asrh/2013/index.html

U.S. Census Bureau. (2013b). Educational attainment in the United States: 2013Detailed tables. Retrieved from http://www.census.gov/hhes/socdemo/education/data/cps/2013/tables.html

U.S. Department of Education, National Center for Education Statistics. (2015). Table 216.20: Number and enrollment of public elementary and secondary schools, by school level, type, and charter and magnet status: Selected years, 1990-91 through 2013-14. Retrieved from http://nces.ed.gov/programs/digest/d15/tables/dt15_216.20.asp

Cite This Article as: Teachers College Record Volume 119 Number 11, 2017, p. 1-34 https://www.tcrecord.org ID Number: 21992, Date Accessed: 12/23/2021 11:43:51 PM Purchase Reprint Rights for this article or review