Status,Growth,and Perceptions of School Quality

Abstract

States and districts are increasingly incorporating measures of achievement growth into their school accountability systems, but there is little research on how these changes affect the public’s perceptions of school quality. We conduct a nationally representative online survey experiment to identify the effects of providing participants with information about their local public schools’ average achievement status and/or average achievement growth. Prior to receiving any information, participants already possess a modest understanding of how their local schools perform in terms of status, but they are largely unaware of how these schools perform in terms of growth. Participants who live in higher status districts tend to grade their local schools more favorably. The provision of status information does not fundamentally change this relationship. The provision of growth information, however, alters Americans’ views about local educational performance. Once informed, participants’ evaluations of their local schools better reflect the variation in district growth.

Keywords

accountability achievement educational policy policy politics experimental design experimental research survey research

Prior to the passage of the Every Student Succeeds Act of 2015 (ESSA), states’ public K–12 school accountability systems focused almost exclusively on measures of student achievement status (i.e., the level of students’ academic performance at a single point in time). ESSA’s predecessor, the No Child Left Behind Act of 2002, required states to measure achievement status by administering annual assessments in math and reading in Grades 3 to 8 and once in high school. The results of these assessments factored heavily in schools’ accountability ratings, shaping states’ efforts to improve educational outcomes as well as public perceptions of school quality. Critics argued that achievement status was a misleading indicator of school quality (Chingos & West, 2015; Rothstein et al., 2008). Schools that served a larger proportion of disadvantaged students tended to perform poorly in terms of achievement status because of their students’ lower initial test scores and additional out-of-school obstacles. Conversely, schools that served a larger proportion of students from more advantaged backgrounds tended to perform well by this standard, regardless of the schools’ actual contributions to student learning.

ESSA now requires states to use multiple measures to evaluate students and schools (Barone, 2017). The most significant change has been the widespread inclusion of student achievement growth (i.e., the rate of improvement in students’ academic performance over time), which many education policy researchers consider a better—albeit still imperfect—indicator of school quality than achievement status (Stiefel et al., 2011). Aggregate measures of growth also have a much weaker relationship to the racial, ethnic, and economic composition of the student body (Reardon, 2019). Forty-three states and the District of Columbia include growth in their publicly available school report cards, and another five states plan to include growth in future iterations (Data Quality Campaign, 2019, 2020).

There is considerable variation among states with respect to how they measure growth and how they present growth data to the public. Most states use one of five approaches to measuring growth: value-added scores, student growth percentiles, value tables, gain scores, or growth to standard (for a brief description of each, see Data Quality Campaign, 2019; for a more detailed discussion, see Castellano & Ho, 2013). The 2018–2019 Iowa school report cards offer a particularly accessible example. Their measure of school-level growth appears on the school summary page with a simple graphic, a short description (“Growth is measured using student growth percentiles. A student growth percentile describes a student’s growth compared to other students with similar prior test scores”), and a link to a more technical explanation (Iowa Department of Education, 2021). By contrast, the 2018–2019 Alabama school report cards offer a single number to represent school-level growth without additional elaboration (Alabama Department of Education, 2021). Among states that include growth in their summative school accountability scores, the relative weight varies from 20% to 60% (Achieve, Inc., 2019). While only a small subset of the public seeks out school accountability data directly from their state departments of education, many Americans encounter this information on websites like GreatSchools.org or Niche.com that generate school ratings based on state data. GreatSchools.org recently revised its summative school rating formula to place more weight on growth (Barnum, 2020).

There is a robust empirical literature on the effects of achievement status information on the public’s perceptions of school quality (Barrows et al., 2016; Chingos et al., 2012; Clinton & Grissom, 2015; Jacobsen et al., 2013, 2014), students’ and families’ school choices (Corcoran et al., 2018; Hastings & Weinstein, 2008), housing prices (Black, 1999; Figlio & Lucas, 2004; Fiva & Kirkebøen, 2011), school board elections (Berry & Howell, 2007; Holbein, 2016; Payson, 2016), and school tax referenda (Kogan et al., 2016). There is a nascent but growing literature on the effects of achievement growth information on students’ and families’ school choices (Ainsworth et al., 2020; Valant & Weixler, 2020) and housing prices (Imberman & Lovenheim, 2016). However, to our knowledge, there is no existing research on the effects of growth information on the public’s perceptions of school quality more generally. This gap is important because perceptions of school quality are presumably the mediating factor between the distribution of academic performance information and the educational, economic, and political outcomes listed above.

To address this gap, we conducted an experiment embedded in a nationally representative online survey of U.S. adults.¹ We first asked participants to estimate their local school district’s performance in terms of status and their local school district’s performance in terms of growth. Next, we randomly assigned participants to receive one or more elements of academic performance information: either the district’s national percentile in terms of status, the district’s national percentile in terms of growth, both, or neither (to serve as a control group). We then asked participants to evaluate the quality of their local public schools.

This research design allows us to answer multiple, related research questions. First, we can gauge the accuracy of participants’ prior beliefs about local status and growth. We can then identify the average effects of providing status information, growth information, or both (representing the fact that when states include growth in their school accountability systems, they generally do so as a supplement to and not a replacement for status) on participants’ perceptions of local school quality. The results of this analysis can help us think through the potential consequences of a few different scenarios: the public’s perceptions of school quality if (a) states release no new academic performance information, (b) states return to the pre-ESSA approach to accountability focused solely on status, (c) states shift wholesale to a growth-based model of accountability, or (d) states continue to collect and disseminate both types of academic performance information.

We can observe whether these effects vary according to the content of the information (lower vs. higher performance). This allows us to assess whether participants respond differently to good or bad news about their local schools when it comes in the form of status, growth, or both. Moreover, because growth has a much weaker underlying relationship with student demographics than status, we can also examine the extent to which the effects of distributing different kinds of academic performance information vary by the demographic composition of participants’ local schools. Our approach does not, however, allow us to assess participants’ comprehension of status and growth. Additional research is necessary to understand how the public interprets—or, in some cases, misinterprets—these concepts.

We find that Americans are more familiar with status (which states have used for many years to evaluate schools) than growth (which states have only recently begun to incorporate into their school accountability systems). Regardless of their experimental condition, individuals living in higher status districts tend to grade their local schools more favorably. On average, we observe a small negative effect of giving participants information about local status. However, the magnitude of this effect is roughly the same for participants living in both lower and higher status districts. In other words, Americans already have a rough comprehension of average achievement status in their communities. Confronting this information directly may have a minor depressing effect, but it does not fundamentally change the public’s understanding of the distribution of school quality.

By contrast, the provision of information about local achievement growth alters the conventional wisdom about school quality. Among participants who receive growth information alone, the relationship between district status and perceptions of school quality becomes weaker, while the analogous relationship between district growth and perceptions of school quality becomes stronger. When we provide both types of academic performance information, the relationship between district status and perceptions of school quality is generally unaffected, while the relationship between district growth and perceptions of school quality is enhanced. In short, providing information about growth reorients the public’s perceptions of school quality to be more in line with a measure that many scholars consider a more accurate indicator of schools’ contributions to student learning. Furthermore, the provision of growth information weakens the relationship between the public’s perceptions of school quality and the economic background of the student body.

When designing our experiment, we also sought to identify one of the cognitive mechanisms potentially responsible for the effects of status and/or growth information on perceptions of school quality. After asking participants to evaluate their local schools, we also asked them about the importance of academic performance relative to other educational objectives. We expected a priori that the provision of academic performance information would raise the salience of academic outcomes. The results, on the contrary, do not conform neatly to these expectations. On average, the provision of academic performance information has no meaningful effect on the relative importance of academic performance. Among those in lower growth districts, however, the provision of growth information causes participants to indicate that they think schools should focus less on academic outcomes. The reverse is true in higher growth districts. It may be the case that many participants in lower growth districts do not appreciate or accept this new, negative depiction of their local schools, and they respond by de-emphasizing the importance of academic performance. Similarly, many participants in higher growth districts may be surprised to receive such a positive portrayal, and they respond with additional attention to academic outcomes. It could also be the case that many participants living in lower growth districts are more skeptical of the measures of academic performance featured in our experiment than their peers in higher growth districts. Their responses may reflect this wariness.

To summarize, the public’s current perceptions of school quality are largely consistent with the predominant indicator of academic performance over the last few decades: average achievement status as measured by state standardized tests. The provision of district-level information about average growth can shift the public’s perceptions of school quality to be more in line with schools’ contributions to student learning. However, portions of the public may be disinclined to embrace growth as a valuable metric. Especially among those living in lower growth districts, the provision of this information may reduce support for schools’ academic objectives and/or this particular method of measuring success toward those objectives.

The Effects of Public Service Performance Information

In the last decade, there has been a surge of empirical research on the attitudinal effects of public service performance information in policy domains such as health care, policing, mail delivery, recycling/waste removal, and education (e.g., Baekgaard & Serritzlew, 2015; James, 2011; Marvel, 2016; Walker & Archbold, 2014). Three major findings are particularly relevant to our inquiry. First, recipients of public service performance information respond more decisively to negative reports than to positive reports. The public’s satisfaction with local services declines with the provision of evidence about low performance, but the public is generally unmoved by the provision of evidence about high performance (James & Moseley, 2014). Second, information from an independent source (rather than the service provider itself) and information about performance relative to similar institutions (rather than an absolute level of performance) appear to be particularly influential (Barrows et al., 2016; James & Moseley, 2014; James & Van Ryzin, 2017). Finally, individuals’ prior beliefs about the quality of local public services shape their interpretation of the evidence they receive. When new information is inconsistent with these prior beliefs, recipients are more likely to misinterpret or discard it (Baekgaard & Serritzlew, 2015).

Achievement Status

Many Americans already possess a nontrivial understanding of achievement status in their communities. Chingos et al. (2012) asked a nationally representative sample of U.S. adults to evaluate the quality of their local public schools. They find that these ratings are positively associated with the percentage of students who scored above the proficiency threshold on their states’ standardized tests. This relationship is 2 to 3 times stronger among parents of school-age children, who might be expected to be more familiar with local schools. The provision of new information about achievement status can also shift attitudes toward schools. Researchers have examined the effects of status information on perceptions of school quality in the context of online surveys (Barrows et al., 2016; Clinton & Grissom, 2015; Jacobsen et al., 2014), official school letter grades released by the state (Chingos et al., 2012), and shifts in performance outcomes following the introduction of new state tests (Jacobsen et al., 2013). In most cases, when individuals encounter new information about achievement status, they tend to revise their appraisals of local education institutions downward.

Other scholars have considered the effects of status information on outcomes such as students’ and families’ school choices, housing prices, school board elections, and school tax referenda. Analyses of school application data in districts with centralized enrollment systems suggest that students and families place a high priority on achievement status (Glazerman & Dotter, 2017; Harris & Larsen, 2015). In large-scale field experiments, the distribution of information about average test scores and graduation rates to low-income students tends to increase enrollment in higher status schools (Corcoran et al., 2018; Hastings & Weinstein, 2008). Moreover, housing values reflect status differences in nearby schools (Bayer et al., 2007; Black, 1999; Kane et al., 2006), and the release of new information about achievement status also appears to influence housing prices (Figlio & Lucas, 2004; Fiva & Kirkebøen, 2011). In the political realm, improving or declining status—which can be influenced by changing student demographics and is not equivalent to growth—can influence vote choice and turnout in both school board elections and school tax referenda (Berry & Howell, 2007; Holbein, 2016; Kogan et al., 2016; Payson, 2016).

Achievement Growth

The analogous literature on the effects of growth information is smaller but growing rapidly. When exploring the relationship between perceptions of school quality and achievement status, Chingos et al. (2012) also establish that individuals’ ratings of local schools are weakly related to differences in growth, but this relationship is largely explained by the fact that school-level average status and school-level average growth are moderately correlated. After controlling for achievement status, the relationship between growth and ratings is not statistically significant. This is consistent with work by Abdulkadiroglu et al. (2020) and Beuermann et al. (2020), indicating that families generally prioritize attributes other than growth when ranking their school options in a centralized school enrollment system. However, survey and field experiments suggest that the provision of growth information can steer participants toward schools and districts that exhibit higher growth rates (Ainsworth et al., 2020; Houston & Henig, 2021; Schneider et al., 2018; Valant & Weixler, 2020). On the contrary, the release of Los Angeles Unified School District teacher and school value-added data in the Los Angeles Times had no effect on housing prices—although this situation may have been atypical given the controversial nature of the data release (Imberman & Lovenheim, 2016). In sum, the available evidence suggests that Americans possess little prior knowledge about school performance in terms of growth, and the provision of this information may have considerable influence on their attitudes toward those educational institutions.

Multiple Educational Objectives

A consistent challenge with respect to measuring educational performance is the multiplicity of objectives that schools are expected to pursue: cultivating students’ academic skills, civic values, social and emotional well-being, artistic appreciation, athletic ability, and much more (Jacobsen, 2009; Ladd & Loeb, 2013; Rothstein et al., 2008). Previous work by Jacobsen et al. (2015) indicates that individuals with different normative expectations for schools—either a heavy emphasis on academic outcomes or a more equal balance among multiple educational objectives—react differently to academic performance information. Those who place greater emphasis on students’ academic development tend to respond more negatively to indications of lackluster performance on standardized tests. By contrast, those who prefer more balance between academic and nonacademic objectives appear to be less critical of schools that underperform on standardized tests if they are strong in other areas. We are unaware of research that examines the converse relationship: how the provision of academic performance information can influence attitudes about the optimal balance between various educational objectives.

Priming Versus Learning

When studying the effects of public service performance information, it is important to consider whether the results we observe are due to participants learning something new or if they are merely the consequences of priming. Priming refers to the process through which individuals become temporarily attuned to different considerations when answering questions, making decisions, or performing actions (Sherman et al., 1990). Priming occurs when a stimulus (like a survey question) briefly increases the salience of one consideration (such as the importance of academic performance when evaluating school quality) at the expense of other relevant considerations (such as the importance of students’ social and emotional well-being). The effects of priming disappear quickly as the newly salient consideration wanes in prominence. To differentiate between learning and priming, previous studies examined whether the effects of information were larger for individuals who underestimated or overestimated the value in question—a pattern that would be more consistent with learning than with priming (Clinton & Grissom, 2015; Schueler & West, 2016). We employ the same approach in our analysis. We also test the priming hypothesis directly by identifying the effects of status and/or growth information on the importance of academic performance relative to other educational objectives.

Method

Preregistration

This experiment has been preregistered on the American Economic Association’s registry for randomized controlled trials. The research questions and the accompanying analyses presented here are consistent with the preanalysis plan posted on the registry.

Research Questions

We divide our research questions into two categories: primary and secondary. As the number of statistical tests necessary to answer these questions increases, so does the likelihood of false positives. The reader should place more confidence in the results of the analyses associated with the primary research questions. The results of the analyses associated with the secondary research questions should be viewed as exploratory.

Primary Research Questions

Research Question 1: To what extent are individuals able to estimate average achievement status and average achievement growth in their school districts?

Research Question 2: Does the provision of status and/or growth information affect the grades that participants assign to their local public schools?

Research Question 3: Does the provision of this information affect the importance of academic performance relative to other educational objectives?

Research Question 4: Do these effects vary by the academic performance of participants’ districts?

Secondary Research Questions

Research Question 5: Do these effects vary by the racial/ethnic and economic compositions of participants’ districts?

Research Question 6: Could these effects be the result of participants updating their prior beliefs about academic performance in their districts?

Data

We embedded an experiment in the 2019 EducationNext Poll, an annual survey of Americans’ attitudes toward education issues. The survey was conducted from May 14 to May 25, 2019, by the polling firm Ipsos Public Affairs via its KnowledgePanel®. In the KnowledgePanel®, Ipsos Public Affairs maintains a nationally representative panel of more than 50,000 adults (obtained via address-based sampling techniques) who agree to participate in a limited number of online surveys, providing noninternet households with internet access and a device with which to participate. Ipsos then samples from this panel to obtain participants for particular surveys, such as the EducationNext Poll. This survey features a sample of 3,046 respondents, including a nationally representative, stratified sample of adults (age 18 and older) in the United States as well as representative oversamples of the following subgroups: teachers (667), African Americans (597), and Hispanics (648). Survey weights are employed to account for nonresponse and the oversampling of specific groups. Respondents could elect to complete the survey in English or Spanish.

Ipsos Public Affairs provided us with extensive demographic information for each participant: race/ethnicity, teacher status, parent status, Spanish language status, political party identification, political ideology, household income, U.S. Census region, age, educational attainment, gender, head of household status, housing type, marital status, and employment status. In addition, Ipsos provided the census block identifier for each respondent. We used U.S. Census files linking block identifiers to school districts to match each respondent to his or her local school district. For participants living in areas with separate elementary and secondary districts, we link them to their elementary district. In all, our respondents reside in 1,893 school districts. Importantly, Ipsos provided census block identifiers for the total sample prior to fielding the survey, allowing us to incorporate locally tailored information about school districts in the experiment.

For measures of district-level average status, average growth, free and reduced-price lunch (FRPL) eligibility, and racial/ethnic composition, we use the Stanford Education Data Archive v2.1 (SEDA). SEDA contains data from state standardized tests in reading and math in Grades 3 to 8 administered from 2009 to 2015 for almost every school district in the United States. For each district, SEDA contains average status and growth in reading and math as well as the average across both subjects (we employ these combined values in our experiment). SEDA defines school districts in geographic terms. The dataset contains student performance data for all public schools located in the geographic boundaries of the district, including charter schools. The student test score data have been converted to a common scale that allows district-to-district comparisons across the country (Fahle et al., 2018).

SEDA’s academic performance measures are derived from the U.S. Department of Education’s EDFacts Data Initiative, which contains district-level achievement data by grade, year, and subject. The structure of the EDFacts data has an important drawback for the estimation of district growth. Ideally, growth measures the rate at which individual students’ achievement improves over time. The aggregated nature of the EDFacts data allows only for the estimation of grade–year–subject cohort gains over time, which can be biased by within-cohort shifts in student demographics. However, comparisons of SEDA’s district growth estimates and those generated by state longitudinal student data systems (which would be preferable but are neither widely available nor, in their raw form, directly comparable across states) show that the two are closely correlated (Reardon et al., 2019).

We use SEDA’s empirical Bayes Grade Cohort Scale estimates for the measures of status and growth. To aid in the interpretability of these values for participants, we provide status and growth scores in terms of national percentiles. For example, we present growth information in the survey as follows: “The rate of growth in student academic performance in your school district is better than in [growth percentile] percent of districts and worse than in [100 – growth percentile] percent of districts” (see the following section for more details about the survey text). Prior research suggests that even minor differences in the presentation of school information (different phrasings, graphical representations, sequencing of information, etc.) can influence recipients’ reactions and subsequent behavior (Glazerman et al., 2020). We readily acknowledge that our presentation of status and growth information is only one way that this content could be conveyed to the public. Additional research is warranted on the effects of different presentations of academic performance information.

Experimental Design

Participants are randomly assigned with equal probability to one of four experimental groups:

Participants in the status group receive their district’s national percentile in terms of average achievement status.

Participants in the growth group receive their district’s national percentile in terms of average achievement growth.

Participants in the both group receive both their district’s national percentile in terms of average achievement status and their district’s national percentile in terms of average achievement growth.

Participants in the control group do not receive academic performance information for their district.

At the beginning of the survey, all participants are asked to estimate how their local school district performs in terms of average achievement status. They receive the following prompt:

The next few questions are about the current level of student academic performance and the rate of growth or improvement in student academic performance in your school district from one year to the next.

Enter any number from 0 to 100.

I think the current level of student academic performance in my school district is better than [number box, range 0–100] percent of other districts in the United States.

Next, they estimate how their district performs in terms of average achievement growth:

Enter any number from 0 to 100.

I think the rate of growth in student academic performance in my school district is better than [number box, range 0–100] percent of other districts in the United States.

Depending on their experimental assignment, some participants receive information about their district’s academic performance. Those assigned to the status group receive:

According to the most recent information available, the current level of student academic performance in your school district is better than in [achievement percentile] percent of districts and worse than in [100 – achievement percentile] percent of districts.

Those assigned to the growth group receive:

According to the most recent information available, the rate of growth in student academic performance in your school district is better than in [growth percentile] percent of districts and worse than in [100 – growth percentile] percent of districts.

Those assigned to the both group receive both pieces of information displayed above, while those assigned to the control group receive neither.

All participants then receive the following question about the quality of their local public schools:

Students are often given the grades A, B, C, D, and Fail to denote the quality of their work. Suppose the public schools themselves were graded in the same way. What grade would you give the public schools in your community? (Answer options: A, B, C, D, or Fail)

This question employs the standard wording for measuring confidence in the public schools as tracked by Loveless (1997) and Bali (2016).

Last, all participants receive the following question about the relative importance of student academic performance versus student social and emotional well-being (the sequence of “student academic performance” and “student social and emotional well-being” is randomized to eliminate ordering effects):

How much should schools focus on student academic performance versus student social and emotional well-being?

Please give a percentage for each. Your answers should add to 100%.

1. Student academic performance [number box, 0–100] %

2. Student social and emotional well-being [number box, 0–100] %

Total [show sum of boxes]

Analytic Approach

To check for balance between experimental groups, we compare the demographic composition of the control group with the demographic compositions of each of the other randomly assigned groups. To accomplish this, we use a series of weighted least squares (WLS) regressions:

X_{i} = a + b S_{i} + c G_{i} + d B_{i} + u_{i},

where $X_{i}$ is one of the available demographic covariates (from the set of individual-level demographic characteristics provided by Ipsos Public Affairs); $S_{i}$ , $G_{i}$ , and $B_{i}$ are indicators for each experimental group (the status group, the growth group, and the both group); and $u_{i}$ is the error term for individual $i$ . Ipsos Public Affairs computed poststratification statistical weights by age, gender, race/ethnicity, U.S. Census region, metropolitan status, education, and household income. We incorporate these weights into all regression models to improve generalizability to the American public as a whole. The results are substantively similar with or without the use of weights.

To answer Research Question 1, we calculate a range of descriptive statistics for participants’ estimates of status and growth as well as actual status and growth in their districts.

When calculating average treatment effects (Research Questions 2 and 3), we rely on the following general model:

Y_{i} = a + b S_{i} + c G_{i} + d B_{i} + e X_{i} + u_{i},

where $Y_{i}$ is the outcome (local school grades or the relative importance of academic performance) and $X_{i}$ is a vector of demographic characteristics. The default comparison is to the control group, but we also calculate average treatment effects with the status group as the comparison group.

When calculating heterogeneous treatment effects by individual-level and district-level characteristics (Research Questions 4–6), we rely on the following general model:

\begin{array}{l} Y_{i} = a + b S_{i} + c G_{i} + d B_{i} + e X_{i} + f Z_{i d} \\ + g (S_{i} Z_{i d}) + h (G_{i} Z_{i d}) + j (B_{i} Z_{i d}) + u_{i}, \end{array}

where $Z_{i d}$ is the value of the heterogeneous dimension for individual $i$ in district $d$ .²

The treatment effect heterogeneity analyses associated with Research Question 5 are based on the premise that average status and average growth have different underlying relationships with districts’ racial/ethnic and economic compositions. The relationships between student demographics and average growth are much weaker than the analogous relationships between student demographics and average status. To corroborate this finding within our sample, we calculate a series of bivariate relationships between district-level demographic characteristics and academic performance.

Findings

Balance and Missing Data

Table 1 displays the frequencies of participants’ demographic characteristics by experimental condition. Our use of random assignment establishes groups with similar demographic compositions. There are seven instances (out of 78 total comparisons) in which the demographic profile of an experimental group is statistically different from the control group. This rate is marginally higher than we would expect by chance alone. To adjust for these observable differences between groups, we include all individual-level covariates in subsequent analyses.

Table 1

Balance and Missing Data

%	Control(n = 724)	Status(n = 763)	Growth(n = 788)	Both(n = 771)
Female	48.81	52.94	50.94	53.32
White	63.31	65.91	63.03	63.06
Black	12.86	11.22	11.31	12.60
Hispanic	16.82	17.04	14.96	16.73
Other race	7.00	5.83	10.70*	7.60
Less than high school	10.71	12.04	8.15	11.51
High school	29.60	28.23	28.49	28.47
Some college	25.30	28.06	30.02*	29.16
College	34.39	31.67	33.34	30.87
Household income <US$25,000	13.25	13.99	14.74	15.71
Household income US$25,000–US$85,000	46.83	41.57*	41.08*	43.96
Household income >US$85,000	39.92	44.45	44.19	40.33
Took survey in Spanish	5.16	6.69	5.21	5.98
Parent	28.90	28.72	34.40*	32.00
Teacher	3.10	2.78	2.93	2.51
Head of household	80.79	77.56	80.10	79.40
Owns home	70.02	71.77	71.54	69.09
Employed	66.89	61.98*	65.29	65.64
Married	56.52	54.66	58.89	61.09
Northeast	18.97	17.32	18.12	16.72
Midwest	18.93	20.05	24.39*	19.54
South	37.49	37.68	34.36	41.76
West	24.61	24.95	23.14	21.98
Age (years)	47.74	47.89	47.58	47.34
Party ID (1–7)	3.63	3.70	3.71	3.74
Ideology (1–7)	4.03	3.99	4.13	4.06
Missing outcomes	4.80	6.80	4.47	4.15
Missing covariates	0.85	1.43	1.56	1.98

Note. Status, Growth, and Both compared with Control; analyses incorporate survey weights.

p < .05.

Roughly, 4% to 7% of each group is missing answers to one or more of the survey questions that serve as outcomes in our study. If participants do not answer the district status estimation question, the district growth estimation question, the local public school grades question, or the relative importance of academic performance question, they are dropped from the analyses that rely on those values. About 1% to 2% of each group is missing one or more of the demographic covariates. If participants are missing demographic information, we recode the missing data with an arbitrary value and control for an indicator of missingness in subsequent analyses. Thirty-one participants (approximately 1% of the sample) live in areas where we do not have data on district status or growth. Depending on their experimental assignment, these participants are informed that their local schools perform at the 50th percentile in terms of status and/or growth.

Estimating Status and Growth

Our first research question asks about the extent to which participants are able to estimate status and growth in their districts. The first and second plots in Figure 1 display the distributions of participants’ estimates of their district status and district growth percentiles versus the actual district status and district growth percentiles. Participants’ estimates of status and growth percentiles range from 0 to 100. The modal estimate for both status and growth is the 50th percentile. For both forms of academic performance, there is also a second, smaller spike at the 75th percentile. Because the survey features a nationally representative sample, the distributions of actual status and growth percentiles are roughly uniform.

Table 2 describes the distributions of estimated and actual academic performance in greater detail. The average estimated status percentile is 54.05 while the average actual status percentile is 48.46, suggesting that participants are somewhat overoptimistic about their districts’ performance with respect to status. There is considerable variation in participants’ responses. Their status estimates have a standard deviation of 24.97 percentile points (similar to the 29.22 percentile point standard deviation in actual status). Overall, participants’ estimates of district status are related to and moderately predictive of actual district status. The correlation between the two is .29. This relationship is displayed visually in the third plot in Figure 1. In short, participants’ estimates of achievement status reveal a modest understanding of how their districts perform in this regard.

Table 2

Estimating District Status and District Growth (n = 2,928)

Value	Mean	Standard Deviation	Correlation
Estimated status	54.05*	24.97
Actual status	48.46*	29.22
Estimated S − Actual S	5.57*	32.52
Estimated growth	48.66*	25.77
Actual growth	49.10	26.59
Estimated G − Actual G	−0.57	35.95
Estimated S and Actual S			.29*
Estimated G and Actual G			.06*
Estimated S and Estimated G			.79*
Actual S and Actual G			.32*

Note. Estimated Status, Actual Status, Estimated Growth, and Actual Growth compared with 50th percentile; analyses incorporate survey weights.

p < .05.

The pattern with respect to growth is quite different. While the average estimated growth percentile (48.66) and the average actual growth percentile (49.10) are close, this is the product of participants underestimating and overestimating growth in roughly equal proportion. The correlation between the two is .06—essentially a precisely estimated zero. This relationship is displayed visually in the fourth plot in Figure 1. While participants demonstrate some understanding of their districts’ performance in terms of achievement status, they are largely unaware of how their districts perform in terms of achievement growth.

We also explore the extent to which district status and growth—both actual and estimated—are related (see the fifth and sixth plots in Figure 1). Among the participants in our sample, actual status and actual growth are correlated at .32. On average, higher status districts are also higher growth districts, but the relationship is modest. Estimated status and estimated growth, on the contrary, are correlated at a much stronger .79. This mismatch lends itself to two different interpretations. Perhaps participants understand the distinction between status and growth, but they incorrectly believe that a district that is strong on one dimension of academic performance is also overwhelmingly likely to be strong on the other. Alternatively, it is possible that participants simply do not distinguish between the two concepts. They may incorrectly view status and growth as different ways of measuring the same underlying construct. Our analysis is unable to adjudicate between these two possibilities.

The Effects of Academic Performance Information

Table 3 displays the results of the analyses associated with Research Questions 2 to 4, in which we estimate the effects of providing academic performance information on (a) the grades that participants assign to their local public schools and (b) participants’ preferences about how much schools should focus on academic performance.³ Model 1 displays the average effects of providing information about status, growth, or both on participants’ local school grades. Participants in the control group give an average grade of 3.64 on a 5-point scale (roughly a B-) with a standard deviation of 0.91 points. For every experimental group, the receipt of academic performance information is a sobering experience. Compared with the grades in the control group, the grades in the status and growth groups decline on average by 0.28 points and 0.29 points, respectively. The grades in the both group decline by 0.18 points on average. When considering the sample as a whole, the provision of academic performance information of any kind reduces the grades that participants give to their local public schools.

Table 3

The Effects of Status and/or Growth Information

	Local school grades (1–5)				Importance of academic performance (0–100)
	1	2	3	4	5	6	7	8
Status	−0.28* (0.05)		−0.22* (0.09)	−0.30* (0.10)	−2.13* (0.94)		−1.87(1.83)	−6.21* (1.97)
Growth	−0.29* (0.05)	0.00(0.05)	−0.03(0.09)	−0.57* (0.10)	−1.06(0.94)	1.06(0.91)	−4.28* (1.81)	−5.61* (1.99)
Both	−0.18* (0.05)	0.10* (0.05)	−0.28* (0.09)	−0.49* (0.11)	−0.60(0.94)	1.52(0.91)	−1.03(1.84)	−6.07* (2.01)
Control		0.28* (0.05)				2.13* (0.94)
Z			0.14* (0.01)	0.04* (0.01)			−0.09(0.24)	−0.64* (0.27)
Z × S			−0.01(0.02)	0.00(0.02)			−0.05(0.32)	0.83* (0.35)
Z × G			−0.05* (0.02)	0.06* (0.02)			0.67* (0.32)	0.92* (0.36)
Z × B			0.02(0.02)	0.06* (0.02)			0.09(0.32)	1.11* (0.36)
Z =			SPercentile(10s)	GPercentile(10s)			SPercentile(10s)	GPercentile(10s)
Observations	3,012	3,012	3,012	3,012	3,045	3,045	3,045	3,045

Note. Values are WLS coefficients (standard errors in parentheses); analyses incorporate survey weights and include all individual-level covariates. WLS = weighted least squares.

p < .05.

A more complex story emerges when we consider how the effects of academic performance information vary by participants’ local context. Figure 2 displays the relationships between actual district performance and participants’ evaluations of their local public schools, disaggregated by experimental condition (equivalent to Models 3 and 4 in Table 3). The first three plots compare the control group with each of the other experimental groups at every point in the district status distribution. Participants in higher status districts tend to give higher grades, regardless of their experimental assignment. The provision of status information tends to reduce these grades at all points in the district status distribution, but the relationship between district status and participants’ perceptions of school quality is generally unaffected. The provision of both status and growth information generates similar results. By contrast, the provision of growth information alone weakens the relationship between district status and perceptions of school quality. For every 10-percentile-point increase in district status, the grades in the growth group decrease by an additional 0.05 points relative to the control group. In higher status districts (not all of which are also higher growth districts), the negative effect of receiving growth information is large: about half a letter grade on average. In lower status districts (some of which are relatively higher growth districts), there is no effect. This is consistent with the greater responsiveness to negative information identified by James and Moseley (2014). Satisfaction tends to decline with the receipt of bad performance information, but there is no analogous positive effect of good performance information. This pattern of results suggests that participants in the growth group incorporate the new information into their evaluations in ways that temper the conventional wisdom that higher status districts are therefore higher quality districts.

The fourth, fifth, and sixth plots in Figure 2 compare the control group with each of the other experimental groups at every point in the district growth distribution. In the control group, there is a weak positive relationship between district growth and participants’ grades, indicating that these evaluations only loosely reflect the variation in district growth. Given participants’ unfamiliarity with growth, the presence of any relationship between district growth and perceptions of school quality is likely due to the fact that district growth is also correlated with district status (when regressing grades on both status and growth among participants in the control group, only the relationship between status and grades is significant). The provision of growth information—alone or in combination with status information—strengthens the relationship between district growth and participants’ grades considerably. For every 10-percentile-point increase in district growth, the grades in the growth group and the both group increase by an additional 0.06 points relative to the control group. In line with the expectation that negative news carries more weight, the provision of growth information produces downward effects on perceptions of school quality for participants living in lower growth districts and no effects for participants living in higher growth districts. This pattern of results is also consistent with the finding that participants know less about their districts’ performance in terms of growth than in terms of status. As a result, the effects of growth information vary based on the news—positive or negative—that it contains.

We also explore whether the provision of status and/or growth information affects participants’ views about how much schools should focus on academic performance relative to other educational objectives (in this case: students’ social and emotional well-being). Model 5 of Table 3 displays the average effects for the status group, the growth group, and the both group. On average, participants in the control group suggest that schools should place about 65% of their focus on academic performance with a standard deviation of about 19 percentage points. Receiving status information reduces the relative importance of academic performance by 2.13 percentage points. Participants in the growth group and the both group also say that schools should focus less on academic outcomes than their peers in the control group, but these differences are not statistically significant.

Figure 3 displays the relationships between actual district performance and participants’ attitudes about how much schools should focus on academic outcomes, disaggregated by experimental condition (equivalent to Models 7 and 8 in Table 3). The first three plots compare the control group with each of the other experimental groups at every point in the district status distribution. In the control group, there is no relationship between district status and participants’ educational priorities. Participants in both lower status and higher status districts come to the same general conclusion: Schools ought to spend about two thirds of their time and resources on academic matters. The provision of status information—alone or in combination with growth information—tends to reduce participants’ emphasis on academic performance slightly, but the negative effect is constant across the entire district status distribution. However, the effect of growth information alone varies by district status. For every 10-percentile-point increase in district status, the importance of academic performance in the growth group increases by an additional 0.67 percentage points relative to the control group. In lower status districts, the provision of growth information (which may contain unexpected “good news” for some) prompts participants to indicate that schools could focus a little less on academic outcomes. Meanwhile, in higher status districts, the provision of growth information (which may contain unexpected “bad news” for some) shifts the importance of academic performance upward.

The fourth, fifth, and sixth plots in Figure 3 compare the control group with each of the other experimental groups at every point in the district growth distribution. In the control group, there is a clear negative relationship between district growth and participants’ educational priorities. These attitudes are remarkably intuitive considering the absence of growth information. Participants in lower growth districts want schools to focus more on students’ academic development, while their peers in higher growth districts suggest that schools should focus a little more on students’ social and emotional well-being. The provision of academic performance information of any kind appears to reverse this relationship. For every 10-percentile-point increase in district growth, the importance of academic performance increases by an additional 0.83 percentage points (the status group), 0.92 percentage points (the growth group), and 1.11 percentage points (the both group), relative to the control group. Among those in lower growth districts, the provision of academic performance information—either status, growth, or both—causes participants to assign less importance to students’ academic development. The reverse is true in higher growth districts. Our research design does not offer insight into the mechanism for these somewhat counterintuitive results. It may be the case that participants in lower growth districts are more skeptical of the measures of academic performance provided by our experiment, and they respond by placing relatively more emphasis on students’ social and emotional well-being.

Heterogeneous Effects by District Racial/Ethnic and Economic Composition

Next, we present the results of the analyses associated with our secondary research questions. Given the increasing number of statistical tests that accompany each additional research question, the reader should view the following results as exploratory.

The first of our two secondary research questions asks whether the effects described above vary by the racial/ethnic and economic composition of participants’ districts. This question is based on the premise that district-level student racial/ethnic composition and district-level student economic composition have different underlying relationships with average status and average growth. Figure 4 displays these relationships for the participants in our sample. There is a strong, positive relationship between the percentage of White students and the district status percentile (r = .62). The analogous relationship is much weaker with respect to the district growth percentile (r = .18). Similarly, the relationship between the percentage of FRPL-eligible students and the district status percentile (r = −.87) is much stronger than the analogous relationship with the district growth percentile (r = −.35). Based on the differences in these underlying relationships, we speculated that the effects of the provision of academic performance information might vary for participants living in districts with different racial/ethnic and economic compositions.

Models 1 and 2 of Table 4 display the effects of academic performance information on local school grades as they vary by the percentage of White students and the percentage of FRPL-eligible students in participants’ districts. We do not observe evidence of treatment effect heterogeneity by the percentage of White students. However, the effects of growth information vary by the percentage of FRPL-eligible students. In the control group, the relationship between the percentage of FRPL-eligible students and perceptions of school quality is sharply negative: Participants in less affluent district tend to give lower grades to their local schools. In the growth group, this relationship is weakened. For every 10-percentage-point increase in FRPL-eligible students, the grades in the growth group increased by an additional 0.06 points relative to the control group. In short, upon receiving growth information, participants’ perceptions of school quality are less likely to be a simple function of the affluence of a community.

Table 4

Heterogeneous Treatment Effects

	Local school grades (1–5)				Importance of academic performance (0–100)
	1	2	3	4	5	6	7	8
Status	−0.27* (0.10)	−0.45* (0.12)	−0.29* (0.05)	−0.28* (0.05)	−3.27(1.92)	−1.72(2.39)	−1.38(0.93)	−1.14(0.92)
Growth	−0.12(0.10)	−0.58* (0.12)	−0.30* (0.05)	−0.30* (0.05)	−6.77* (1.89)	4.23(2.38)	−0.40(0.93)	−0.66(0.91)
Both	−0.12(0.10)	−0.13(0.12)	−0.16* (0.05)	−0.18* (0.05)	−1.28(1.91)	0.83(2.39)	−0.46(0.93)	−0.54(0.91)
Z	0.09* (0.01)	−0.19* (0.02)	0.02(0.01)	0.07* (0.01)	−0.52* (0.24)	0.28(0.33)	0.23(0.21)	0.46* (0.19)
Z × S	0.00(0.02)	0.03(0.02)	0.01(0.02)	0.00(0.01)	0.24(0.32)	−0.05(0.44)	0.04(0.28)	−0.41(0.25)
Z × G	−0.03(0.02)	0.06* (0.02)	0.01(0.02)	−0.05* (0.01)	1.06* (0.31)	−1.03* (0.43)	−0.53(0.29)	−0.47(0.26)
Z × B	−0.01(0.02)	−0.01(0.02)	−0.04* (0.02)	−0.04* (0.01)	0.14(0.31)	−0.28(0.44)	−0.02(0.29)	−0.75* (0.25)
Z =	% White(10s)	% FRPL(10s)	Est. S – Act. S(10s)	Est. G – Act. G(10s)	% White(10s)	% FRPL(10s)	Est. S – Act. S(10s)	Est. G – Act. G(10s)
Observations	3,003	3,003	2,934	2,931	3,036	3,036	2,943	2,939

Note. Values are WLS coefficients (standard errors in parentheses); analyses incorporate survey weights and include all individual-level covariates. FRPL = free and reduced-price lunch; WLS = weighted least squares.

p < .05.

Models 5 and 6 of Table 4 display the effects of status and/or growth information on the relative importance of academic performance as they vary by the percentage of White students and the percentage of FRPL-eligible students in participants’ districts. In the control group, the relationship between the percentage of White students and the relative importance of academic performance is negative, and the analogous relationship with the percentage of FRPL-eligible students is positive. In other words, participants in Whiter and more affluent districts tend to say that schools should focus a little less on academic outcomes and a little more on students’ social and emotional well-being than their peers in districts that serve more low-income students and students of color. The provision of growth information reverses these relationships. For every 10-percentage-point increase in the percentage of White students, the importance of academic performance in the growth group increases by an additional 1.06 percentage points relative to the control group. For every 10-percentage-point increase in FRPL-eligible students, the importance of academic performance in the growth group decreases by an additional 1.03 percentage points relative to the control group. The provision of growth information tends to prompt participants in richer, Whiter districts to place more emphasis on students’ academic development. At the same time, the provision of growth information prompts participants in less affluent, more diverse districts to place more emphasis on students’ social and emotional well-being.

Updating Prior Beliefs About Academic Performance

The final research question asks whether these effects could be the result of participants updating their prior beliefs about academic performance in their districts. To answer this question, we test whether the effects vary by the extent to which participants incorrectly estimate academic performance in their districts. Model 3 of Table 4 displays the effects of academic performance information on local school grades as they vary by the amount that participants overestimate their districts’ status percentiles. The only evidence for differential updating with respect to status appears among participants in the both group. For every 10-percentile-point overestimation of district status, the grades in the both group decrease by an additional 0.04 points relative to the control group. The evidence for differential updating with respect to growth—about which participants’ prior beliefs are less well informed—is more robust. Model 4 displays the effects of academic performance information on local school grades as they vary by the amount that participants overestimate their districts’ growth percentiles. For every 10-percentile-point overestimation of district growth, the grades in the growth group and the both group decrease by an additional 0.05 points and 0.04 points relative to the control group, respectively. In short, as the overestimation of district growth increases, so does the negative effect of growth information on local school grades.

Model 7 of Table 4 displays the effects of status and/or growth information on the relative importance of academic performance as they vary by the amount that participants overestimate their districts’ status percentiles. We do not observe evidence of treatment effect heterogeneity along this dimension. Alternatively, Model 8 indicates that, for every 10-percentile-point overestimation of district growth, the importance of academic performance in the both group decreases by an additional 0.75 percentage points relative to the control group. In other words, as the overestimation of district growth increases, so does the negative effect of receiving both forms of academic performance information on the importance that they assign to students’ academic development.

Conclusion

States increasingly include measures of both achievement status and achievement growth in their school accountability systems. States and school districts use this information to guide their efforts to support struggling students, redirect resources where they are most needed, and even shut down chronically underperforming schools. Families also use this information—either via states’ official school report cards or via secondary sources, such as GreatSchools.org or Niche.com, that draw on state data—as they make school and housing decisions.

Previous scholarship suggests that the American public already possesses a modest understanding of how their local schools perform in terms of achievement status. We observe the same pattern in our own analysis. However, we also find that Americans are largely unfamiliar with how their local schools perform in terms of growth. This should not be altogether surprising, given the fact that many states have only recently incorporated growth data into their school accountability systems. To understand how the public responds to these new measures of educational performance, we conducted an online survey experiment with a nationally representative sample that identifies the effects of disseminating status and/or growth information on participants’ perceptions of school quality.

Because of Americans’ existing familiarity with achievement status in their local schools, the provision of status information does not fundamentally change the underlying relationship between district status and the public’s perceptions of school quality. The provision of growth information, however, alters Americans’ views about educational performance. The effects of growth information are quite different for participants living in lower growth districts (significantly reducing the grades that participants assign to their local schools) and higher growth districts (no effect). Consequently, the provision of growth information strengthens the underlying relationship between district growth and the public’s perceptions of school quality. In short, when participants learn about student growth, their personal evaluations of their local schools become more in line with a measure that many researchers consider a better measure of schools’ contributions to student learning. Moreover, because district growth bears a weaker relationship to the economic composition of the student body than district status, the provision of growth information reorients Americans’ perceptions of school quality away from the conventional wisdom that more affluent school districts are almost always higher quality districts.

This pattern of results does not necessarily imply that the provision of growth information improves participants’ conceptual understanding of the causal effects of schools on student learning. We did not ask participants to describe what status and growth meant to them, nor did we ask participants to explain how they arrived at their appraisals of their local public schools. Our results indicate that participants who receive growth information placed at least some weight on it when evaluating school quality, but we cannot ascertain whether it changed their beliefs about the validity of growth as a measure of school effectiveness.

The increasing prevalence of student growth information may have some unanticipated consequences. We also asked participants to opine on how much schools should focus on academic performance relative to other educational objectives (in this case, students’ social and emotional well-being). We find that the provision of growth information to participants living in lower growth districts not only lowers their perceptions of school quality, it also causes them to say that their schools should focus less on academic outcomes. It may be the case that distributing information about lackluster student growth will not, as one might expect, induce a call-to-arms for a greater emphasis on academic goals. Rather, it may end up reducing support for schools’ academic objectives and/or this particular method of measuring success toward those objectives.

These results have important implications for our understanding of the social, economic, and political consequences of how we measure educational performance. While the vast majority of states now include or plan to include growth in their school accountability systems, there is a great deal of variation in their approaches to measuring growth, the weight they assign to it in school and district ratings, and the accessibility of this information to the public. The results of our analysis help us think through some of the potential consequences of these policy choices. As states begin to rely less on traditional indicators of achievement status when evaluating school quality, we might expect to see heightened demand for higher growth schools and nearby housing. Because average growth bears a weaker relationship to student demographics than average status, this shift could alternatively benefit many low-income communities and communities of color (by increasing housing values for existing homeowners) or further disadvantage them (by attracting relatively affluent newcomers who can afford higher housing costs). With metrics that better reflect schools’ contributions to student learning—and not merely the characteristics of the students they serve—districts and states may be able to allocate resources more efficiently and offer more targeted interventions to underperforming schools. We might also expect to see political advantages accrue to school board members and other elected officials who preside over periods of high growth.

The COVID-19 pandemic and subsequent school closures have created new complications for the measurement of academic performance. Annual, standardized tests are necessary to calculate both status and growth for students, schools, and districts. In Spring 2020, the U.S. Department of Education allowed states to forgo testing elementary and secondary students for a year, and, at the time of writing, discussions are underway to determine whether to reinstate these tests in 2021 (Gewertz, 2020; Strauss, 2020). Some researchers suggest that it may be possible to skip a year of testing and calculate student-level growth over a 2-year period (Betebenner & Van Iwaarden, 2020; Fazlul et al., 2021). If 2 years without testing elapse, this approach becomes less tenable. Even if testing resumes in Spring and/or Fall 2021, estimates of either status or growth derived from these data may not be directly comparable with previous years. During such a crisis, the composition of test-takers can change in unpredictable ways, and the students who are present on test day may face atypical physical, psychological, and environmental obstacles that can introduce additional noise into the data (Klugman & Ho, 2020). Both educators and the public should apply caution when making decisions based on estimates of academic performance using data collected during the pandemic.

We see a variety of important avenues for future research stemming from this work. First, our estimates of average district status and average district growth obscure considerable variation in both measures. It may be the case that individuals are particularly sensitive to information about the status and growth of local students who share their racial, ethnic, and/or socioeconomic identity. Moreover, evidence of large status-based or growth-based inequalities between students—or the lack thereof—may have a significant influence on perceptions of school quality. Similar survey-based information experiments with more fine-grained treatments could help illuminate this issue.

Additional research is also necessary to understand the extent to which individuals comprehend and trust new measures of academic performance such as growth. We examine the effects of distributing status and growth information on perceptions of school quality, but our analysis does not capture the nuances of how participants make sense of these data. Some participants may view growth as a superior measure of schools’ contributions to student learning, but they may also consider status to be an important element of school quality (e.g., the value of high-performing peers). Other participants may have little interest in the underlying constructs that either metric attempts to measure, but care deeply about their districts’ rankings. Still others may be skeptical of any metric derived from standardized test scores. Both quantitative and qualitative research approaches would be valuable in mapping out the public’s reactions to academic performance information and to identify potential barriers to understanding.

Finally, the variation in states’ school accountability systems offers an opportunity to study how these policy choices may affect behavioral outcomes such as families’ school enrollment decisions, housing prices, and political outcomes in school board elections and local school funding referenda. The staggered rollout of growth measures—both in official state data portals and via popular school rating websites—may make it possible for researchers to identify the effects of disseminating this information. Our analysis suggests that these policy choices can alter the public’s attitudes toward the public schools, but much work remains to be done to understand whether these revised attitudes translate into different behaviors.

Supplemental Material

sj-docx-1-epa-10.3102_01623737211030505 – Supplemental material for Status, Growth, and Perceptions of School Quality

Supplemental material, sj-docx-1-epa-10.3102_01623737211030505 for Status, Growth, and Perceptions of School Quality by David M. Houston, Michael Henderson, Paul E. Peterson and Martin R. West in Educational Evaluation and Policy Analysis

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Authors

DAVID M. HOUSTON is an assistant professor of education policy at George Mason University. He studies education policy and politics.

MICHAEL HENDERSON is an associate professor of political communication in the Manship School of Mass Communication at Louisiana State University. His research areas include public opinion, media, and public policy.

PAUL E. PETERSON is the Henry Lee Shattuck Professor of Government at Harvard University. His research focuses on education policy, federalism, and urban policy.

MARTIN R. WEST is the Henry Lee Shattuck Professor of Education at the Harvard Graduate School of Education and a faculty research fellow of the National Bureau of Economic Research. He studies K–12 education policy and politics in the United States.

References

Abdulkadiroglu

Pathak

P. A.

Schellenberg

Walters

C. R.

(2020). Do parents value school effectiveness? American Economic Review, 110(5), 1502–1539.

Achieve, Inc. (2019). Accountability in state ESSA plans [Tracking tool]. https://www.achieve.org/accountability-in-essa

Ainsworth

Dehejia

Pop-Eleches

Urqiola

(2020). Information, preferences, and household demand for school value added (NBER Working Paper 28267). https://www.nber.org/papers/w28267

Alabama Department of Education. (2021). Education report card. https://reportcard.alsde.edu/alsde/selectschool

Baekgaard

Serritzlew

(2015). Interpreting performance information: Motivated reasoning or unbiased comprehension. Public Administration Review, 76(1), 73–82.

Bali

(2016). Evolving trends in public opinion on the quality of local schools. Educational Policy, 30(5), 688–720.

Barnum

(2020, September 24). GreatSchools overhauls ratings in bid to reduce link with race and poverty. Chalkbeat. https://www.chalkbeat.org/2020/9/24/21453357/greatschools-overhauls-ratings-reduce-link-race-poverty

Barone

(2017). What ESSA says: Continuities and departures. In Hess

F. M.

Eden

(Eds.), The Every Student Succeeds Act: What it means for schools, systems, and states (pp. 59–73). Harvard Education Press.

Barrows

Henderson

Peterson

P. E.

West

M. R.

(2016). Relative performance information and perceptions of public service quality: Evidence from American school districts. Journal of Public Administration Research and Theory, 26(3), 571–583.

10.

Bayer

Ferreira

McMillan

(2007). A unified framework for measuring preferences for schools and neighborhoods. Journal of Political Economy, 115(4), 588–638.

11.

Berry

C. R.

Howell

W. G.

(2007). Accountability and local elections: Rethinking retrospective voting. Journal of Politics, 69(3), 844–858.

12.

Betebenner

Van Iwaarden

(2020). Issues and considerations that the COVID-19 pandemic presents for measuring student growth. Center for Assessment. https://www.nciea.org/blog/sgp/issues-and-considerations-covid-19-pandemic-presents-measuring-student-growth

13.

Beuermann

Jackson

C. K.

Navarro-Sola

Pardo

(2020). What is a good school, and can parents tell? Evidence on the multidimensionality of school output (NBER Working Paper 25342). https://www.nber.org/papers/w25342

14.

Black

S. E.

(1999). Do better schools matter? Parental valuation of elementary education. Quarterly Journal of Economics, 114(2), 577–599.

15.

Castellano

K. E.

A. D.

(2013). A practitioner’s guide to growth models. Council of Chief State School Officers.

16.

Chingos

M. M.

Henderson

West

M. R.

(2012). Citizen perceptions of government service quality: Evidence from public schools. Quarterly Journal of Political Science, 7(4), 411–445.

17.

Chingos

M. M.

West

M. R.

(2015, January 20). Why annual statewide testing is critical to judging school quality. Brookings Institution. https://www.brookings.edu/research/why-annual-statewide-testing-is-critical-to-judging-school-quality/

18.

Clinton

J. D.

Grissom

J. A.

(2015). Public information, public learning, and public opinion: Democratic accountability in education policy. Journal of Public Policy, 35(3), 355–385.

19.

Corcoran

Jennings

Cohodes

Sattin-Bajaj

(2018). Leveling the playing field for high school choice: Results from a field experiment of informational interventions (NBER Working Paper 24471). https://www.nber.org/papers/w24471

20.

Data Quality Campaign. (2019). Growth data: It matters, and it’s complicated. https://dataqualitycampaign.org/resource/growth-data-it-matters-and-its-complicated/

21.

Data Quality Campaign. (2020). Show me the data: There is no finish line for report cards. https://dataqualitycampaign.org/wp-content/uploads/2020/10/2020-DQC-Show-Me-the-Data.pdf

22.

Fahle

E. M.

Shear

B. R.

Kalogrides

Reardon

S. R.

DiSalvo

A. D.

(2018). Stanford Education Data Archive: Technical documentation (Version 2.1). Stanford University Center for Education Policy Analysis.

23.

Fazlul

Koedel

Parsons

Qian

(2021). Bridging the COVID divide: How states can measure student achievement growth in the absence of 2020 test scores. Thomas B. Fordham Institute. https://fordhaminstitute.org/national/research/bridging-covid-divide-how-states-can-measure-student-achievement-growth-absence

24.

Figlio

D. N.

Lucas

M. E.

(2004). What’s in a grade? School report cards and the housing market. American Economic Review, 94(3), 591–604.

25.

Fiva

J. H.

Kirkebøen

L. J.

(2011). Information shocks and the dynamics of the housing market. Scandinavian Journal of Economics, 113(3), 525–552.

26.

Gewertz

(2020, April 2). It’s official: All states have been excused from statewide testing this year. Education Week. https://www.edweek.org/teaching-learning/its-official-all-states-have-been-excused-from-statewide-testing-this-year/2020/04

27.

Glazerman

Dotter

(2017). Market signals: Evidence on the determinants and consequences of school choice from a citywide lottery. Educational Evaluation and Policy Analysis, 39(4), 593–619.

28.

Glazerman

Nichols-Barrer

Valant

Chandler

Burnett

(2020). The choice architecture of school choice websites. Journal of Research on Educational Effectiveness, 13(2), 322–350.

29.

Harris

D. N.

Larsen

M. F.

(2015). What do families want (and why)? New Orleans families and their school choices before and after Katrina. Education Research Alliance for New Orleans.

30.

Hastings

J. S.

Weinstein

J. M.

(2008). Information, school choice, and academic achievement: Evidence from two experiments. Quarterly Journal of Economics, 123(4), 1373–1414.

31.

Holbein

(2016). Left behind? Citizen responsiveness to government performance information. American Political Science Review, 110(2), 353–368.

32.

Houston

D. M.

Henig

J. R.

(2021). The effects of student growth data on school district choice: Evidence from a survey experiment. American Journal of Education. Advance online publication.

33.

Imberman

S. A.

Lovenheim

M. F.

(2016). Does the market value value-added? Evidence from housing prices after the public release of school and teacher value-added. Journal of Urban Economics, 91(1), 104–121.

34.

Iowa Department of Education. (2021). Iowa school performance profiles. https://www.iaschoolperformance.gov/ECP/Home/Index

35.

Jacobsen

(2009). The voice of the people in education. In Sykes

Schneider

Plank

D. N.

(Eds.), Handbook of education policy research (pp. 307–318). Routledge.

36.

Jacobsen

Saultz

Snyder

J. W.

(2013). When accountability strategies collide: Do policy changes that raise accountability standards also erode public satisfaction? Educational Policy, 27(2), 360–389.

37.

Jacobsen

Snyder

J. W.

Saultz

(2014). Informing or shaping public opinion? The influence of school accountability data format on public perceptions of school quality. American Journal of Education, 121(1), 1–27.

38.

Jacobsen

Snyder

J. W.

Saultz

(2015). Understanding satisfaction with schools: The role of expectations. Journal of Public Administration Research and Theory, 25(3), 831–848.

39.

James

(2011). Performance measures and democracy: Information effects on citizens in field and laboratory experiments. Journal of Public Administration Research and Theory, 21(3), 399–418.

40.

James

Moseley

(2014). Does performance information about public services affect citizens’ perceptions, satisfactions, and voice behavior? Field experiments with absolute and relative performance information. Public Administration, 92(2), 493–511.

41.

James

Van Ryzin

G. G.

(2017). Incredibly good performance: An experimental study of source and level effects on the credibility of government. American Review of Public Administration, 47(1), 23–35.

42.

Kane

T. J.

Cellini

S. R.

Staiger

D. O.

(2006). School quality, neighborhoods and housing prices. American Law and Economics Review, 8(2), 183–212.

43.

Klugman

E. M.

A. D.

(2020). How can released state test items support interim assessment purposes in an educational crisis? Educational Measurement: Issues and Practice, 39(3), 65–69.

44.

Kogan

Lavertu

Peskowitz

(2016). Performance federalism and local democracy: Theory and evidence from school tax referenda. American Journal of Political Science, 60(2), 418–435.

45.

Ladd

Loeb

(2013). The challenges of measuring school quality: Implications for educational equity. In Reich

Allen

(Eds.), Education, justice, and democracy (pp. 22–55). Chicago University Press.

46.

Loveless

(1997). The structure of public confidence in education. American Journal of Education, 105(2), 127–159.

47.

Marvel

J. D.

(2016). Unconscious bias in citizens’ evaluations of public sector performance. Journal of Public Administration Research and Theory, 26(1), 143–158.

48.

Payson

J. A.

(2016). When are local incumbents held accountable for government performance? Evidence from US school districts. Legislative Studies Quarterly, 42(3), 421–448.

49.

Reardon

S. F.

(2019). Educational opportunity in early and middle childhood: Using full population administrative data to study variation by place and age. RSF: The Russell Sage Foundation Journal of the Social Sciences, 5(2), 40–68.

50.

Reardon

S. F.

Papay

J. P.

Kilbride

Strunk

K. O.

Cowen

Donohue

(2019). Can repeated aggregate cross-sectional data be used to measure average student learning rates? A validation study of learning rate measures in the Stanford Education Data Archive (Working Paper No. 19-08). Center for Education Policy Analysis, Stanford University.

51.

Rothstein

Jacobsen

Wilder

(2008). Grading education: Getting accountability right. Teachers College Press.

52.

Schneider

Jacobsen

White

R. S.

Gehlbach

(2018). The (mis)measure of schools: How data affect stakeholder knowledge and perceptions of quality. Teachers College Record, 120(6), 1–40.

53.

Schueler

West

M. R.

(2016). Sticker shock: How information affects citizen support for public school funding. Public Opinion Quarterly, 80(1), 90–113.

54.

Sherman

S. J.

Mackie

D. M.

Driscoll

D. M.

(1990). Priming and the differential use of dimensions in evaluation. Personality and Social Psychology Bulletin, 16(3), 405–418.

55.

Stiefel

Schwartz

A. E.

Rotenberg

(2011). What do AEFA Members say? Summary of results of an education finance and policy survey. Education Finance and Policy, 6(2), 267–292.

56.

Strauss

(2020, December 31). Calls are growing for Biden to do what DeVos did: Let states skip annual standardized tests this spring. The Washington Post. https://www.washingtonpost.com/education/2020/12/30/calls-are-growing-biden-do-what-devos-did-let-states-skip-annual-standardized-tests-this-spring/

57.

Valant

Weixler

L. B.

(2020). Informing school-closing families about their options: A field experiment from New Orleans. Education Research Alliance for New Orleans.

58.

Walker

S. E.

Archbold

C. A.

(2014). The new world of police accountability (2nd ed.). SAGE.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.05 MB