Education Policy Center
Dynamic Participation in Inter-disrict Open Enrollment:
Evidence from Michigan 2005-2013
Joshua M. Cowen, Ph.D.+
College of Education
Michigan State University
Benjamin Creed
College of Education
Michigan State University
Venessa A. Keesler, Ph.D.
Michigan Department of Education
July 2015
Author Note:
Corresponding author: jcowen@msu.edu; funding for this project provided in part by the Walton Family Foundation (Cowen) and the Institute for Education Science at the U.S Department of Education (Creed). We are grateful for the assistance with data access and data questions provided by the Michigan Department of Education, in particular Erika Bolig and Joseph Martineau, and Melissa Bisson at the Michigan Center for Educational Performance and Information. We are also grateful to Christina Mazuca Ebmeyer, Barbara Schneider, and the Educational Policy Center at Michigan State University for ongoing internal support for this project.
Abstract
This paper considers the determinants of student participation in Michigan’s large inter-district open enrollment system, known as Schools of Choice. Employing a rich dataset from the Michigan Department of Education, we examine the population of public school students in resident and non-resident school districts between the 2005-06 and 2012-13 academic years. We find substantial evidence that historically disadvantaged students are those most likely to participate in Schools of Choice: African American students and low-income students are more likely than their peers in their resident districts to make an inter-district transfer; they are, however, also the most likely among other Schools of Choice participants to exit the program. In addition, students who are relatively low performing on the state’s standardized exam—especially in mathematics—are most likely to both participate in Schools of Choice and, among participants, the most likely to exit. We conclude by noting that these patterns are similar to those found among the determinants of more general forms of student mobility.
Introduction
Although charter schools and, in recent years, vouchers for private school tuition are highly visible alternatives to traditional public schooling, open enrollment programs represent an important form of school choice policy across the United States. Twenty-one states require schools to accept students from outside their home districts (inter-district open enrollment) under certain circumstances, while 23 require districts to allow students to choose within their boundaries (intra-district open enrollment) (Mikulecky, 2013). In other states, open enrollment is voluntary, but in only Alabama and Maryland are both intra-district and inter-district choice entirely non-existent (National Center for Education Statistics, 2013). Since 2002, federal policy under the No Child Left Behind Act began to require open enrollment options within districts for students in schools failing repeatedly to make Adequate Yearly Progress. The law also expanded public school choice for students with certain physical, mental or emotional needs (Cowen & Toma, 2015).
Within these broad national parameters, considerable variation exists at the state level. Districts with financial or capacity-related barriers to accepting new students are often exempt. In many cases, the rules governing entry into high-demand schools are set at local levels, often in the form of lotteries for over-subscribed schools or means-tests prioritizing students with socioeconomic or academic disadvantages. Historically, open enrollment and particularly magnet school programs have been used to address racial segregation (Schneider, Schiller, & Coleman, 1996), and in the contemporary context some states issue explicit guidelines intended to maintain racial or ethnic balances. Transportation costs associated with open enrollment are borne by individual districts, or shared by multiple districts trading students, while in others parents are responsible for transportation across catchment areas or district boundaries. Finally, the implications of these plans for district finances varies widely (Cowen and Toma, 2015; Mikulecky, 2013).
Despite the prevalence of open enrollment plans, however diverse, research on these policies is scant relative to the extensive scholarship devoted to charter schools. A handful of studies have been devoted to considering the determinants of student transfer under open enrollment-type programs, either by comparing sending and receiving districts (e.g. Reback 2008; Welsch, Statz, and Skidmore 2010; Carlson, Lavery, and Witte 2011) or by examining the characteristics of individual students themselves (e.g. Phillips, Hausman, and Larson 2012; Lavery and Carlson 2014; Bifulco, Cobb, and Bell 2009). Several studies have considered the impacts of public school choice on student achievement, (Cullen, Jacob and Levitt 2005, 2006; Hastings and Weinstein 2007; Bifulco, Cobb and Bell 2009; Deming, et al. 2011; Welsch and Zimmer 2012; Carlson, Lavery and Hughes 2014), although of these the majority are focused entirely on transfers within rather than between school districts. Finally, a number of other studies have considered the more general effects of student transfers and student mobility (e.g. Alexander, Entwisle and Dauber 1996; Hanushek, Kain and Rivkin 2004; Ingersoll, Scamman and Eckerling 1989; Rumberger et al. 1999; South, Haynie and Bose 2007; Temple and Reynolds 1999; Xu, Hannaway and D’Souza 2009)
The present paper contributes to this relatively underdeveloped literature in the school choice field by considering the characteristics of individual students who opt to leave their resident districts for a public school elsewhere through an open enrollment policy. Drawing on a rich set of administrative data on students in Michigan between 2005 and 2013, we estimate models predicting which individuals participate in the state’s large-scale open enrollment program. We find consistent evidence that participation in Michigan’s Schools of Choice program is dynamic, with similar student demographic and academic characteristics predicting both entry into and exit from the program. In particular, African Americans are more likely than their peers in the same district to exit the district using Schools of Choice, as are students participating in free/reduced lunch and students with lower state test scores. These students are also the most likely to exit the program, however. Prior school characteristics appear to play a weaker role in determining student entry into Schools of Choice, but a number of demographic and academic characteristics at the school level are related to student exit. We conclude by noting that these patterns are comparable to those found in studies of broader forms of student mobility: in particular, we note that the evidence here suggests that one practical result of Michigan’s Schools of Choice program is to open district borders to disadvantaged students who are likely more mobile in the first place.
Background
Previous Research: Who Chooses, Where Do They Go, and What are the Effects?
As in studies of charter or private school choice, one of the critical questions for research on open enrollment programs focuses on the characteristics of students who participate. Among the studies employing student-level data, race, academic background and socioeconomic background have all appeared among the determinants of intra or inter-district choice. Studies of Colorado’s program have indicated that on some measures, open enrollment students are more advantaged—they are less likely to be eligible for free/reduced lunch, designated for special educational services, and English Language Learning, have higher reading achievement scores prior to transferring (Carlson, Lavery and Hughes 2014; Lavery and Carlson 2014). On the other hand, prior math achievement may be lower for open enrollment participants (Lavery and Carlson 2014) and race plays differential roles depending on student grade: Hispanics are disproportionately less likely to transfer, but African Americans in younger grades are more likely to transfer than white students (Lavery and Carlson 2014). Descriptive statistics from one large district in the same region have found less ambiguity, with white students and those not eligible for free/reduced lunch more likely to participate (Phillips, Hausman and Larsen 2012).
Among studies focusing on district-level differences in open enrollment participation in either Colorado, Minnesota or Wisconsin, the evidence has indicated that districts with higher levels of income are considerably more likely to have high rates of choice (Holme and Richards 2009; Welsch, Statz and Skidmore 2010; Carlson, Lavery and Witte 2011; Lavery and Carlson 2014), and that higher levels of achievement may also be associated with higher levels of public school choice. A handful of district-level studies have also considered where students who opt into open enrollment actually attend. Generally, these studies have found that schools and districts receiving transfer students tended to be better advantaged, as measured by higher test scores, higher levels of resident income or housing value, or in some cases lower levels of non-white minority students (Fossey 1994; Armore and Peiser 1998; Reback 2008; Carlson, Lavery and Witte 2011) Taken as a whole, because there is evidence that districts with higher levels of achievement send students to districts with even higher student outcomes, and because students appear to be choosing within and out of comparably high income districts, the available evidence indicates that open enrollment programs are utilized primarily in areas where socioeconomic and academic advantages are already substantial (Carlson, Lavery and Witte 2011).
The effects of public school choice have also been generally considered in a variety of geographic and programmatic settings. Evidence from Chicago’s lottery-based intra-district open enrollment system has shown positive impacts of choice (Cullen, Jacob and Levitt 2005; 2006), a pattern consistent with data from Charlotte, North Carolina (Hastings and Weinstein 2007; Deming, et al. 2011) and evidence from magnet school lotteries in Connecticut (Bilfulco, Cobb and Bell 2009). Non-experimental evidence from Colorado has been mixed, with Carlson, Lavery and Hughes (2014) finding no immediate effects of transferring but, for students who maintained participation over several years, eventual gains to achievement, particularly in reading. There is also district-level evidence from Wisconsin that indicates indirect effects of open enrollment on outcomes for students remaining in districts with high enrollment losses: these districts may be responding to competitive pressures to maintain enrollment through improved performance (Welsch and Zimmer 2012).
Finally, it is important to note that open enrollment programs are not the only source of student transfer within the public school system. A larger literature on student mobility has consistently found student transfers outside of formal choice programs (or, at least, not explicitly accounting for the existence of open enrollment) are a phenomenon associated with substantial socioeconomic disadvantage. As much as half of all urban students may transfer schools for “non-structural” reasons (i.e. reasons other than meeting a school’s terminal grade) within three years (Kerbow, Azcoitia and Buell 2003), including many who transfer during the school year (Alexander, Entwisle and Dauber 1996). Such numbers may be as high as 75 percent of students across the first and twelfth grades (Rumberger, et al. 1999). All else equal, race, academic ability, income and family background appear related to student mobility and, with few exceptions, the literature has consistently demonstrated negative academic consequences for students who move and students in classrooms with high rates of turnover (Alexander, Entwisle and Dauber 1996; Hanushek, Kain and Rivkin 2004; Ingersoll, Scamman and Eckerling 1989; Kerbow, Azcoitia and Buell 2003; Rumberger et al. 1999; South, Haynie and Bose 2007; Temple and Reynolds 1999; Xu, Hannaway and D’Souza 2009). As we discuss further below, these patterns may help explain in ways otherwise underappreciated in the broader school choice literature, why substantial variation exists in both the determinants of participation and its effects on outcomes.
Michigan’s Schools of Choice Program
In Michigan, various forms of modern school choice have existed since 1994. The state’s Schools of Choice program established by the state legislature governs inter-district student enrollment. Michigan school districts are permitted but not required to participate in Schools of Choice. (1) Under Section 105 of the State School Aid Act, non-resident parents may choose to enroll their children in a participating local district within the same regionally determined (typically county-based) intermediate school district (ISD) as their resident district; or, under Section 105c, within a different intermediate district altogether if the choice district shares a border with the resident ISD. Each participating district determines specific provisions, including: caps on non-resident enrollment; which schools and grades are eligible to receive non-residents; whether specific academic programs are available to non-residents; and the timeline for applying for enrollment. Although most districts in the state nominally accept non-resident students, the provisions for local control imply substantial statewide variation in the extent to which school choice is realistically available to individual students. (2) Subject to these provisions, individual students may choose to leave their residential district, and in areas with more than one participating district may choose between schools in different districts. In addition to the Schools of Choice program authorized at the state level, neighboring districts may establish local cooperative agreements to permit other forms of student transfers between their borders. For the present paper, we focus exclusively on students participating in inter-district open enrollment via Schools of Choice.
Data and Methods
Data
We utilized the Michigan Department of Education’s (MDE) administrative dataset from the 2005-2006 school year through the 2012-2013 school year. This dataset contains demographic information and enrollment history for the universe of Michigan students. The enrollment history contains information on whether a student participates in Michigan’s School of Choice, attends a Public School Academy (charter school), or attends school in the district of residency in every year. The data set also provides Michigan Educational Assessment Program (MEAP) test scores for students in grades 3 through 8. Each student receives a unique identification code upon entry into Michigan public schools enabling us to follow a student’s enrollment choices across years. In total, we are able create a panel spanning 7 years with 14,346,368 unique student-year observations which exhaustively covers Michigan students attending both traditional public schools and public school academies. Table 1 depicts the number of and percentage of enrolled students in either Section 105 or 105c of Michigan’s Schools of Choice program (hereafter SoC students), as well as the number and percentage of charter students (PSA) for comparison against the total enrollment numbers. As Table 1 indicates, participation in both SoC and in the charter sector has grown steadily over the years for which we have data, while overall grade 3-8 enrollment has declined.
In our primary analyses below, which focus on differences between SoC and traditionally assigned public school students, we exclude charter school students from these comparisons. Future work will consider differences between students who choose SoC options and charter school students. Further, we limit our final sample to exclude those students attending alternative schooling options, such as vocational education schools. After removing duplicate entries for students our final sample consisted of 13,647,204 student-year observations for nearly 3 million individual students in schools serving traditional K-12 students. (3)
Variables
We constructed a series of enrollment dummy variables for each student-year observation which represent our key outcome variables of interest. Using the enrollment codes provided by MDE, we created a series of dummy variables for use of the schools of choice policy: in this paper, we focus on students who enrolled in Schools of Choice under either Section 105 or 105c described above. As we are interested in the switching decisions of students and families, we created two sets of variables to capture whether a student switched into or out of Schools of Choice. This requires comparing students’ enrollment choices in year t-1 to year t, effectively removing the 2005-2006 year of observations when we explore switching decisions. If a student is not in Schools of Choice in t-1 but is flagged as a School of Choice student in t, we count them as switching into schools of choice at time t. Conversely, if they are in schools of choice at t-1 but not at time t we code them as switching out of schools of choice at time t.
The richness of the MDE data set enabled us to develop a series of student, school, and district variables to use in our analysis. At the student level, we created variables from the data capturing student gender, race/ethnicity, grade of attendance, free or reduced price lunch status (FRL), limited English proficiency status (LEP), and special needs status. In addition to these demographic characteristics, we were able to include the MEAP scores for students in grades 3 – 8. We used the MEAP math and reading scores to generate two variables for each subject: the student’s MEAP scores standardized for a given year at the district level, and at the state level.
The school and district level variables were generated by aggregating the individual student-year variables to the respective levels. Thus, we created variables capturing the proportion of various racial/ethnic groups, proportion of FRL students, proportion of LEP students, and proportion of special needs students at the school and district levels. Further, we have the school mean standardized district and state MEAP scores as well as the district mean standardized state MEAP scores. These variables at the school and district level let us explore what, if any, characteristics of schools and districts are associated with students opting in or opting out of Schools of Choice.
Methods
Initial Participation Statistics
As no baseline descriptive information about Michigan’s School of Choice program exists in the literature, we focus considerable attention to a careful descriptive analysis. We begin with a series of bivariate analyses that provide simple comparisons between SoC students and those remaining in their residential schools. There are two ways to construct these comparisons. The first considers students who either are enrolled in SoC or enrolled in their district of residency at any given point in time, t, regardless of when they first entered the program. The other comparison, which we call a switcher analysis, compares students who leave their residential districts after t-1 to enter a new district under SoC at t with students who remain in their residential district at t. Because our modeling strategy below is focused on students who switch into SoC, we similarly focus on switchers for the preliminary bivariate analyses here. In the Appendix to this paper, we present the first such comparisons between SoC and traditionally assigned public school students.
We also consider the characteristics of schools and districts of schools serving the SoC students in our panel. As with the decision to participation in SoC in the first place, there are at least two ways to consider school and district characteristics. The first simply takes a population-level view of the schools and districts attended by SoC students compared to those attended by stayer students. Such a comparison addresses the policy question: “do schools/districts attended by choice students differ, on average, than those attended by stayer students?” A second, more complicated approach contains relative comparisons: a.) between sending and destination schools/districts for each student i or b.) between different potential destinations for each student i. These two sets of comparisons address the policy questions: “how do schools/districts that students select under SoC differ from their resident schools/districts?” and “what characteristics of schools/districts make them more likely to be selected by students as destinations under SoC?”, respectively. These latter two formulations are important in their own right, but speak largely to a school or district-level study focus. In this initial paper, where we are focused largely on individual participants, we simply consider school and district differences by students’ SoC status. We note also that the estimated relationships between the S characteristics in Equation 1 provide an initial analysis of the relationship between students’ school characteristics and the decision to opt out of a residential district in the following year.
Finally, we consider the characteristics of students who exit the Schools of Choice program at some point during their observed careers in the data. Mechanically such analyses in our data are difficult to conduct for all students, since despite the longevity of our panel—the eight academic years between 2005-06 and 2012-13.Only those in early grades in the early years of the panel are at least in principle observed from 2005 through 2012. We are able to simply calculate the average number of years in Schools of Choice observed in our data for all students who switch into the program at some point in our data, but these averages only provide a partial picture of the extent to which School of Choice is a permanent or more transitory feature of a student’s education. In our multivariate framework below, we are able to simply estimate the probability that a student exits School of Choice conditional on participating in a given year. To obtain other descriptive statistics related to the duration of a student’s participation, we necessarily focus on the students who we observe in Schools of Choice at the beginning of their academic careers.
Modeling Switchers – Multivariate Analysis
We model the choice for student i to opt into the School of Choice program in year t as a function of student, school, and district characteristics at time t-1 (Equation 1). O represents the outcome of interest, whether student i, in school s, in district d, at time t opts out of their assigned district through Schools of Choice:
(1) 
The opting out decision in time t is a function of a vector of student characteristics C, and school characteristics S at time t-1, a residential district-by-year fixed effect δ and an idiosyncratic error term (
). The district-by-year fixed effects imply that we are comparing students within the same resident district in the same year on the basis of whether they choose to leave under Schools of Choice or remain in that district. This is important not only to net out local variation in educational quality and access to other choice options (such as charter schools or private schools) but also because, as described above, intermediate school districts within which each student’s residential districts are clustered make their own participation rules for the School of Choice program. We use the linear probability model (LPM) to estimate (1). We utilize the LPM approach for three main reasons: 1) the results are straightforward to interpret, 2) the LPM does a good job of approximating the partial effects (Wooldridge, 2010, p. 563) and the marginal effects (Angrist & Pishke, 2009, p. 107), and 3) we are interested in looking at partial correlations, not making causal claims.
The decision to leave the School of Choice program, L, is a function of the same vector of student and school-level controls found in (1). However, in specifying the decision to leave the program and return to a resident district, we estimate:
(2) 
across all students in SoC at time t, where the dependence on time of students’ continued participation in Schools of Choice, g(ti) is distributed Weibull. Equation 2 allows us to explicitly account for the possibility that the longer a student remains in SoC the less (or more) likely it is that she will continue her participation in the following year. Such survival models are increasingly common in the school choice literature, as well as those in education policy more generally (e.g. Lavery and Carlson 2014).
Results
Participation in Schools of Choice
Table 2 compares students who stayed in their resident district (stayer) to SoC students who left their residential district after time t (leaver). Several differences and similarities are particularly apparent. The proportion of white students is considerably smaller among SoC participants than among students who remain in their assigned schools (0.65 compared to 0.74). This difference is explained almost entirely by disproportionate participation in Schools of Choice among African American students, as Hispanic and multiracial or other students participate at a rate similar to the proportion of these students who remain in public schools. School of Choice students are also more likely to receive free or reduced price lunch. The program has somewhat more balance between high school, middle school and elementary school students. Similar proportions of students with limited English proficiency or special academic needs participate in SoC relative to stayer students. Perhaps most striking in Table 2 are the math and reading MEAP scores of SoC students, which are considerably lower than those for stayer students. When standardized against their own residential district means (by grade), students who left their district to participate in SoC scored 0.08 of a standard deviation lower in math in their last year prior to switching than did students who did not switch. That difference was similar in reading. When standardized against the statewide means, these differences are even more stark, around 1/5 of a standard deviation. The bottom panel of Table 2 indicates that students who leave their resident districts under SoC attend schools with higher proportions of African Americans and higher levels of free/reduced price participation in the year prior to exit. Similarly, they also attend schools at t - 1 with lower levels of academic achievement relative to state averages (but not, unlike with student-level, lower levels relative to district averages).
The student-level patterns for switching into SoC at time t are largely present in Table 3, which provides estimates of Equation 1 both with and without school-level covariates included in the student-level model. In this multivariate framework, the role of race is less apparent than in the bivariate results in Table 2, but many of the other relationships are still evident. Lower income students (as indicated by free/reduced price lunch designation) are more likely to participate in Schools of Choice than non-F/RL students in the same district; students with special academic needs are less likely. Students with higher mathematics MEAP scores (standardized at the district level) are less likely to participate in SoC, while no relationship is apparent between MEAP reading and SoC participation (math and reading scores are highly correlated, at 0.65, so this is driven partly by collinearity). These results generally all hold when school-level attributes are added to the model. Few of these school-level predictors appear related to student-level participation, perhaps due to the absence of within-district school-level variation across much of the state—recall that each model already includes district fixed effects. The exception appears to be in school-level MEAP results: students attending schools with higher MEAP scores in math relative to other schools in the district are actually more likely to opt out. That result aside, the important comparison between both specifications in Table 3 is that student-level predictors of SoC participation are not affected by the inclusion or exclusion of school-level measures of those variables. Of note, the disproportionate rate of African American participation in Schools of Choice is explained away in the multivariate model once prior achievement and free/reduced lunch eligibility is taken into account.
Exiting Schools of Choice
To obtain a basic indication of a student’s duration in Schools of Choice, we simply calculate the average number of years of participation in SoC that we observe for each participant in the data. Among all participants, the average number of years in SoC is slightly less than 3 years (2.86). Among participants whom we observe transferring into the program at some point in the data—i.e. those we were not SoC in their first year in our data—this number is somewhat lower, at 2.29 years of SoC participation. As noted above, however, these numbers are somewhat affected by both the range of the data (2005-06 through 2012-13) and the fact that only a sub-group of our students—those whose kindergarten year coincided with our first year of data—are those we observe throughout the entirety of the panel. For those students, as well as those in kindergarten in either the fall of 2006 or 2007, we are able to observe at least the entirety of a student’s elementary school career (for the 2005 and 2006 cohorts we can observe at least one year of middle school as well, but we limit this description to elementary school for consistency across the three cohorts). For each of these students, the maximum number of years in School of Choice is 6—kindergarten plus grades 1 through 5. Table 4 reports the percentage of School of Choice participants who began in kindergarten and completed fifth grade in Schools of Choice. As Table 4 indicates, just under 40 percent of students who began kindergarten in Schools of Choice remained in the program through 5th grade. That figure is considerably lower for African American students, free/reduced lunch students and LEP students, and somewhat lower for Hispanics and students with special needs.
Table 5 depicts similar information for all students who exit in any given year, and includes standardized MEAP scores. Similar patterns emerge in the overall population of students: African American and free/reduced price lunch students are more prevalent among students who leave Schools of Choice than among those who stay. Students who leave the program also have lower math and reading MEAP test scores in their chosen school. Among school-level characteristics, the bottom panel of Table 5 also indicates that students in schools with more African Americans and free/reduced price lunch students are more likely to exit the program. On the other hand, students who leave SoC tend to leave schools with average district MEAP scores but slightly lower than the statewide school average. Taken alongside the student-level results, these patterns suggest that lack of sustained participation in SoC may be partly explained by attendance at schools with marginally higher levels of at-risk or minority students as well.
Table 6 provides estimates of Equation 2, the model of student duration in SoC. Among SoC participants at t-1, generally the same individual characteristics that predicted SoC in the first place also predict exit from the open enrollment program—and in the same directions as well. The hazard coefficient of 0.08 for African American students corresponds to a hazard ratio of about 1.09—the odds of exit are nearly 10 percent higher for African Americans. Lower income students are also more likely to exit: the reported hazard coefficient corresponds to a hazard ratio 1.5 times (or 50 percent greater) for free/reduced lunch students. Higher performing SoC participants on the MEAP exam—both math and reading subjects—are less likely to exit as well. A number of school characteristics also appear to predict exit. Students in schools with higher MEAP (reading) scores are less likely to exit, while students in schools with higher rates of special needs and LEP students are more likely to exit SoC. Although all of the coefficients on school-level race demographics are positive, the coefficient for school-level African American means stands out: going from a school with no African Americans to a school with all African Americans increases would more than triple the hazard of exit from SoC. To summarize: the results in Tables 4, 5, and 6 indicate that more advantaged students stay in the SoC program longer, as do students in schools with lower levels of at-risk or minority students.
Additional Specifications and Robustness Checks
The primary results above indicate that students who leave their resident districts under schools of choice tend to be lower performing in mathematics than their peers in the same district, and more likely to be from low-income families, as measured by free/reduced lunch participation. They are also disproportionately African American, but that difference disappears in the multivariate framework. Although we believe that the primary models provide adequate controls for observed local and temporal factors—the district-by-year fixed effect in all models and the additional time-varying school characteristics in some specifications account for locality-specific trends at a given point in time—there are still a number of other possibilities to consider. We test each of these possibilities in Table 7.
The first is that our results are explained by a small number of districts with disproportionate participation in Schools of Choice. In the appendix, we show a sample of districts with particularly high and low rates of student exit or entry, and the point is that there is considerable local variation in these transfers. We might especially worry that our results were driven by student mobility in the metropolitan Detroit area, which not only serves the greatest number of students in absolute terms, but with the presence of charter schools is a particularly choice-rich environment. To test this possibility, we re-estimate Equation 1 (the full specification, Column 2, presented in Table 3) excluding all students in Wayne County, which includes students in Detroit Public Schools and the immediate area. Column 1 of Table 7 reports these results, which are essentially unchanged.
More generally, it is possible that larger regional or geographic variation is driving these results. For example, large-urban centers are typically the focus of much of the literature on school choice due to the density of both students and schools. If our results are not confined simply to students in the Detroit-area, they may still be driving by mobility between districts in urban or suburban settings. To test this, we again re-estimate the full specification of Equation 1 (again, Column 2 of Table 3) separately for rural school districts and report these results in Column 2 of Table 7. Yet again, we see these results substantively unchanged from Table 3.
A separate explanation concerns the possibility that low performing or low-income students select into Schools of Choice at higher rate because other potential choosers—i.e. higher performing and higher-income students—select other options. For example, critics of charter schools have suggested that such a sector “cream-skims” more academically advantaged students away from traditional public schools. If, on the other hand, students attending charter schools are qualitatively similar to students participating in Schools of Choice the main estimates here should be attenuated. To consider these possibilities, we re-estimated Equation 1 including all charter students in the comparison group, and report these results in Table 7, column 3. The student-level results are materially unchanged from the earlier specification. On the other hand, the proportion of both LEP and Black/African American students in a school become statistically significant when including charter students in the comparison group. This suggests that charter school students attend schools with lower proportions of LEP and Black/African American students.
Finally, although our primary focus is specifically on students who transfer between districts under the Schools of Choice program provided by state statute, we have noted above that many districts operate local versions of open enrollment programs in partnerships with their neighboring communities. It is possible that, for whatever reason, students who select into Schools of Choice are different than inter-district transfers who switch under separate, local partnerships. We cannot observe all of these students directly in the data, but we can provide an “upper bound” estimate of their participation by simply considering a “local chooser” any student who is attending a district that is not listed as the student’s own residential district and is not listed as a School of Choice student in our administrative data. The district of residence is only available for the school years 2009-10 through 2012-13 which necessarily restricts the following analyses to those years. Table 8 depicts these students, as well as students we observe as formal participants in Schools of Choice, to provide a full estimate of all students participating in any kind of inter-district choice in Michigan. As indicated, when non-SoC choosers are included, the overall rate of inter-district choice in our sample jumps to 8.84 % between 2009-10 and 2012-13 compared with 6.27% for just SoC. We also re-estimate Equation 1 (column 2 of Table 3) in two new ways: first, by simply adding the local choosers to a more comprehensive indicator of inter-district choice, and predicting participation overall; second, by excluding School of Choice students and simply predicting participation in some other form of locally defined open enrollment. The results are presented in columns 1 and 2 of table 9, respectively. Column 3 of Table 9 presents a re-estimation of Equation 1 for the years 2009-2012 for students using SoC to enable a comparison within the same set of years. The results suggest some differences between local choosers and users of SoC. Black/African American students and students flagged as special needs are more likely to utilize local choice options that white students. In comparison, special needs students are less likely to utilize SoC options in both the 2009-2012 subsample and the full 2005-2012 panel. Black/African American students appear to be no more or less likely to use SoC than white students, although this is also the case in the primary results above, and is partially explained by the inclusion of prior test scores. It also appears that users of local choice policies are leaving schools with low average reading performance, which runs counter to those who are using SoC. While we cannot directly explore the nature of the local choice policies, students utilizing SoC differ in potentially systematic ways from students using local choice policies. It is possible local districts may coordinate with neighboring districts to provide specialized services to students with special needs, creating economies of scale. Further research is needed to understand the rules and rationale of the local choice policies in Michigan.
Discussion
The objective of this paper is to document student-level explanations for participation in Michigan’s Schools of Choice inter-district open enrollment program. From a purely descriptive standpoint, the results provided here are straightforward: participation is directly related to student-level attributes traditionally considered by both scholars and policymakers to indicate socioeconomic disadvantage. Students who choose to exit their resident districts are more likely to be African American, and have lower levels of income as measured by free/reduced lunch participation. They are also considerably lower performing, on average, on Michigan’s statewide math and reading exams, both relative to district and statewide averages. For the most part, these patterns hold when considered in a multivariate framework that not only holds constant each relationship between individual student characteristics and SoC participation but, importantly, also after netting out school-level explanations and comparisons to students within the same Michigan school district. Students who opt to leave their resident districts under Michigan’s open enrollment program are less advantaged on a number of measures than their peers who stay in the same district.
Taken in isolation, such results would indicate that Schools of Choice is a program that directly appeals to low-performing and other at-risk students in Michigan. But the same characteristics also help explain student exits from open enrollment. African Americans, free/reduced lunch participants, and students with lower MEAP scores are all more likely to exit the SoC program. They are also more likely to be leaving schools with higher concentrations of at-risk or minority students. The schools they leave have similar math and reading test scores to their district averages, but somewhat lower school test scores relative to the state average. Broadly, the notion that at-risk or low achieving students are among those most likely to transfer out of a choice program after earlier participation has been demonstrated in other choice systems—notably those employing school vouchers (Howell, et al. 2004; Cowen, et al. 2012)—as well as in other studies of open enrollment in other states (Carlson, Lavery and Hughes 2014).
These patterns indicate two broad patterns associated with participation in Michigan’s open enrollment program. The first is that participation itself is dynamic: to choose is not necessarily to stay. The second is the open enrollment participation rates imply the program provides a form of school choice to students from less advantaged backgrounds; however, those who exit the program are also the least advantaged relative to other SoC participants. If to choose is not necessarily to stay, it is especially not so for African Americans, free/reduced lunch participants, and students with lower test scores. In this, participation in and out of Michigan’s Schools of Choice program resembles patterns identified more broadly in studies of student mobility. In those studies, the most mobile students—those who transfer between public schools between and within academic years—are generally the least advantaged along a number of socioeconomic and academic measures. That such a pattern holds here suggests at least in part that one practical result of Michigan’s Schools of Choice is to open inter-district borders to students who are more mobile in the first place. Whether open enrollment influences the number of switches for this already mobile population is left to future research.
School choice remains a field that draws considerable controversy among scholars, policymakers and the general public as well. Much of that debate concerns access to high quality education—whether provided by traditionally organized public schools, charter schools, or non-public alternatives—for historically disadvantaged students. Given that such students are precisely those we have shown here to both participate and exit from this particular choice option, it is important to state directly what this paper does not show—at least in its present iteration. First, this paper does not attempt to link participation in Schools of Choice to changes in student achievement. We do not, in other words, identify or estimate the effect of open enrollment on outcomes, despite our consideration of the relationship between achievement among SoC participants and the decision to stay in or leave the program. Second, this paper does not directly consider which characteristics of a particular school or district may draw a student toward it under Schools of Choice: without considering the characteristics of receiving schools or districts, we cannot draw firm conclusions about the role of a school’s demographic or academic characteristics and its ability to draw or retain choice students. Finally, for the same reason, our results do not directly inform questions about race-based school or district segregation. Whether Michigan’s School of Choice program alleviates, exacerbates, or has little implication for segregation in the state’s public school system is not under study here. That question, along with the others outlined above, are subjects for our next studies of this choice program.
References
Alexander, K. L., Entwisle, D. R., & Dauber, S. L. (1996). Children in Motion: School Transfers and Elementary School Performance. Journal of Educational Research, 90(1), 3-12.
Armor, D. J., & Peiser, B. M. (1998). Interdistrict Choice in Massachusetts. In P. E. Peterson, & B. C. Hassel (Eds.), Learning from School Choice. (pp. 157-186). Washington D.C: Brookings Institution Press.
Bifulco, R., Cobb, C. D., & Bell, C. (2009). Can interdistrict choice boost student achievement? The case of Connecticut’s interdistrict magnet school program. Educational Evaluation and Policy Analysis 31(4), 323-345.
Carlson, D., Lavery, L., & Witte, J. F. (2011). The determinants of interdistrict open enrollment flows evidence from two states. Educational Evaluation and Policy Analysis, 33(1), 76-94
Carlson, D. Lavery, L., Hughes, T. (2014) Schooling Decisions and Student Achievement Outcomes under Interdistrict Open Enrollment Working Paper University of Oklahoma.
Cowen, J. M., Fleming, D. J., Witte, J. F., & Wolf, P. J. (2012). Going Public Who Leaves a Large, Longstanding, and Widely Available Urban Voucher Program?. American Educational Research Journal, 49(2), 231-256.
Cowen, J. M., & Toma, E. F. (2015). Emerging Alternatives to Neighborhood-Based Public Schooling. In H. Ladd, & M. Goertz, Handbook of Education Finance and Policy. Association for Education Finance and Policy.
Cullen, J. B., Jacob, B. A., & Levitt, S. D. (2005). The impact of school choice on student outcomes: an analysis of the Chicago Public Schools. Journal of Public Economics, 89(5), 729-760.
Deming, D., Hastings, J., Kane, T., & Staiger, D. (2011). School choice, school quality and academic achievement. NBER Working Paper, 17438.
Fossey, R. (1994). Open Enrollment In Massachusetts: Why Families Choose. Educational Evaluation And Policy Analysis, 16(3), 320.
Hanushek, E. A., Kain, J. F., & Rivkin, S. G. (2004). Disruption versus Tiebout Improvement: The Costs and Benefits of Switching Schools. Journal of Public Economics, 88 (9-10), 1721-1746.
Hastings, J. S., & Weinstein, J. M. (2007). Information, school choice, and academic achievement: Evidence from two experiments (No. w13623). National Bureau of Economic Research.
Holme, J. J., & Richards, M. P., (2009). School Choice and Stratification In Regional Context: Examining The Role Of Inter-District Choice. Peabody Journal Of Education, 84, 150-171.
Howell, W. G. (2004). Dynamic selection effects in means‐tested, urban school voucher programs. Journal of Policy Analysis and Management, 23(2), 225-250.
Ingersoll, G. M., Scamman, J. P., & Eckerling, W. D. (1989). Geographic Mobility and Student Achievement in an Urban Setting. Educational Evaluation and Policy Analysis, 11(2), 143-149.=
Kerbow, D. (1996). Patterns of Urban Student Mobility and Local School Reform. Journal of Education for Students Placed at Risk, 1(2), 1147-169.
Kerbow, D., Azcoitia, C., & Buell, B. (2003). Student Mobility and Local School Improvement in Chicago. Journal of Negro Education, 72(1), 158-164.
Lavery, L., & Carlson, D. (2014). Dynamic Participation in Interdistrict Open Enrollment. Educational Policy (forthcoming).
Mikulecky, M. T. (2013). Open Enrollment is on the Menu--But Can You Order It? Education Commission of the States.
National Center for Education Statistics. (2013). State Education Reforms: Table 4.2. Retrieved November 24, 2014, from http://nces.ed.gov/programs/statereform/tab4_2.asp.
Phillips, K. J., Hausman, C., & Larsen, E. S. (2012). Students who choose and the schools they leave: Examining participation in intradistrict transfers. The Sociological Quarterly, 53(2), 264-294.
Reback, R. (2008). Demand (and supply) in an inter-district public school choice program. Economics of Education Review, 27(4), 402-416
Rumberger, R. W., Larson, K. A., Ream, R. K., & Palardy, G. J. (1999). The Educational Consequences of Mobility for California Students and Schools. Berkeley, CA: Policy Analysis for California Education.
Schneider, B., Schiller, K. S., & Coleman, J. S. (1996). Public School Choice: Some Evidence from the National Longitudinal Study of 1988. Educational Evaluation and Policy Analysis , 18 (1), 19-29.
South, S. J., Haynie, D. L., & Bose, S. (2007). Student Mobility and School Dropout. Social Science Research, 36(1), 68-94.
Temple, J. A., & Reynolds, A. J. (1999). School Mobility and Achievement: Longitudinal Findings from an Urban Cohort. Journal of School Psychology, 37(4), 355-377.
Welsch, D.M., Statz, B., & Skidmore, M. (2010). An Examination Of Inter-District Public School Transfers In Wisconsin. Economics of Education Review, 29(1): 126-137.
Welsch, D. M., & Zimmer, D. M. (2012). Do student migrations affect school performance? Evidence from Wisconsin's inter-district public school program. Economics of Education Review, 31(1), 195-207.
Xu, Z., Hannaway, J., & D'Souza, S. (2009). Student Transience in North Carolina: The Effect of School Mobility on Student Outcomes Using Longitudinal Data. Washington, DC: Urban Institute.
Notes:
(1) Michigan State School Aid Act [MCL 388.1705; 388.1705c; Sections 105, 105c]; Michigan Department of Education (2013; http://mi.gov/documents/mde/choice1_279579_7.pdf)
(2) For example, some districts participate at minimum levels required to receive Best Practices funding from the state
http://www.michigan.gov/documents/mde/Best_Practice_Guidance_463861_7.pdf
(3) Approximately 5.75% of student-years (806,341) had multiple records for a given year in the raw data. An examination of the data revealed this occurred typically when students switched schools midyear. We used a three step process to deal with duplicate student observations in time 1) if a student participated in SoC at any point in time t, all non-SoC entries were dropped; 2) if a student had multiple observations associated with SoC enrollment, we kept the record associated with the earliest SoC enrollment (i.e. we selected the October enrollment record over a December enrollment date); 3) for students with multiple records in a given t and never enrolled in SoC, we randomly selected an observation for the student in year t.
TABLES
Table 1: Participation Rates for Schools of Choice and Charter Schools 2005-6 to 2012-13
Year |
SoC |
PSA |
Total K-12 Enrollment |
2005-06 |
66560 |
94529 |
1781866 |
3.74% |
5.31% |
100% |
|
2006-07 |
71587 |
101787 |
1766148 |
4.05% |
5.76% |
100% |
|
2007-08 |
83177 |
103839 |
1755975 |
4.74% |
5.91% |
100% |
|
2008-09 |
86341 |
103983 |
1718093 |
5.03% |
6.36% |
100% |
|
2009-10 |
91995 |
109348 |
1693019 |
5.43% |
6.74% |
100% |
|
2010-11 |
99982 |
114057 |
1660234 |
6.02% |
7.22% |
100% |
|
2011-12 |
107970 |
119943 |
1642437 |
6.57% |
7.91% |
100% |
|
2012-13 |
115209 |
129942 |
1629432 |
7.07% |
7.97% |
100% |
|
Total sample |
809162 |
981411 |
13647204 |
5.93% |
7.19% |
100% |
Source: Author calculations from data provided by the Michigan Department of Education
Table 2: Characteristics of Students Transferring into Schools of Choice 2005-06 to 2011-12
Variables (all at t) |
Stayer in District at t |
Enter Schools of Choice at t |
|
Student characteristics |
|||
Female |
0.484 |
0.497 |
|
White |
0.742 |
0.649 |
|
Black/African American |
0.165 |
0.273 |
|
Hispanic |
0.053 |
0.049 |
|
Multiracial |
0.014 |
0.015 |
|
Elementary school |
0.476 |
0.420 |
|
Middle school |
0.253 |
0.245 |
|
High School |
0.271 |
0.335 |
|
Free/Reduced lunch |
0.387 |
0.540 |
|
Limited English Proficiency |
0.038 |
0.017 |
|
Special Needs |
0.129 |
0.136 |
|
Standardized at district level |
|||
MEAP Math z-score at t |
0.005 |
-0.080 |
|
MEAP Reading z-score at t |
0.003 |
-0.053 |
|
Standardized at state level |
|||
MEAP Math z-score at t |
0.0400 |
-0.243 |
|
MEAP Reading z-score at t |
0.0311 |
-0.188 |
|
Proportion switched or not switched |
0.984 |
0.016 |
|
Student year N |
8,910,850 |
142,145 |
|
School Characteristics |
|||
School Mean Female |
0.484 |
0.483 |
|
School Mean White |
0.738 |
0.662 |
|
School Mean Black/African American |
0.166 |
0.255 |
|
School Mean Hispanic |
0.054 |
0.052 |
|
School Mean Multiracial |
0.014 |
0.013 |
|
School Mean Free/Reduced lunch |
0.383 |
0.475 |
|
School Mean Limited English Proficiency |
0.037 |
0.025 |
|
School Mean Special Needs |
0.129 |
0.136 |
|
Standardized at district level |
|||
School Mean MEAP Math z-score |
0.012 |
0.012 |
|
School Mean MEAP Reading z-score |
0.007 |
0.008 |
|
Standardized at state level |
|||
School Mean MEAP Math z-score |
0.002 |
-0.197 |
|
School Mean MEAP Reading z-score |
-0.007 |
-0.170 |
|
Student-year N |
10,301,574 |
170,174 |
Source: Author calculations from data provided by the Michigan Department of Education; transfer defined as participation at t conditional on non-participation at t-1; 2012-13 excluded.
Table 3: Estimated Models of Student Transfer into Schools of Choice at t
Variables (all at t-1) |
1 |
2 |
|
||
female |
0.001*** |
0.001*** |
(0.000) |
(0.000) |
|
blackafam |
0.000 |
0.001 |
(0.001) |
(0.001) |
|
hispanic |
0.000 |
0.000 |
(0.000) |
(0.000) |
|
multiracial |
0.001** |
0.001* |
(0.001) |
(0.001) |
|
amerindian |
0.002** |
0.002** |
(0.001) |
(0.001) |
|
asianamer |
0.000 |
0.000 |
(0.001) |
(0.001) |
|
pacificislander |
0.001 |
0.001 |
(0.002) |
(0.002) |
|
middle |
0.013*** |
0.013*** |
(0.003) |
(0.003) |
|
frl |
0.005*** |
0.005*** |
(0.000) |
(0.000) |
|
lep |
-0.006*** |
-0.006*** |
(0.001) |
(0.001) |
|
specialneeds |
-0.001*** |
-0.001*** |
(0.000) |
(0.000) |
|
district_standardizedMEAPmath |
-0.001*** |
-0.001*** |
(0.000) |
(0.000) |
|
district_standardizedMEAPreading |
-0.000 |
-0.000 |
(0.000) |
(0.000) |
|
schmeandistrictMEAPmath |
-0.003** |
|
(0.001) |
||
schmeandistrictMEAPreading |
0.007*** |
|
(0.002) |
||
schmeanspecialneeds |
0.001 |
|
(0.004) |
||
schmeanlep |
0.002 |
|
(0.003) |
||
schmeanfrl |
-0.002 |
|
(0.002) |
||
schmeanwhite |
0.026** |
|
(0.013) |
||
schmeanpacificislander |
0.026 |
|
(0.019) |
||
schmeanblackafam |
0.019 |
|
(0.013) |
||
schmeanasianamer |
0.031** |
|
(0.014) |
||
schmeanamerindian |
0.007 |
|
(0.008) |
||
schmeanmultiracial |
0.037** |
|
(0.014) |
||
schmeanhispanic |
0.020 |
|
(0.013) |
||
state_standardizedMEAPmath |
||
state_standardizedMEAPreading |
||
Constant |
-0.002 |
-0.026** |
(0.003) |
(0.013) |
|
Observations |
4,188,476 |
4,188,476 |
Number of distxyear |
3,986 |
3,986 |
***p<0.01 **p<0.05 *p<0.10; robust standard errors in parantheses clustered by school at t-1. Estimates are linear probability coefficients; Models include district by year fixed effects and grade fixed effects.
Table 4: % of School of Choice Participants Remaining in Program Through Elementary School
|
% in School of Choice K through 5th Grade |
All Students in Schools of Choice |
39.49 |
Female |
40.43 |
Male |
38.68 |
White |
41.34 |
Black/African American |
28.75 |
Hispanic |
34.47 |
Free/Reduced Lunch |
28.47 |
LEP |
28.95 |
Special Needs |
33.68 |
Note: Based on kindergarten cohorts from the 2005, 2006 and 2007 academic years.
Table 5: Characteristics of Students Transferring Out of Schools of Choice 2005-06 to
2011-12
Variables (all at t-1) |
Stayer in SoC at t |
Exit SoC at t |
|
Female |
0.499 |
0.492 |
|
White |
0.723 |
0.666 |
|
Black/African American |
0.198 |
0.257 |
|
Hispanic |
0.048 |
0.048 |
|
Multiracial |
0.018 |
0.017 |
|
Elementary school |
0.460 |
0.449 |
|
Middle school |
0.229 |
0.216 |
|
High School |
0.298 |
0.305 |
|
Free/Reduced lunch |
0.405 |
0.480 |
|
Limited English Proficiency |
0.012 |
0.010 |
|
Special Needs |
0.107 |
0.121 |
|
Standardized at district level |
|||
MEAP Math z-scorea |
-0.029 |
-0.128 |
|
MEAP Reading z-scorea |
0.000 |
-0.092 |
|
Standardized at state level |
|||
MEAP Math z-scorea |
-0.028 |
-0.241 |
|
MEAP Reading z-scorea |
0.008 |
-0.188 |
|
Student-year N |
434,056 |
104,298 |
|
|
|
|
|
School Characteristics |
|
|
|
School Mean Female |
|
0.487 |
0.486 |
School Mean White |
|
0.761 |
0.705 |
School Mean Black/African American |
|
0.158 |
0.217 |
School Mean Hispanic |
|
0.047 |
0.047 |
School Mean Multiracial |
|
0.014 |
0.013 |
School Mean Free/Reduced lunch |
|
0.392 |
0.426 |
School Mean Limited English Proficiency |
|
0.019 |
0.018 |
School Mean Special Needs |
|
0.119 |
0.122 |
Standardized at district level |
|
|
|
School Mean MEAP Math z-scorea |
|
0.017 |
0.006 |
School Mean MEAP Reading z-scorea |
|
0.015 |
0.008 |
Standardized at state level |
|
|
|
School Mean MEAP Math z-scorea |
|
-0.015 |
-0.113 |
School Mean MEAP Reading z-scorea |
|
-0.006 |
-0.085 |
Student-year N |
|
434,056 |
104,298 |
Source: Author calculations from data provided by the Michigan Department of Education; transfer defined as non-participation at t conditional on participation at t-1; 2012-13 excluded.
Table 6: Hazard Estimates of Student Exit from Schools of Choice at t
Variables (all at t-1) |
(1) |
female |
0.018* |
(0.011) |
|
blackafam |
0.084*** |
(0.032) |
|
hispanic |
-0.009 |
(0.031) |
|
multiracial |
0.031 |
(0.038) |
|
amerindian |
0.178*** |
(0.067) |
|
asianamer |
-0.148** |
(0.073) |
|
pacificislander |
0.157 |
(0.179) |
|
middle |
-0.412*** |
(0.033) |
|
frl |
0.382*** |
(0.016) |
|
lep |
0.058 |
(0.063) |
|
specialneeds |
-0.028 |
(0.019) |
|
district_standardizedMEAPmath |
-0.063*** |
|
(0.008) |
district_standardizedMEAPreading |
-0.027*** |
|
(0.007) |
schmeandistrictMEAPmath |
0.018 |
(0.204) |
|
schmeandistrictMEAPreading |
-0.222 |
(0.224) |
|
schmeanspecialneeds |
1.331*** |
(0.205) |
|
schmeanlep |
1.008*** |
(0.313) |
|
schmeanfrl |
-0.631*** |
(0.130) |
|
schmeanwhite |
0.396 |
(0.302) |
|
schmeanblackafam |
1.143*** |
(0.305) |
|
schmeanhispanic |
0.289 |
(0.371) |
|
Constant |
-5.540*** |
(0.297) |
|
ln_p |
0.718*** |
(0.006) |
|
Student-year N |
400,213 |
***p<0.01 **p<0.05 *p<0.10; robust standard errors in parantheses clustered by school at t-1. Estimates are hazard coefficients
Table 7: Estimated Models of Student Transfer into Schools of Choice at t for Different Sample Specifications.
|
Excluding Wayne County |
Rural |
Including PSA |
Variables (all at t-1) |
|
|
|
|
|
||
female |
0.001*** |
0.001*** |
0.001*** |
(0.000) |
(0.000) |
(0.000) |
|
blackafam |
-0.000 |
0.000 |
0.000 |
(0.001) |
(0.001) |
(0.001) |
|
hispanic |
-0.000 |
-0.002** |
0.000 |
(0.000) |
(0.001) |
(0.000) |
|
multiracial |
0.001 |
0.002 |
0.001 |
(0.001) |
(0.002) |
(0.001) |
|
amerindian |
0.002** |
0.002 |
0.002*** |
(0.001) |
(0.002) |
(0.001) |
|
asianamer |
-0.001 |
-0.001 |
0.000 |
(0.001) |
(0.001) |
(0.001) |
|
pacificislander |
0.001 |
0.002 |
0.001 |
(0.002) |
(0.004) |
(0.002) |
|
middle |
0.011*** |
0.007 |
0.007 |
(0.004) |
(0.007) |
(0.004) |
|
frl |
0.006*** |
0.007*** |
0.004*** |
(0.000) |
(0.000) |
(0.000) |
|
lep |
-0.006*** |
-0.008*** |
-0.006*** |
(0.000) |
(0.001) |
(0.001) |
|
specialneeds |
-0.001*** |
-0.001 |
-0.001*** |
(0.000) |
(0.000) |
(0.000) |
|
district_standardizedMEAPmath |
-0.001*** |
-0.002*** |
-0.001*** |
(0.000) |
(0.000) |
(0.000) |
|
district_standardizedMEAPreading |
-0.000 |
-0.000** |
-0.000 |
(0.000) |
(0.000) |
(0.000) |
|
schmeandistrictMEAPmath |
-0.002 |
0.001 |
-0.003** |
(0.001) |
(0.003) |
(0.001) |
|
schmeandistrictMEAPreading |
0.006*** |
0.003 |
0.007*** |
(0.002) |
(0.003) |
(0.002) |
|
schmeanspecialneeds |
-0.001 |
0.006 |
0.001 |
(0.005) |
(0.007) |
(0.004) |
|
schmeanlep |
-0.001 |
0.017 |
0.006* |
(0.004) |
(0.014) |
(0.003) |
|
schmeanfrl |
0.004* |
0.003 |
-0.000 |
(0.002) |
(0.005) |
(0.002) |
|
schmeanwhite |
0.008 |
-0.027 |
0.033** |
(0.013) |
(0.023) |
(0.014) |
|
schmeanpacificislander |
-0.001 |
-0.164 |
0.029 |
(0.018) |
(0.114) |
(0.019) |
|
schmeanblackafam |
-0.004 |
-0.029 |
0.027** |
(0.014) |
(0.030) |
(0.014) |
|
schmeanasianamer |
0.009 |
-0.047* |
0.039*** |
(0.013) |
(0.026) |
(0.015) |
|
schmeanamerindian |
0.000 |
-0.048 |
0.007 |
(0.008) |
(0.033) |
(0.008) |
|
schmeanmultiracial |
0.003 |
-0.013 |
0.057*** |
(0.014) |
(0.024) |
(0.016) |
|
schmeanhispanic |
0.006 |
-0.036 |
0.026* |
(0.013) |
(0.023) |
(0.013) |
|
Constant |
-0.009 |
0.031 |
-0.028** |
(0.013) |
(0.024) |
(0.014) |
|
Observations |
3,430,297 |
990,524 |
4,530,376 |
Number of Districts x Year |
0.001 |
0.001 |
0.001 |
***p<0.01 **p<0.05 *p<0.10; robust standard errors in parantheses clustered by school at t-1. Estimates are linear probability coefficients; Models include district by year fixed effects and grade fixed effects.
Table 8: Summary of All Inter-district Transfers in Michigan 2009 - 2012
|
N (% of all Inter-district Choosers) |
Students transferring under Schools of Choice |
415,156 (70.9%) |
Non-PSA “local choosers” |
170,424 (19.1%) |
Total |
585,580 (100%) |
(% of All Michigan Students) |
8.84% |
Note: “local choosers” defined as students with non-matching resident and attending district codes, excluding students in Schools of Choice.
Table 9: Estimated Models of Student Transfer into Alternate Definitions of Inter-district Choice at t (2009-2012)
Variables (all at t-1) |
(1) |
(2) |
(3) |
Any Non-Resident Switch |
Switch Under Local Policy |
SoC |
|
female |
0.001*** |
-0.000 |
0.001*** |
(0.000) |
(0.000) |
(0.000) |
|
blackafam |
0.001 |
0.002*** |
-0.001 |
(0.001) |
(0.000) |
(0.001) |
|
hispanic |
0.001 |
0.001 |
-0.001 |
(0.001) |
(0.001) |
(0.001) |
|
multiracial |
0.001 |
0.002** |
0.000 |
(0.001) |
(0.001) |
(0.001) |
|
amerindian |
0.001 |
0.001 |
-0.000 |
(0.002) |
(0.001) |
(0.001) |
|
asianamer |
0.003** |
0.001** |
0.001 |
(0.001) |
(0.001) |
(0.001) |
|
pacificislander |
0.003 |
0.001 |
0.002 |
(0.004) |
(0.003) |
(0.003) |
|
middle |
0.047*** |
0.020** |
0.026*** |
(0.010) |
(0.009) |
(0.008) |
|
frl |
0.008*** |
0.003*** |
0.005*** |
(0.000) |
(0.000) |
(0.000) |
|
lep |
-0.010*** |
-0.003*** |
-0.006*** |
(0.001) |
(0.001) |
(0.001) |
|
specialneeds |
0.003*** |
0.004*** |
-0.001*** |
(0.001) |
(0.000) |
(0.000) |
|
district_standardizedMEAPmath |
-0.002*** |
-0.001*** |
-0.001*** |
(0.000) |
(0.000) |
(0.000) |
|
district_standardizedMEAPreading |
-0.000*** |
-0.000** |
-0.000 |
(0.000) |
(0.000) |
(0.000) |
|
schmeandistrictMEAPmath |
-0.005 |
-0.002 |
-0.002 |
(0.004) |
(0.003) |
(0.002) |
|
schmeandistrictMEAPreading |
-0.008 |
-0.016** |
0.008*** |
(0.008) |
(0.007) |
(0.002) |
|
schmeanspecialneeds |
0.009 |
0.003 |
0.000 |
(0.014) |
(0.012) |
(0.005) |
|
schmeanlep |
-0.026** |
-0.023** |
-0.002 |
(0.011) |
(0.009) |
(0.004) |
|
schmeanfrl |
-0.003 |
-0.008 |
0.003 |
(0.007) |
(0.006) |
(0.003) |
|
schmeanwhite |
0.049* |
0.022 |
0.027 |
(0.029) |
(0.018) |
(0.017) |
|
schmeanpacificislander |
0.113 |
0.056 |
0.036 |
(0.119) |
(0.113) |
(0.050) |
|
schmeanblackafam |
0.018 |
0.001 |
0.019 |
(0.029) |
(0.020) |
(0.017) |
|
schmeanasianamer |
0.078** |
0.050** |
0.031* |
(0.034) |
(0.025) |
(0.018) |
|
schmeanamerindian |
0.006 |
0.003 |
0.003 |
(0.023) |
(0.020) |
(0.009) |
|
schmeanmultiracial |
0.012 |
0.002 |
0.013 |
(0.027) |
(0.018) |
(0.019) |
|
schmeanhispanic |
0.029 |
0.007 |
0.025 |
(0.028) |
(0.016) |
(0.016) |
|
Constant |
-0.072** |
-0.028 |
-0.039** |
(0.030) |
(0.020) |
(0.019) |
|
Observations |
1,636,522 |
1,637,799 |
1,703,473 |
Number of Districts x Year |
1,631 |
1,631 |
1,702 |
***p<0.01 **p<0.05 *p<0.10; robust standard errors in parantheses clustered by school at t-1. Estimates are linear probability coefficients; Models include district by year fixed effects and grade fixed effects. Note: The sample for these estimates are restricted to the years 2009 -2012. This is limited due to data availability for students' district of residence.
APPENDICES:
A1 Descriptive statistics of students based on enrollment in Schools of Choice (2005 – 2012)
Variables (all at t) |
Non-School of Choice students (at t) |
Students in any type of School of Choice (at t) |
|
Student characteristics |
|||
Female |
0.485 |
0.498 |
|
White |
0.713 |
0.707 |
|
Black/African American |
0.191 |
0.214 |
|
Hispanic |
0.056 |
0.050 |
|
Multiracial |
0.015 |
0.019 |
|
Elementary school |
0.444 |
0.424 |
|
Middle school |
0.232 |
0.213 |
|
High School |
0.324 |
0.363 |
|
Free/Reduced lunch |
0.400 |
0.427 |
|
Limited English Proficiency |
0.039 |
0.012 |
|
Special Needs |
0.123 |
0.107 |
|
Standardized at district level |
|||
MEAP Math z-score at t |
0.004 |
-0.045 |
|
MEAP Reading z-score at t |
0.001 |
-0.016 |
|
Standardized at state level |
|||
MEAP Math z-score at t |
0.010 |
-0.068 |
|
MEAP Reading z-score at t |
0.005 |
-0.027 |
|
Proportion switched or not switched |
0.946 |
0.054 |
|
Student year N |
12,251,258 |
695,329 |
|
School Characteristics |
|||
School Mean Female |
0.486 |
0.487 |
|
School Mean White |
0.710 |
0.745 |
|
School Mean Black/African American |
0.193 |
0.174 |
|
School Mean Hispanic |
0.056 |
0.048 |
|
School Mean Multiracial |
0.015 |
0.016 |
|
School Mean Free/Reduced lunch |
0.399 |
0.408 |
|
School Mean Limited English Proficiency |
0.038 |
0.020 |
|
School Mean Special Needs |
0.125 |
0.119 |
|
Standardized at district level |
|||
School Mean MEAP Math z-score |
0.015 |
0.017 |
|
School Mean MEAP Reading z-score |
0.010 |
0.013 |
|
Standardized at state level |
|||
School Mean MEAP Math z-score |
-0.033 |
-0.040 |
|
School Mean MEAP Reading z-score |
-0.038 |
-0.020 |
|
Student-year N |
12,251,258 |
695,329 |
A2: SoC Participation by Students Classified as SoC 2005-2013
A2. Top and bottom quintiles of districts based on proportion of students in district that are SoC based on average per year proportion. |
||||
Quintile 1 |
Quintile 5 |
|||
Allegan Area Educational Service |
0 |
Sodus Township S/D #5 |
0.8908529 |
|
Allendale Public Schools |
0 |
Oneida Township S/D #3 |
0.874958 |
|
Alpena-Montmorency-Alcona ESD - |
0 |
Colfax Township S/D #1F |
0.8601665 |
|
Atherton Community Schools |
0 |
Bangor Township S/D #8 |
0.8098626 |
|
Barry ISD - District created fro |
0 |
Glenn Public School District |
0.7691527 |
|
Bay-Arenac ISD - District create |
0 |
Sigel Township S/D #3F |
0.7138045 |
|
Beaver Island Community School |
0 |
Church School District |
0.6482862 |
|
Beecher Community School Distric |
0 |
Verona Township S/D #1F |
0.6347761 |
|
Belding Area School District |
0 |
Hagar Township S/D #6 |
0.6189465 |
|
Bendle Public Schools |
0 |
Moran Township School District |
0.5952899 |
|
Berrien RESA - District created |
0 |
Sigel Township S/D #6 |
0.580763 |
|
Bloomfield Hills Schools |
0 |
Highland Park City Schools |
0.4718556 |
|
Bois Blanc Pines School District |
0 |
Mar Lee School District |
0.4621197 |
|
Branch ISD - District created fr |
0 |
Clintondale Community Schools |
0.450133 |
|
C.O.O.R. ISD - District created |
0 |
Bloomfield Township S/D #7F |
0.4416667 |
|
Caledonia Community Schools |
0 |
Free Soil Community Schools |
0.4387443 |
|
Calhoun ISD - District created f |
0 |
Sigel Township S/D #4F |
0.4102028 |
|
Charlevoix-Emmet ISD - District |
0 |
Madison School District (Lenawee |
0.408216 |
|
Cheb-Otsego-Presque Isle ESD - D |
0 |
Vandercook Lake Public Schools |
0.4078569 |
|
Chelsea School District |
0 |
Westwood Community School Distri |
0.407191 |
|
Clare-Gladwin Regional Education |
0 |
School District of the City of I |
0.3967995 |
|
Clinton County RESA - District c |
0 |
Grant Township S/D #2 |
0.39375 |
|
Clio Area School District |
0 |
Oak Park, School District of the |
0.3886454 |
|
Comstock Public Schools |
0 |
AuTrain-Onota Public Schools |
0.3876815 |
|
Copper Country ISD - District cr |
0 |
Madison District Public Schools |
0.3842303 |
|
Dearborn City School District |
0 |
Dearborn Heights School District |
0.3836385 |
|
Delta-Schoolcraft ISD - District |
0 |
Carrollton Public Schools |
0.3798007 |
|
Dexter Community School District |
0 |
Eau Claire Public Schools |
0.3723673 |
|
Dickinson-Iron ISD - District cr |
0 |
Galien Township School District |
0.3595136 |
|
Dundee Community Schools |
0 |
Britton Deerfield Schools |
0.3518681 |
|
Eastern Upper Peninsula ISD - Di |
0 |
Jonesville Community Schools |
0.3451546 |
|
Easton Township S/D #6 |
0 |
River Rouge, School District of |
0.3436198 |
|
Eaton ISD - District created fro |
0 |
Corunna Public Schools |
0.3384913 |
|
Education Achievement System-EAS |
0 |
Arvon Township School District |
0.3271329 |
|
Genesee ISD - District created f |
0 |
Beal City Public Schools |
0.3196734 |
|
Godfrey-Lee Public Schools |
0 |
Riverview Community School Distr |
0.3191296 |
|
Gogebic-Ontonagon ISD - District |
0 |
Sand Creek Community Schools |
0.3139615 |
|
Grand Rapids Public Schools |
0 |
Wells Township School District |
0.3106419 |
|
Grant Public School District |
0 |
Caseville Public Schools |
0.3023363 |
|
Gratiot-Isabella RESD - District |
0 |
Pittsford Area Schools |
0.3012538 |
|
Grosse Pointe Public Schools |
0 |
Essexville-Hampton Public School |
0.2940496 |
|
Harbor Springs School District |
0 |
Houghton-Portage Township School |
0.291479 |
|
Hillsdale ISD - District created |
0 |
Bridgman Public Schools |
0.2811575 |
|
Huron ISD - District created fro |
0 |
School District of Ypsilanti |
0.2707297 |
|
Ida Public School District |
0 |
Dollar Bay-Tamarack City Area Sc |
0.2688 |
|
Ingham ISD - District created fr |
0 |
Lakeview Public Schools (Macomb) |
0.2650043 |
|
Ionia ISD - District created fro |
0 |
Pennfield Schools |
0.2642007 |
|
Ionia Public Schools |
0 |
Swan Valley School District |
0.2637673 |
|
Iosco RESA - District created fr |
0 |
Bark River-Harris School Distric |
0.2543038 |
|
Jackson ISD - District created f |
0 |
Western School District |
0.2535662 |
|
Kalamazoo Public Schools |
0 |
Carney-Nadeau Public Schools |
0.2526017 |
|
Kalamazoo RESA - District create |
0 |
Coloma Community Schools |
0.2488133 |
|
Kelloggsville Public Schools |
0 |
Mason County Eastern Schools |
0.2486079 |
|
Kent ISD - District created from |
0 |
Hanover-Horton School District |
0.2475046 |
|
Lake Orion Community Schools |
0 |
New Buffalo Area Schools |
0.2437747 |
|
Lapeer ISD - District created fr |
0 |
Michigan Center School District |
0.2426145 |
|
Lenawee ISD - District created f |
0 |
Melvindale-North Allen Park Scho |
0.2415673 |
|
Lewis Cass ISD - District create |
0 |
Onekama Consolidated Schools |
0.2410066 |
|
Livingston ESA - District create |
0 |
Pentwater Public School District |
0.2375597 |
|
Mackinac Island Public Schools |
0 |
Mason County Central Schools |
0.2365329 |
|
Macomb ISD - District created fr |
0 |
Watervliet School District |
0.2359947 |
|
Manistee ISD - District created |
0 |
Bangor Township Schools |
0.2346192 |
|
Marquette-Alger RESA - District |
0 |
Clawson Public Schools |
0.2324029 |
|
Mecosta-Osceola ISD - District c |
0 |
Ecorse Public Schools |
0.2319877 |
|
Menominee ISD - District created |
0 |
East Jackson Community Schools |
0.2306659 |
|
Michigan Department of Human Ser |
0 |
West Bloomfield School District |
0.228474 |
|
Midland County Educational Servi |
0 |
North Adams-Jerome Public School |
0.2281646 |
|
Montcalm Area ISD - District cre |
0 |
Burt Township School District |
0.2273935 |
|
Muskegon Area ISD - District cre |
0 |
Ferndale Public Schools |
0.2268047 |
|
Newaygo County RESA - District c |
0 |
Saugatuck Public Schools |
0.2232514 |
|
Northview Public Schools |
0 |
Peck Community School District |
0.2181366 |
|
Novi Community School District |
0 |
Bear Lake Schools |
0.2130223 |
|
Oakland Schools - District creat |
0 |
Big Jackson School District |
0.2086378 |
|
Oakridge Public Schools |
0 |
Palo Community School District |
0.2072192 |
|
Ottawa Area ISD - District creat |
0 |
Fulton Schools |
0.2058943 |
|
Portland Public Schools |
0 |
Berrien Springs Public Schools |
0.2051829 |
|
Rochester Community School Distr |
0 |
Elm River Township School Distri |
0.2041667 |
|
Saginaw ISD - District created f |
0 |
Fraser Public Schools |
0.2041353 |
|
Sanilac ISD - District created f |
0 |
Owendale-Gagetown Area School Di |
0.2015345 |
|
Saranac Community Schools |
0 |
Buena Vista School District |
0.1994212 |
|
School District of the City of B |
0 |
Concord Community Schools |
0.1990893 |
|
Shiawassee Regional ESD - Distri |
0 |
New Lothrop Area Public Schools |
0.1947801 |
|
St. Clair County RESA - District |
0 |
East Lansing School District |
0.1921928 |
|
St. Johns Public Schools |
0 |
Mount Clemens Community School D |
0.1914591 |
|
St. Joseph County ISD - District |
0 |
Allen Park Public Schools |
0.190322 |
|
Summerfield Schools |
0 |
Carsonville-Port Sanilac School |
0.1902555 |
|
Traverse Bay Area ISD - District |
0 |
Bullock Creek School District |
0.1902307 |
|
Tuscola ISD - District created f |
0 |
Walkerville Public Schools |
0.189461 |
|
Van Buren ISD - District created |
0 |
Lake Shore Public Schools (Macom |
0.1876862 |
|
Washtenaw ISD - District created |
0 |
Milan Area Schools |
0.1869294 |
|
Wayne RESA - District created fr |
0 |
Homer Community School District |
0.1835621 |
|
West Shore Educational Service D |
0 |
Port Hope Community Schools |
0.1826151 |
|
Wexford-Missaukee ISD - District |
0 |
Reese Public Schools |
0.182154 |
|
Whiteford Agricultural School Di |
0 |
Boyne Falls Public School Distri |
0.1806984 |
|
Flint, School District of the Ci |
0.000013 |
Southgate Community School Distr |
0.1803922 |
|
Northville Public Schools |
0.0000157 |
Leland Public School District |
0.1788193 |
|
Monroe Public Schools |
0.000018 |
Ellsworth Community School |
0.1742559 |
|
Forest Hills Public Schools |
0.0000361 |
Brandywine Community Schools |
0.1730274 |
|
Coldwater Community Schools |
0.0000387 |
Chassell Township School Distric |
0.172523 |
|
Orchard View Schools |
0.0000432 |
Mackinaw City Public Schools |
0.1722113 |
|
Kentwood Public Schools |
0.0000543 |
Northport Public School District |
0.1714843 |
|
East Grand Rapids Public Schools |
0.0000832 |
Pickford Public Schools |
0.1713319 |
|
Bentley Community School Distric |
0.0001524 |
Clinton Community Schools |
0.1684491 |
|
Glen Lake Community Schools |
0.000159 |
Lakeview Sch. District (Calhoun) |
0.167474 |
|
Grand Blanc Community Schools |
0.0001671 |
Berkley School District |
0.1663842 |
|
Livonia Public Schools School Di |
0.000169 |
Potterville Public Schools |
0.1656566 |
|
Rockford Public Schools |
0.0001887 |
Springport Public Schools |
0.1649686 |
|
Baldwin Community Schools |
0.0002073 |
Tekonsha Community Schools |
0.1648171 |
|
Alpena Public Schools |
0.00024 |
Marcellus Community Schools |
0.1643792 |
|
Crestwood School District |
0.0003084 |
McBain Rural Agricultural School |
0.1643627 |
|
Freeland Community School Distri |
0.0003393 |
Alcona Community Schools |
0.1641555 |
|
Mt. Morris Consolidated Schools |
0.0003483 |
Iron Mountain Public Schools |
0.1605676 |
|
Zeeland Public Schools |
0.0003835 |
Big Bay De Noc School District |
0.1596613 |
|
Byron Center Public Schools |
0.0004095 |
Holt Public Schools |
0.159248 |
|
Wyoming Public Schools |
0.0004306 |
Camden-Frontier School |
0.1586438 |
|
Reeths-Puffer Schools |
0.0004502 |
Wyandotte, School District of th |
0.1585435 |
|
Monroe ISD - District created fr |
0.0004647 |
Warren Woods Public Schools |
0.1578923 |
|
Flushing Community Schools |
0.0005379 |
Waverly Community Schools |
0.1561055 |
|
Centreville Public Schools |
0.0005517 |
Napoleon Community Schools |
0.1559368 |
|
Kenowa Hills Public Schools |
0.0006973 |
Leslie Public Schools |
0.155762 |
|
Kent City Community Schools |
0.0007109 |
Webberville Community Schools |
0.1555311 |
|
Trenton Public Schools |
0.0008478 |
|||
Total |
0.0000661 |
Total |
0.2936637 |
|
A3: District Rates of Student Exit Under SoC
A3 Top and bottom quintiles of districts based on proportion of students exiting the district through SoC based on average per year proportion. |
||||
Quintile 1 |
Quintile 5 |
|||
Allegan Area Educational Service |
0 |
Grant Township S/D #2 |
0.2083333 |
|
Allendale Public Schools |
0 |
Bloomfield Township S/D #7F |
0.1875 |
|
Alpena Public Schools |
0 |
Oneida Township S/D #3 |
0.1861284 |
|
Alpena-Montmorency-Alcona ESD - |
0 |
Highland Park City Schools |
0.1521386 |
|
Atherton Community Schools |
0 |
Bangor Township S/D #8 |
0.1423525 |
|
Barry ISD - District created fro |
0 |
Sigel Township S/D #3F |
0.1406712 |
|
Bay-Arenac ISD - District create |
0 |
Free Soil Community Schools |
0.1274565 |
|
Beaver Island Community School |
0 |
Verona Township S/D #1F |
0.1177929 |
|
Beecher Community School Distric |
0 |
Glenn Public School District |
0.1118257 |
|
Belding Area School District |
0 |
Colfax Township S/D #1F |
0.1054088 |
|
Bendle Public Schools |
0 |
Moran Township School District |
0.1049691 |
|
Bentley Community School Distric |
0 |
Westwood Community School Distri |
0.0986202 |
|
Berrien RESA - District created |
0 |
Mar Lee School District |
0.0965102 |
|
Bloomfield Hills Schools |
0 |
Sodus Township S/D #5 |
0.0922846 |
|
Bois Blanc Pines School District |
0 |
Hagar Township S/D #6 |
0.0834192 |
|
Branch ISD - District created fr |
0 |
Sigel Township S/D #6 |
0.0785985 |
|
Byron Center Public Schools |
0 |
Clintondale Community Schools |
0.0748343 |
|
C.O.O.R. ISD - District created |
0 |
River Rouge, School District of |
0.0719892 |
|
Caledonia Community Schools |
0 |
AuTrain-Onota Public Schools |
0.0712547 |
|
Calhoun ISD - District created f |
0 |
Wells Township School District |
0.0683517 |
|
Carman-Ainsworth Community Schoo |
0 |
Church School District |
0.0658505 |
|
Centreville Public Schools |
0 |
Oak Park, School District of the |
0.0634621 |
|
Charlevoix-Emmet ISD - District |
0 |
School District of the City of I |
0.0621733 |
|
Cheb-Otsego-Presque Isle ESD - D |
0 |
Madison District Public Schools |
0.0612982 |
|
Chelsea School District |
0 |
Galien Township School District |
0.0600117 |
|
Clare-Gladwin Regional Education |
0 |
School District of Ypsilanti |
0.0562567 |
|
Clinton County RESA - District c |
0 |
Burt Township School District |
0.055753 |
|
Clio Area School District |
0 |
Ecorse Public Schools |
0.052441 |
|
Coldwater Community Schools |
0 |
Vandercook Lake Public Schools |
0.0520655 |
|
Comstock Public Schools |
0 |
Mount Clemens Community School D |
0.0516361 |
|
Copper Country ISD - District cr |
0 |
Dearborn Heights School District |
0.050126 |
|
Dearborn City School District |
0 |
Buena Vista School District |
0.0496493 |
|
Delta-Schoolcraft ISD - District |
0 |
Carrollton Public Schools |
0.0478989 |
|
Dexter Community School District |
0 |
Pennfield Schools |
0.0459872 |
|
Dickinson-Iron ISD - District cr |
0 |
Mason County Eastern Schools |
0.0453666 |
|
Dundee Community Schools |
0 |
Pittsford Area Schools |
0.0452079 |
|
East Grand Rapids Public Schools |
0 |
Caseville Public Schools |
0.045162 |
|
Eastern Upper Peninsula ISD - Di |
0 |
North Muskegon Public Schools |
0.0445727 |
|
Easton Township S/D #6 |
0 |
Eau Claire Public Schools |
0.0431992 |
|
Eaton ISD - District created fro |
0 |
Willow Run Community Schools |
0.0424224 |
|
Education Achievement System-EAS |
0 |
Tekonsha Community Schools |
0.0419866 |
|
Excelsior Township S/D #1 |
0 |
Elm River Township School Distri |
0.040625 |
|
Flint, School District of the Ci |
0 |
Michigan Center School District |
0.0393156 |
|
Forest Hills Public Schools |
0 |
Britton Deerfield Schools |
0.0391212 |
|
Freeland Community School Distri |
0 |
Jonesville Community Schools |
0.0384725 |
|
Genesee ISD - District created f |
0 |
Madison School District (Lenawee |
0.0382965 |
|
Glen Lake Community Schools |
0 |
Elk Rapids Schools |
0.0371671 |
|
Godfrey-Lee Public Schools |
0 |
Walkerville Public Schools |
0.0350796 |
|
Godwin Heights Public Schools |
0 |
Clawson Public Schools |
0.0349824 |
|
Gogebic-Ontonagon ISD - District |
0 |
Mason County Central Schools |
0.0347798 |
|
Grand Blanc Community Schools |
0 |
Coloma Community Schools |
0.034508 |
|
Grand Rapids Public Schools |
0 |
Frankfort-Elberta Area Schools |
0.0339422 |
|
Grandville Public Schools |
0 |
East Jackson Community Schools |
0.0335773 |
|
Grant Public School District |
0 |
Ellsworth Community School |
0.033437 |
|
Gratiot-Isabella RESD - District |
0 |
North Adams-Jerome Public School |
0.0330536 |
|
Greenville Public Schools |
0 |
Boyne Falls Public School Distri |
0.0318803 |
|
Grosse Pointe Public Schools |
0 |
Pentwater Public School District |
0.0316826 |
|
Harbor Springs School District |
0 |
Corunna Public Schools |
0.0316355 |
|
Hart Public School District |
0 |
Ashley Community Schools |
0.0315964 |
|
Hillsdale ISD - District created |
0 |
Fulton Schools |
0.0313476 |
|
Huron ISD - District created fro |
0 |
Mancelona Public Schools |
0.0311067 |
|
Ida Public School District |
0 |
Alba Public Schools |
0.030784 |
|
Ingham ISD - District created fr |
0 |
Beal City Public Schools |
0.0306298 |
|
Ionia ISD - District created fro |
0 |
Ionia Township S/D #2 |
0.0305556 |
|
Ionia Public Schools |
0 |
Alcona Community Schools |
0.0304709 |
|
Iosco RESA - District created fr |
0 |
Kaleva Norman Dickson School Dis |
0.0302392 |
|
Jackson ISD - District created f |
0 |
Peck Community School District |
0.0300397 |
|
Kalamazoo Public Schools |
0 |
Reading Community Schools |
0.029991 |
|
Kalamazoo RESA - District create |
0 |
Palo Community School District |
0.0297695 |
|
Kelloggsville Public Schools |
0 |
Melvindale-North Allen Park Scho |
0.0295577 |
|
Kenowa Hills Public Schools |
0 |
Watervliet School District |
0.0291021 |
|
Kent ISD - District created from |
0 |
Houghton-Portage Township School |
0.0286834 |
|
Kentwood Public Schools |
0 |
St. Louis Public Schools |
0.028435 |
|
Lake Orion Community Schools |
0 |
Riverview Community School Distr |
0.0284168 |
|
Lakeview Community Schools (Mont |
0 |
Waldron Area Schools |
0.0283662 |
|
Lapeer ISD - District created fr |
0 |
Onekama Consolidated Schools |
0.0279136 |
|
Lenawee ISD - District created f |
0 |
Sand Creek Community Schools |
0.0277243 |
|
Lewis Cass ISD - District create |
0 |
Webberville Community Schools |
0.0275374 |
|
Livingston ESA - District create |
0 |
Bullock Creek School District |
0.0272022 |
|
Mackinac Island Public Schools |
0 |
Owendale-Gagetown Area School Di |
0.0271156 |
|
Macomb ISD - District created fr |
0 |
Homer Community School District |
0.0269304 |
|
Manistee ISD - District created |
0 |
Clare Public Schools |
0.0264578 |
|
Marquette-Alger RESA - District |
0 |
St. Ignace Area Schools |
0.0260264 |
|
Mecosta-Osceola ISD - District c |
0 |
Covert Public Schools |
0.0259624 |
|
Menominee ISD - District created |
0 |
Concord Community Schools |
0.0258617 |
|
Michigan Department of Human Ser |
0 |
Engadine Consolidated Schools |
0.0257527 |
|
Midland County Educational Servi |
0 |
Swan Valley School District |
0.0255322 |
|
Monroe Public Schools |
0 |
Brandywine Community Schools |
0.0255211 |
|
Montcalm Area ISD - District cre |
0 |
Ferndale Public Schools |
0.025222 |
|
Mt. Morris Consolidated Schools |
0 |
New Buffalo Area Schools |
0.0248937 |
|
Muskegon Area ISD - District cre |
0 |
Johannesburg-Lewiston Area Schoo |
0.0244121 |
|
Newaygo County RESA - District c |
0 |
Carsonville-Port Sanilac School |
0.0243989 |
|
Northview Public Schools |
0 |
Bridgman Public Schools |
0.0243344 |
|
Northville Public Schools |
0 |
Sigel Township S/D #4F |
0.0240275 |
|
Nottawa Community School |
0 |
Essexville-Hampton Public School |
0.0236242 |
|
Novi Community School District |
0 |
Marcellus Community Schools |
0.0235769 |
|
Oakland Schools - District creat |
0 |
Saugatuck Public Schools |
0.0235635 |
|
Oakridge Public Schools |
0 |
Bark River-Harris School Distric |
0.0235208 |
|
Ottawa Area ISD - District creat |
0 |
Kingsley Area Schools |
0.0234447 |
|
Portland Public Schools |
0 |
Kingston Community School Distri |
0.0232658 |
|
Rochester Community School Distr |
0 |
Hanover-Horton School District |
0.0231532 |
|
Rockford Public Schools |
0 |
Morrice Area Schools |
0.0230438 |
|
Saginaw ISD - District created f |
0 |
Lakeview Public Schools (Macomb) |
0.0230112 |
|
Sanilac ISD - District created f |
0 |
Port Hope Community Schools |
0.0227382 |
|
Saranac Community Schools |
0 |
Camden-Frontier School |
0.0227123 |
|
School District of the City of B |
0 |
Southgate Community School Distr |
0.0225549 |
|
Shiawassee Regional ESD - Distri |
0 |
Hale Area Schools |
0.0225312 |
|
Sparta Area Schools |
0 |
Fruitport Community Schools |
0.0224993 |
|
St. Clair County RESA - District |
0 |
Western School District |
0.0224721 |
|
St. Johns Public Schools |
0 |
Milan Area Schools |
0.022368 |
|
St. Joseph County ISD - District |
0 |
Fraser Public Schools |
0.0220231 |
|
Summerfield Schools |
0 |
Hillsdale Community Schools |
0.0219118 |
|
Traverse Bay Area ISD - District |
0 |
Alma Public Schools |
0.0219048 |
|
Trenton Public Schools |
0 |
Lake Shore Public Schools (Macom |
0.0214821 |
|
Tuscola ISD - District created f |
0 |
Litchfield Community Schools |
0.0214821 |
|
Van Buren ISD - District created |
0 |
Mid Peninsula School District |
0.0213597 |
|
Washtenaw ISD - District created |
0 |
Bear Lake Schools |
0.0212758 |
|
Wayne RESA - District created fr |
0 |
River Valley School District |
0.0212126 |
|
West Shore Educational Service D |
0 |
Hartford Public Schools |
0.0211516 |
|
Wexford-Missaukee ISD - District |
0 |
Carney-Nadeau Public Schools |
0.0210167 |
|
Whiteford Agricultural School Di |
0 |
Bad Axe Public Schools |
0.020982 |
|
Wyoming Public Schools |
0 |
Total |
0.0463663 |
|
A4: District Rates of Student Entry Under SoC
A4 Top and bottom quintiles of districts based on proportion of students exiting the district through SoC based on average per year proportion. |
||||
Quintile 1 |
Quintile 5 |
|||
Alpena-Montmorency-Alcona ESD - |
0 |
Bloomfield Township S/D #7F |
0.2083333 |
|
Bay-Arenac ISD - District create |
0 |
Free Soil Community Schools |
0.2051317 |
|
Beaver Island Community School |
0 |
Grant Township S/D #2 |
0.1791667 |
|
Bois Blanc Pines School District |
0 |
Big Jackson School District |
0.1024743 |
|
C.O.O.R. ISD - District created |
0 |
Hillsdale ISD - District created |
0.1016743 |
|
Charlevoix-Emmet ISD - District |
0 |
Galien Township School District |
0.0953223 |
|
Clare-Gladwin Regional Education |
0 |
Washtenaw ISD - District created |
0.0855716 |
|
Delta-Schoolcraft ISD - District |
0 |
Wexford-Missaukee ISD - District |
0.0810229 |
|
Eastern Upper Peninsula ISD - Di |
0 |
Hagar Township S/D #6 |
0.0794033 |
|
Easton Township S/D #6 |
0 |
Arvon Township School District |
0.0788691 |
|
Eaton ISD - District created fro |
0 |
Allegan Area Educational Service |
0.0645549 |
|
Education Achievement System-EAS |
0 |
Sigel Township S/D #3F |
0.0625 |
|
Gogebic-Ontonagon ISD - District |
0 |
School District of the City of I |
0.0616338 |
|
Ionia ISD - District created fro |
0 |
Vanderbilt Area Schools |
0.0613781 |
|
Iosco RESA - District created fr |
0 |
Jackson ISD - District created f |
0.0605359 |
|
Kent ISD - District created from |
0 |
Excelsior Township S/D #1 |
0.0594691 |
|
Lapeer ISD - District created fr |
0 |
Wells Township School District |
0.0590909 |
|
Livingston ESA - District create |
0 |
Gratiot-Isabella RESD - District |
0.0582909 |
|
Manistee ISD - District created |
0 |
Willow Run Community Schools |
0.055369 |
|
Michigan Department of Human Ser |
0 |
Palo Community School District |
0.0544572 |
|
Muskegon Area ISD - District cre |
0 |
North Adams-Jerome Public School |
0.0540321 |
|
Newaygo County RESA - District c |
0 |
AuTrain-Onota Public Schools |
0.0535644 |
|
Oakland Schools - District creat |
0 |
Mar Lee School District |
0.0515622 |
|
Sanilac ISD - District created f |
0 |
Verona Township S/D #1F |
0.0500129 |
|
St. Clair County RESA - District |
0 |
Mount Clemens Community School D |
0.049348 |
|
West Shore Educational Service D |
0 |
Waldron Area Schools |
0.0492695 |
|
East Grand Rapids Public Schools |
0.0002106 |
East Detroit Public Schools |
0.0489211 |
|
Forest Hills Public Schools |
0.0003276 |
Carney-Nadeau Public Schools |
0.0477004 |
|
Caledonia Community Schools |
0.000501 |
Ecorse Public Schools |
0.0473695 |
|
Ida Public School District |
0.0005922 |
Mason County Central Schools |
0.0460346 |
|
Rockford Public Schools |
0.0007086 |
East Jackson Community Schools |
0.0459338 |
|
Bloomfield Hills Schools |
0.0007737 |
Stanton Township Public Schools |
0.0456461 |
|
Northview Public Schools |
0.000792 |
River Rouge, School District of |
0.0454181 |
|
Nottawa Community School |
0.0008013 |
Litchfield Community Schools |
0.04403 |
|
Zeeland Public Schools |
0.0009188 |
Madison District Public Schools |
0.0439859 |
|
Rochester Community School Distr |
0.0010444 |
Albion Public Schools |
0.0433691 |
|
West Ottawa Public School Distri |
0.0010609 |
Sigel Township S/D #4F |
0.0424936 |
|
Oakridge Public Schools |
0.0010834 |
Pittsford Area Schools |
0.0410968 |
|
Northville Public Schools |
0.0010964 |
Madison School District (Lenawee |
0.0407942 |
|
Kentwood Public Schools |
0.0011102 |
Hillsdale Community Schools |
0.0398638 |
|
School District of the City of B |
0.0011379 |
Camden-Frontier School |
0.0394626 |
|
Troy School District |
0.0011626 |
Clintondale Community Schools |
0.0393992 |
|
Byron Center Public Schools |
0.0013038 |
Buena Vista School District |
0.0390851 |
|
Grosse Ile Township Schools |
0.001312 |
School District of Ypsilanti |
0.0388958 |
|
Holland City School District |
0.0013193 |
Ithaca Public Schools |
0.0383295 |
|
Grand Rapids Public Schools |
0.0013519 |
Glenn Public School District |
0.038086 |
|
Coopersville Area Public School |
0.0013962 |
Tekonsha Community Schools |
0.0380133 |
|
Dexter Community School District |
0.0013981 |
Oak Park, School District of the |
0.0379928 |
|
Bedford Public Schools |
0.0015203 |
River Valley School District |
0.0378148 |
|
Novi Community School District |
0.001539 |
Pennfield Schools |
0.0376898 |
|
Monroe Public Schools |
0.0015905 |
Michigan Center School District |
0.0374102 |
|
Mackinac Island Public Schools |
0.0016026 |
Highland Park City Schools |
0.0373719 |
|
Chelsea School District |
0.0016031 |
Kaleva Norman Dickson School Dis |
0.0370596 |
|
Godfrey-Lee Public Schools |
0.001696 |
Van Dyke Public Schools |
0.0369943 |
|
Dickinson-Iron ISD - District cr |
0.0017007 |
Cheb-Otsego-Presque Isle ESD - D |
0.0367884 |
|
Kenowa Hills Public Schools |
0.00171 |
Bellevue Community Schools |
0.0365133 |
|
Allendale Public Schools |
0.0017233 |
Westwood Community School Distri |
0.034731 |
|
Summerfield Schools |
0.0018225 |
Ashley Community Schools |
0.0345285 |
|
Portland Public Schools |
0.0018271 |
Baldwin Community Schools |
0.0345259 |
|
Charlevoix Public Schools |
0.0018657 |
Elk Rapids Schools |
0.0344416 |
|
Grosse Pointe Public Schools |
0.0018693 |
Powell Township Schools |
0.0342955 |
|
Reeths-Puffer Schools |
0.001918 |
Martin Public Schools |
0.0342876 |
|
Clarkston Community School Distr |
0.001947 |
Morrice Area Schools |
0.03406 |
|
Plymouth-Canton Community School |
0.0019801 |
Jonesville Community Schools |
0.0330593 |
|
Farmington Public School Distric |
0.0020018 |
Reading Community Schools |
0.0328683 |
|
Lake Orion Community Schools |
0.0020129 |
Akron-Fairgrove Schools |
0.0327112 |
|
Menominee Area Public Schools |
0.0020918 |
Concord Community Schools |
0.0326558 |
|
Spring Lake Public Schools |
0.0021806 |
Roseville Community Schools |
0.0324279 |
|
Portage Public Schools |
0.0022 |
Vandercook Lake Public Schools |
0.0321873 |
|
Saline Area Schools |
0.0022719 |
South Lake Schools |
0.0318487 |
|
Kelloggsville Public Schools |
0.0022796 |
Watervliet School District |
0.0315064 |
|
Alpena Public Schools |
0.0022954 |
Saginaw, School District of the |
0.0313007 |
|
Whiteford Agricultural School Di |
0.0023814 |
Sigel Township S/D #6 |
0.03125 |
|
Harbor Springs School District |
0.0023982 |
Arenac Eastern School District |
0.0311683 |
|
Trenton Public Schools |
0.0024202 |
Caseville Public Schools |
0.0309582 |
|
Flushing Community Schools |
0.0024741 |
Calhoun ISD - District created f |
0.0307025 |
|
Kent City Community Schools |
0.0024803 |
Mason County Eastern Schools |
0.0306627 |
|
Grandville Public Schools |
0.0025205 |
Alba Public Schools |
0.0305725 |
|
Ann Arbor Public Schools |
0.0025613 |
Addison Community Schools |
0.0304298 |
|
Sparta Area Schools |
0.0025876 |
Huron ISD - District created fro |
0.030042 |
|
Grand Blanc Community Schools |
0.0025883 |
Benton Harbor Area Schools |
0.0296801 |
|
Orchard View Schools |
0.0026003 |
Kingston Community School Distri |
0.0295173 |
|
Bendle Public Schools |
0.0026641 |
Deerfield Public Schools |
0.0293734 |
|
Kalamazoo Public Schools |
0.0026757 |
Adrian, School District of the C |
0.0291992 |
|
Mona Shores Public School Distri |
0.0026829 |
Onekama Consolidated Schools |
0.0288534 |
|
Grant Public School District |
0.0026959 |
Chassell Township School Distric |
0.0287887 |
|
Ionia Public Schools |
0.0027327 |
Covert Public Schools |
0.0287607 |
|
Public Schools of Petoskey |
0.0028006 |
Eau Claire Public Schools |
0.0287478 |
|
Walled Lake Consolidated Schools |
0.0028117 |
Lenawee ISD - District created f |
0.0287018 |
|
Wyoming Public Schools |
0.002816 |
Coloma Community Schools |
0.0285753 |
|
Boyne City Public Schools |
0.0028364 |
Wayne RESA - District created fr |
0.0285714 |
|
Shiawassee Regional ESD - Distri |
0.0028409 |
Suttons Bay Public Schools |
0.0283842 |
|
Dearborn City School District |
0.0028415 |
Napoleon Community Schools |
0.0283193 |
|
Centreville Public Schools |
0.0028828 |
Britton Deerfield Schools |
0.0281202 |
|
Dundee Community Schools |
0.0029815 |
Owendale-Gagetown Area School Di |
0.0280282 |
|
Davison Community Schools |
0.0030338 |
Mayville Community School Distri |
0.0279674 |
|
Livonia Public Schools School Di |
0.0031051 |
Hudson Area Schools |
0.0278634 |
|
Whitehall District Schools |
0.0031438 |
Elm River Township School Distri |
0.0277778 |
|
Saranac Community Schools |
0.0031569 |
Vestaburg Community Schools |
0.0275943 |
|
St. Johns Public Schools |
0.0031698 |
St. Louis Public Schools |
0.0272961 |
|
Sturgis Public Schools |
0.0031724 |
Lewis Cass ISD - District create |
0.0270367 |
|
South Lyon Community Schools |
0.0032992 |
Corunna Public Schools |
0.0270328 |
|
DeWitt Public Schools |
0.0032996 |
Bridgeport-Spaulding Community S |
0.0269869 |
|
Jefferson Schools (Monroe) |
0.0033048 |
Athens Area Schools |
0.0268383 |
|
Clio Area School District |
0.003388 |
Saginaw ISD - District created f |
0.0267022 |
|
Fremont Public School District |
0.0034105 |
Carsonville-Port Sanilac School |
0.0264459 |
|
Belding Area School District |
0.0035077 |
Homer Community School District |
0.0261253 |
|
Hudsonville Public School Distri |
0.0036278 |
Ellsworth Community School |
0.0261226 |
|
Avondale School District |
0.0036971 |
Brandywine Community Schools |
0.0258776 |
|
Kalamazoo RESA - District create |
0.0037409 |
Hanover-Horton School District |
0.0258545 |
|
Grand Haven Area Public Schools |
0.0037586 |
Mid Peninsula School District |
0.025787 |
|
Bentley Community School Distric |
0.0037667 |
Webberville Community Schools |
0.0257721 |
|
Coldwater Community Schools |
0.0038676 |
Vassar Public Schools |
0.0255718 |
|
Colfax Township S/D #1F |
0.0039063 |
Fitzgerald Public Schools |
0.0254491 |
|
North Muskegon Public Schools |
0.0039914 |
Pentwater Public School District |
0.0253597 |
|
Macomb ISD - District created fr |
0.0040754 |
Berrien RESA - District created |
0.0252403 |
|
West Bloomfield School District |
0.0041482 |
Buchanan Community Schools |
0.0251859 |
|
Beecher Community School Distric |
0.0042361 |
Ionia Township S/D #2 |
0.025 |
|
Branch ISD - District created fr |
0.0045417 |
Mecosta-Osceola ISD - District c |
0.0247044 |
|
Huron Valley Schools |
0.0045512 |
Rapid River Public Schools |
0.0242996 |
|
Berkley School District |
0.0045864 |
Hale Area Schools |
0.0242859 |
|
St. Joseph County ISD - District |
0.0046296 |
Total |
0.0433109 |
|
Total |
0.0018692 |
|||
COLLEGE LINKS
