One way to understand what students and parents are seeking through Schools of Choice is to examine their choices — what economists call “revealed preference.” The 1999 Michigan State University study teased out student preferences by comparing the characteristics of a school district that a student left to the characteristics of the school district that student enrolled in through Schools of Choice. This type of analysis allows researchers to look at revealed student, or perhaps, parental preferences.
A selection of the 1999 MSU study findings are reproduced in Graphic 9 and compared to findings from the 2011-12 school year, using a similar methodology.[*] Each choice is weighted by the number of students making it; for example, if 11 students used Schools of Choice to transfer from a school in Detroit Public Schools to a school in Dearborn, that transfer would count 11 times toward the averages reported below. If just one student used Schools of Choice to transfer from a school in Dearborn to a DPS school, that transfer would count just once toward the average below.
Averages were computed for Schools of Choice participation statewide. A separate column shows the analysis for all districts, excluding students who left Detroit Public Schools, since that district — by far, the largest in the state — may have a disproportionate impact on the overall results.
During the 2011-12 school year, students and parents appeared to make similar choices to those made more than a decade ago. In general, they moved to districts with better academic outcomes overall. Michigan students and parents used Schools of Choice in 2011-12 to move to districts with higher average proficiency rates on the Michigan Educational Assessment Program tests (measured for seventh graders), higher graduation rates and lower dropout rates. The average difference between MEAP proficiency rates of districts students exited and districts students entered is slightly larger in magnitude to the difference observed in the 1999 MSU study.
Statewide, students tended to choose districts with higher average teacher salaries, but when DPS was excluded, the districts students entered appeared to pay roughly the same amount. Students also tended to leave larger districts for smaller ones, choosing districts, on average, that had 4,622 fewer students. The impact of DPS was large here again: Students and parents chose slightly larger districts on average, when DPS is excluded from the analysis.
Graphic 9: Revealed Student/Parental Preferences Through Schools of Choice, 1999, 2011-12
District Characteristic |
1999 |
2011-12 |
1999, ex. DPS |
2011-12, ex. DPS |
Math proficiency rate |
5.1 |
8.0 |
6.9 |
8.4 |
Reading proficiency rate |
3.2 |
8.9 |
5.1 |
9.6 |
Graduation rate |
9.4 |
5.2 |
6.6 |
6.8 |
Drop-out rate |
-3.6 |
-2.9 |
-2.2 |
-3.4 |
Pupil-teacher ratio |
0.3 |
0.9 |
-0.1 |
0.7 |
Expenditures per pupil |
-$314 |
-$1,773 |
-$377 |
-$899 |
Mean teacher salary |
$1,766 |
$1,385 |
$174 |
$1,063 |
District enrollment |
-15,571 |
-4,682 |
-1,737 |
326 |
Percent free/reduced lunch |
-9.8 |
-10.7 |
-10.9 |
-11.2 |
Percent African-American |
-12.2 |
-11.1 |
-12.0 |
-10.8 |
Source: Arsen, et al., “School Choice Policies in Michigan: The Rules Matter” (1999),CEPI, MDE, NCES.
As observed in the MSU study, students continued to move to districts with a lower proportion of students eligible for free or reduced-price lunch, as well as to districts with a lower percentage of African-American students. This trend may be related to students’ tendency to move to districts with higher average MEAP proficiency rates — students from higher-income families tend to do better on standardized tests on average.[41]
A familiar criticism regarding the expansion of School of Choice is that students will use it to leave city districts for suburban ones. This is an observed trend: Of students using Schools of Choice to leave city districts, 74 percent opted for a suburban district. This proportion drops to 66 percent if DPS students are excluded from the analysis. This is not surprising since a suburban school district is by definition near an urban center.
However, the statewide trend is different. Graphic 10 compares locale categories among educating districts (where Schools of Choice students opted to attend) to resident districts (the district students chose to leave). Most Michigan students are using Schools of Choice to attend a district within a similar locale to their resident district’s.
Graphic 10: Schools of Choice Movement by Locale Type, 2011-12
Resident District |
|||||
Educating District |
City |
Rural |
Suburb |
Town |
Total* |
City |
3,453 |
969 |
4,182 |
245 |
8,849 |
Suburb |
19,871 |
3,169 |
25,302 |
973 |
49,315 |
Town |
725 |
7,187 |
1,156 |
3,263 |
12,331 |
Rural |
2,818 |
14,795 |
2,859 |
8,334 |
28,806 |
Total* |
26,867 |
26,120 |
33,499 |
12,815 |
99,301 |
Source: Center of Educational Performance and Information, National Center for Education Statistics.
* Data from districts without locale codes was excluded.
Of the more than 49,000 Schools of Choice students who attended suburban districts, for example, more than half left a different suburban district. Similarly, more than half of Schools of Choice students attending a rural district left a different rural district.
Many students who left a town district opted for a school in a rural one. Of the 12,815 students who left a town district, 8,334 — 65 percent — left for a rural district. Meanwhile, about 28 percent of students who left a rural district chose a school in a town district.
The tendency of students living in town districts to choose schools located in rural districts may be due simply to the location and number of rural districts. There are more than three times as many rural districts as there are town districts, and town districts frequently border, or are even surrounded by, rural districts. Similarly, most city districts are surrounded, or nearly surrounded, by suburban districts. Of the 34 districts categorized as city, all but two are adjacent to at least one suburban district.
[*] CEPI’s 2011-12 Schools of Choice dataset was matched with data from MDE’s MEAP proficiency file for 2011-12, as well as with graduation and dropout rate data, student-to-teacher ratio data, expenditure data, teacher salary data, enrollment data, free and reduced-price lunch data and student ethnicity data: “Non-Resident Student Research Tool” (Center for Educational Performance and Information, 2012), accessed July 25, 2012, http://goo.gl/HdmaEU; “Fall 2011 MEAP Four Year Comparison (Gap Analysis)” (Michigan Department of Education, 2012), accessed Nov. 15, 2013, http://goo.gl/uexp2; “2012 Cohort Four-Year, 2011 Cohort Five-Year and 2010 Six-Year Graduation and Dropout Rates Including Subgroups” (Center for Educational Performance and Information, 2012), accessed Nov. 15, 2013, http://goo.gl/MGimd; National Center for Education Statistics, “Common Core of Data” (Institute of Education Sciences, 2011), http://goo.gl/i2hJnL; “Public Student Counts (Headcount Data)” (Center for Educational Performance and Information), accessed Oct. 24, 2013, http://goo.gl/Uj3IKA.