About this analysis

The Star Tribune requested data from the Minnesota Department of Education summarizing the number of children open enrolling in another district and the number attending charter schools. Both sets of data covered the 1999-2000 to 2016-2017 school years, and summarized by the race of the students, the district they live in and the district or charter they attended.

The state education department redacted the data for any instance of fewer than three students of a particular race leaving one district to attend another. The Star Tribune used other non-redacted open enrollment data that's not broken down by race to generate more accurate estimates of the total number of children open enrolling. As a result of these redactions, the Star Tribune's analysis is conservative in estimating how many children are opting out of their home district, and the racial breakdowns will be imperfect for smaller school districts, particularly those with small populations within each minority group.

Open enrolling in another district and attending a charter school are viewed differently by policymakers and tracked separately by the state education department. For this analysis, the Star Tribune combined the two in order to estimate the share of students living in a district but not attending the district schools. To estimate the number of children living in each district, we started with the district's total enrollment, subtracted students coming in via open enrollment and then added back in the students leaving for charter schools or open enrolling elsewhere. The analysis did not include children who attend private school or who are homeschooled. Other data maintained by the Department of Education shows the number of children opting for private schools has been declining in recent years, and now accounts for about 7 percent of the state's K-12 students; homeschooled children account for about 2 percent.

We chose to not use U.S. Census Bureau data estimating school-age population because that data, particularly when broken down by race, has high margins of error.