(Percentages represented 0-1)

The Arizona Department of Education provides enrollment data every year, which includes ethnicity breakdowns. The Arizona Center for Investigative Reporting analysis of ethnicity in schools began with this data, and we were able to see that the ethnic profile of charter schools and district schools across the state did not match.

But we wanted to look at schools more individually, which presented a challenge. Because every community is uniquely comprised of different ethnicities, we needed a way to look at how each of Arizona’s nearly 2,000 schools compared to the area around it. In order to do that, we plotted each school in its census tract, then isolated the current school age population from 2010 census data (ages 0 – 13, since the 2010 census data is now about five years old). With ethnicity information for every school, and every school plotted in a census tract with ethnicity data for only the school age population, we could then calculate the ethnic composition differences between each school and its respective census tract.

(Percentages represented 0-1)

We also drew a 10-mile radius around every school, then calculated the average school ethnicity composition for every district school within that radius. This accounts for the sometimes irregular shape of census tracts, or the possibility that a school might be in a corner of one census tract, meaning it might be more accurate to compare it to nearby areas, but not necessarily the tract it is actually in. It also accounts for a reasonable driving distance for parents who take advantage of open enrollment, which is available for both district and charter schools.

These two analyses produced similar results, and showed that, on average, Hispanic students are slightly over-represented in district schools when compared to their census tract (+2.4 percent) and show near parity when compared to district schools within 10 miles (-0.9 percent). But in charter schools, Hispanic students are under-represented on both measures (-3.1 percent when compared to their census tract and -8.2 percent when compared to district schools within 10 miles).

(Percentages represented 0-1)

On average, white students are somewhat under-represented in district schools when compared to their census tract (-4.9 percent) and show near parity when compared to district schools within ten miles (+1.4 percent). But in charter schools, white students show near parity when compared to their census tract (-0.8 percent) and are over-represented when compared to district schools within 10 miles (+6 percent).

We also found a roughly 50 percent stronger relationship between census tract demographics and district school demographics than that for charter schools.

Several sources we spoke with suggested isolating elementary schools from high schools for this kind of analysis, as they thought the trend we observed would be more pronounced in elementary schools, and less so in high schools. This was not possible, though, because schools have so much variety in the grades they include, and ethnicity data is only available by school or grade level statewide, not by school and grade together.

The Arizona Department of Education also redacts student counts for ethnicity when there 11 or fewer students in that category. In order to come up with an approximation for when this occurred, we calculated the total number of students unaccounted for by individual ethnicity, then evenly distributed the missing students to each ethnicity that had a redacted count.

We categorized all of the nearly 600 charter schools based on the mission statements provided on their websites.