AZCIR’s Report for America Fund
AZCIR partnered with Report for America to hire Shaena Montanari as our newsroom’s investigative health reporter. RFA supports half of Shaena’s salary, challenging AZCIR to cover the rest through community donations.
Will you help AZCIR raise $25,000 to cover a portion of Shaena’s salary, and to help offset reporting costs such as statewide travel and public records?
AZCIR analyzed federal and state education data to identify statewide trends in the use of 504 plans, a protection that comes from a federal law designed to help students with disabilities. The analysis tracked the disability plans across schools with different socioeconomic and racial makeups, and uncovered inequalities in access that favor Arizona’s wealthiest and whitest traditional public school students.
AZCIR used data from the U.S. Department of Education’s 2017-2018 Civil Rights Data Collection (CRDC), the most recent available. The CRDC, a federal survey conducted every other year by the department’s Office for Civil Rights, includes information from nearly 100,000 schools across the U.S. The survey covers topics ranging from special education to student ethnicity, which the education department uses to ensure federally-funded schools are in compliance with civil rights laws.
The data collected provides enrollment figures, by race, and rates of 504 plans in Arizona public schools that receive federal funds. AZCIR combined the survey with the Arizona Department of Education’s free and reduced lunch data—an indicator of student wealth. The combined data enabled AZCIR to explore trends between student wealth and access to the disability protection.
Shortcomings at the federal, state and school level make securing a 504 plan easier for students with disabilities who attend Arizona schools that are wealthier and whiter. The disparity gives an advantage to families that can spend time and money advocating on their children’s behalf, while potentially limiting equal access to education for low-income students of color.
AZCIR reporters identified data limitations and irregularities during the analysis, including duplicated free and reduced lunch figures and underreported 504 data. AZCIR consulted national researchers, state and federal officials, and contacted more than 100 school districts to validate and update the data.
The analysis included 1,275 traditional public schools in Arizona, which spanned 126 districts with a combined enrollment of 921,326 students, or 80% of all public school students in the state.
Interviews with dozens of experts, parents, school and district faculty, and disability attorneys informed AZCIR’s reporting. Reporters also reviewed federal laws to provide context about the civil rights protection, including how it is designed to help students learn and who is responsible for ensuring students have access to 504 services.
Help AZCIR do more to keep you informed.
Prefer to send a check? Mail it to AZCIR: PO Box 3665, Phoenix, AZ 85030
Why didn’t AZCIR include all Arizona schools in its analysis?
Exactly 2,000 federally-funded Arizona schools reported civil rights information to the U.S. Department of Education’s 2017-2018 civil rights data survey. The survey contains data for traditional public, charter, magnet, special education and alternative schools, as well as juvenile justice facilities.
School type determines whether data will be representative of Arizona’s public education system. At special education schools where most students have disabilities, for example, the use of 504 plans may look different than it does at a typical public school. The same is true for alternative schools where at-risk youth may require more 504 services, or juvenile justice facilities where children are learning while incarcerated.
AZCIR excluded 124 special education schools, alternative schools and juvenile justice facilities from its analysis.
Free and reduced lunch figures were not available for 15 schools, which were excluded because AZCIR couldn’t compare them to other schools in the analysis when using wealth as a disparity indicator.
AZCIR removed 480 charter schools from the analysis because reporters could not verify their data. A more complete explanation of those problems, and how AZCIR tried to address them, is detailed further below.
In districts with small student populations, a slight change in the number of 504 plans can cause a significant change in the 504 rate overall, potentially skewing AZCIR’s analysis. Reporters removed an additional 106 schools in districts with fewer than 500 students.
How did AZCIR incorporate student wealth?
State data showing the rates of students who are eligible to receive free and reduced school lunch served as a measure of student wealth in AZCIR’s analysis. Eligibility is based on household income, so the data represents a commonly used indicator of student poverty.
AZCIR placed each school into a wealth quartile based on the rate of students eligible for low-cost lunch.The wealthiest schools, for example, were grouped by schools with a free and reduced lunch rate less than 25%. Schools in which 75% or more of students qualified for low-cost lunch fell into the poorest quartile, and the middle quartiles had schools with rates of 25-49% and 50-74%.
Based on this grouping, the number of schools in each quartile were not evenly distributed. This matters because nearly two-thirds of the schools had a low-cost lunch rate greater than 50%, so the majority were placed in the poorest two groups.
Put another way, if AZCIR evenly distributed the schools among its quartile groupings, a school with a low cost lunch rate of 64% would have been classified as moderately wealthy.
AZCIR’s analysis used fixed low-cost lunch rates to create its wealth quartiles, so a school with a 64% low-cost lunch rate would be placed in the moderately poor group. This method allowed AZCIR to define “wealthy” and “poor” schools in a more accurate way.
How did AZCIR handle data discrepancies found in the federal data survey?
The U.S. Department of Education’s Civil Rights Data Survey shows a national 504 rate of 2.7%. Arizona’s rate—which includes all schools that reported to the federal survey—is 1.5%, meaning the statewide 504 rate was less than what AZCIR anticipated.
More than 700 traditional public and charter schools reported zero 504 plans to the CRDC. They had a combined student population of 275,749—about a quarter of the statewide enrollment—so AZCIR expected to find thousands of 504 plans based on the national rate.
AZCIR suspected the zero figures were reported in error. Reporters contacted 513 of those schools through their charter or traditional public school district offices using the following criteria:
- All districts in which every school reported zero 504 plans
- All districts with a student population of at least 500, and at least one school that reported zero 504 plans
Sixty-seven districts didn’t respond to AZCIR. Combined they represent 182 “zero 504 plan” schools. Data for those schools—when they met all other criteria for inclusion—was analyzed as originally published by the Office for Civil Rights.
Districts representing 213 schools, or 42% of schools contacted, said the data was incorrect and provided alternative figures. All other schools verified the data as accurate.
Hundreds of Arizona’s public schools failed to report accurate civil rights data to the U.S. Department of Education for the 2017-2018 school year, a violation that can result in a loss of federal funding.
The updated figures changed Arizona’s rate of 504 plans from 1.46% to 1.66%, and added 2,203 plans to the statewide total. Almost half of those 504 plans were added by schools in the state’s poorest quartile.
AZCIR’s verification was limited because the CRDC is a “snapshot” data collection, meaning districts report data from a specific date—October 1, 2017. The updated data may have been from a different date or period of time during the school year, so some rates could be different.
Why were charter schools excluded?
The majority of “zero 504” charter schools didn’t respond to AZCIR’s emails and phone calls to verify their 504 plan figures. Charters represent 210 of the 710 schools that reported zero 504 plans to the federal survey. AZCIR contacted 89 for verification.
More than two-thirds of traditional public schools responded to AZCIR follow-ups to verify data. For charter schools, just 45% responded, meaning reporters weren’t confident in drawing conclusions from the data.
An earlier version of AZCIR’s analysis did include the unverified charter school data, which showed similar trends in 504 plans when compared to those identified among the state’s traditional public schools. This indicates the same disparities may exist in charter schools. But because AZCIR could not validate those conclusions, they were removed from the overall analysis.
How was race used in AZCIR’s analysis?
Because the federal survey collects both 504 plan and enrollment data by student ethnicity, AZCIR was able to calculate the rate of 504 plans for specific racial groups. It also allowed reporters to calculate 504 rates by grouping schools into quartiles based on the percent share of white students enrolled—similar to how AZCIR analyzed the free and reduced lunch rates.
However, most districts that corrected the federal data during AZCIR’s verification process did not provide 504 figures by specific racial group. This limited the level of specificity in the analysis because rates could not be calculated for specific ethnic groups, meaning trends couldn’t be identified at an individual student level.
Schools were instead placed into quartiles based on the percent share of white students within their total student population. The grouping followed the same pattern as the wealth quartiles: The whitest quartile contained schools with a student population that was at least 75% white, and schools where students of color made up at least 75% of enrollment were placed into the “least white” quartile. The middle two quartiles had schools with white students representing 25-49% and 50-74% of enrollment.
The racial analysis also included a majority-white group that contained schools where at least 50% of the student population is white, and a majority-minority group where students of color comprise at least half of the student population.
Sign up. Get engaged.
Get AZCIR’s latest investigations delivered straight into your inbox.
Other problems AZCIR encountered with state data
The Arizona Department of Education conducts two separate data collections for free and reduced lunch. One is for schools that participate in the National School Lunch Program, a federal low-cost lunch program that doesn’t apply to all public schools. The other is a more general collection for all schools in the state.
State officials said the general data collection contained duplicate student counts, and that some schools may have only reported to the NSLP survey. Reporters obtained a de-duplicated version of the general collection, which covers the most schools, and incorporated the NSLP figures into its analysis in an effort to fix these issues. When the two data sets conflicted for a particular school, the greater of the two rates was used.
To protect student privacy, the education department redacts free and reduced lunch data when fewer than 11 students are eligible at any given school. AZCIR’s analysis treats redacted values as zeros, so some schools that show 0% low-cost lunch may have a slightly higher rate.