How I fought for data on the invisible kids falling through the cracks in one Augusta school district

I could not turn my eyes away, as children exited the homeless shelter every morning. I would watch them walk down the street as I waited in the car line, a block away, to drop my daughter off at school.

I was used to seeing homeless men and women leaving the shelter. This was different. Why so many children? Why aren’t they in school?

Through my reporting, I found Alessia. She was desperate. I met her inside her broken-down SUV where she was living with her four children. “My biggest fear … my kids … I do not want them to be away from me.” Alessia’s fear kept her from telling her son’s teacher the real reason he was missing school.

Another mother also worried about losing her children. Her son was sent to the alternative school but hasn’t been in class. Orientation is mandatory for parents but she’s a single working mom and orientation is only offered during work hours.

The Richmond County School District in Augusta, Georgia is a failing school system. It was rated "F" before the pandemic dried up affordable housing in the city. Most of the students are minorities from low-income families. I knew these children were standing on a fault line in education, just one eviction, flat tire, or missed payment from falling through the cracks.

Deceptive data

I started asking for data from the school district weeks before being accepted into the 2021 Data Fellowship. The district told me their homeless student population decreased by more than 65%. It didn’t add up. I had joined a homeless task force and knew homelessness had increased 150% in the county. The school district told me they had about 900 students unaccounted for. Could those be homeless kids not going to school? 

I was on the right path, but I had the wrong data. I sent a records request for the district’s McKinney Vento grant application to the state. The federal act mandates school systems to eliminate barriers that prevent homeless students from attending school. The grant documents showed the district was denying enrollment to students without a permanent address, losing referrals for homeless students, and not allocating enough resources for the transportation of homeless students without a ride to school.

Several sources contacted me after my first fellowship story aired. They confirmed my suspicion. The school district only gave me part of the data. I would have not known had it not been for my sources telling me the district was miscoding missing students as transfers or homeschoolers. The sources were able to guide me through the wording of another records request which revealed a much larger crisis.

The state of Georgia requires districts to enter a code whenever a student stop coming to school. There are about a dozen codes to explain why a student is no longer in attendance.  Some codes like transfers and homeschooling require documentation, such as a transcript request or Declaration of Intent to Homeschool. The state requires districts to use the code “Unknown” until documentation is received to prove the missing student is in school somewhere.

I learned the district uses the codes for two separate groups of students. No-shows are the kids who didn’t show up on the first day of class. Withdrawals are the kids who stop coming to class during the school year. I requested the total number of no-shows and withdrawals in each school under each code. I also included a request for the number of documents received under each code.

The district provided the data after several email exchanges where I had to include attachments to show them the regulations and data capabilities within their own program. The district provided the data, labeling the miscoded students as “in progress.” I found nearly 3,000 students without the mandated documentation. That’s 3,000 missing students unaccounted for but coded as being in school somewhere else. That’s three times the number of students the district first told us were unknown or missing.

Smoke and mirrors from administrators

Administrators at the school rebutted my findings in the data. It was a game of smoke and mirrors. They told me one group of students “should” have documentation, so the data was incorrect. I followed up with asking for documentation. They did not have documentation.  The district decided to offer an interview a few days before my original airdate. They gave me a sheet with “new” data at the interview. I pushed back the airdate to go over it. 

I had to take a picture and zoom in to read the tiny print under the data they gave me at the last minute. The fine print read the numbers had not been confirmed in their system. I previously had also pulled withdrawal audits from the state. This came in handy when the district spokesperson insisted they are following state regulations. I showed the audits where the state had previously found them noncompliant for the very same thing I had discovered in this project. 

The fallout for students

Miscoded students do not receive intervention or outreach. I was able to show the fallout through interviews with homeless mothers and an inmate search at the county jail. I downloaded the current list of inmates and sorted by age. I pulled the names of all the 17- and 18-year-olds and then requested their booking sheets to confirm their residency within the county. A source was able to run those names through the system to confirm the number of unaccounted for students. More than 90% of the 45 inmates sitting in the county jail are unaccounted-for students. Several of the inmates do not have addresses. Law enforcement confirmed they were homeless.

My findings were validated two weeks before the airdate. A former lead social worker filed a whistleblower lawsuit against the school district. She alleges she was demoted for speaking up about McKinney Vento violations. The lawsuit cites “at least 20 occasions on which she objected to the denial of free public education.” 

The project’s impact

The district began official partnerships with law enforcement and the homeless shelter to help identify homeless children and get them into school. The district put new procedures in place to ensure documentation is received before coding a student exiting the system. The assistant school superintendent overseeing student services and the data unexpectedly resigned after the district turned over the full data. The local state medical college is using data to focus on the gaps identified in a new educational hub for low-income students set to this summer.

This was a big project intensified by normal deadlines, mommy duties and a spouse who believed my laptop was permanently fused to my lap. I came into the project intimidated by data. I finished the project feeling confident and ready to tackle future data projects. The best feeling is seeing all of the hard work and long hours makes a difference in the community that I call home.

**