A journalist’s crash course in data reporting reveals troubling patterns in LA’s child welfare system

When I began researching the ways in which Los Angeles County’s Department of Children and Family Services (DCFS) intervenes in the lives of families, I had a general sense that what I would find would not reflect positively on the county’s child welfare system. This was partly because of what I learned while completing an earlier project through the Center for Health Journalism’s California Fellowship.

While writing that 2021 series, “Pregnant Behind Bars,” which focused on a local program diverting pregnant people from LA’s jails into permanent supportive housing, many of the mothers I interviewed were more anxious for me to know about their disastrous experiences with DCFS and LA’s family courts than about their time in jail.  

The reason one mother wound up in jail at all was due to the fact that she kidnapped her 6-year-old from foster care after he told her the foster parent’s son was sexually assaulting him. Another mother said that at the same time the diversion program was pulling her out of jail, a court hearing to terminate her parental rights for her oldest daughter was taking place. She said she was not notified of the hearing and thus did not get to defend her right to her child. 

Other mothers lost their children to foster care due to their incarceration, or after becoming homeless, or because of drug use.

Spending time with the women I met while researching the “Pregnant Behind Bars” series persuaded me that I needed to hear from other local families entangled with the child welfare system. Around the same time, a new group, the Reimagine Child Safety Coalition, sent the LA County Board of Supervisors a list of 11 demands for overhauling the child welfare system so that families are no longer “ripped apart.”

When I began this most recent series, “Punishing Families,” with the support of the Center for Health Journalism’s Impact Fund, I was clear that the project would center on the stories of impacted families. I knew I would also need a certain amount of data to balance what would be a very zoomed-in view of child welfare in LA County, where approximately 30,000 to 40,000 kids are housed in foster care each year. 

DCFS publishes monthly and annual reports indicating how many abuse and neglect-related investigations and removals have occurred county-wide. Those reports show that the largest portion of maltreatment accusations reported to DCFS fit under the broad category of “general neglect.” 

Many of the factors that have historically indicated “general neglect” — including unstable housing, and insufficient food — are conditions of poverty.

When I submitted a public records request to DCFS seeking deeper layers of data than the readily available basic stats, I asked for five years of neglect allegations and removal numbers, with breakdowns by age, race/ethnicity, gender, and ZIP code. I did not expect to use the ZIP code data, but thought it would be nice to have just in case I could pull out a few interesting facts. 

The series evolved to be much more data heavy than I originally planned, and the ZIP code files turned out to be the most critical component. 

When I began my Impact Fund project, I was not a data journalist. I had never even taken a college statistics class. Yet, when I looked at the numbers for a few specific ZIP codes, and compared them to Wikipedia population numbers, basic calculations revealed what appeared to be vast disparities between areas of the county I knew to be lower-income and ZIP codes known for their wealthy, white neighborhoods. 

Still, I wasn’t sure how to dig deeper. I kept poking around in the data without making much headway until the Center connected me with an experienced investigative data journalist. He generously shared advice on what to look for and how to compare my massive PDF of county-level data to official census records, as well as a dozen YouTube tutorials on relevant Excel functions. Soon, a clearer picture began to emerge. 

The data revealed that neighboring ZIP codes, even those in the same city, often had considerably different experiences with DCFS. If two adjacent ZIP codes had different median household incomes, families in the poorer ZIP code were much more likely to be entangled with the child welfare system. 

For example, there were double the number of general neglect allegations per capita in the 93534 ZIP code in Lancaster, where the median household income was $46,875, as there were in the city’s neighboring 93536, where families made an average of $89,987.

Kids in lower-income ZIP codes were also far more likely than kids in more affluent ZIP codes to be removed once investigated. In addition, DCFS data revealed that between 2018 and 2021, LA County’s social workers removed twice as many Black children from their homes as white children, despite the fact that Black children make up a much smaller share of the population.

Although I’d requested data from government agencies for earlier stories, “Punishing Families” was my first reporting project in which I combined multiple datasets and conducted in-depth analysis. It was daunting, but the information it allowed me to share with the community was well worth the challenge. 

When writing public records requests, it can be useful to ask for far more information than you think you’ll need or will be useful to you. Moreover, there’s often something important to be learned from the data you are not able to obtain.

In this case, DCFS was unable to provide certain metrics because the department did not collect the relevant information.

Learning that DCFS had no way to extract data on the specific reasons children were removed from their homes during neglect removals (i.e. parental drug use, inadequate food or clothing, leaving children unattended, living unsheltered, etc.) was revealing. How could the county see and address the underlying needs of parents and their children — and reduce their contact with the child welfare system in order to keep families together — without that data?